ExoU modulates soluble and membrane-bound ICAM-1 in Pseudomonas aeruginosa-infected endothelial cells.

ExoU, a Pseudomonas aeruginosa cytotoxin injected via the type III secretion system into host cells, possesses eicosanoid-mediated proinflammatory properties due to its phospholipase A(2) (PLA(2)) activity. This report addressed the question whether ExoU may modulate the expression of adhesion molecules in host cells, therefore contributing to the recruitment of leukocyte into infected tissues. ExoU was shown to down-regulate membrane-bound ICAM-1 (mICAM-1) and up-regulate the release of soluble ICAM-1 (sICAM-1) from P. aeruginosa-infected endothelial cells. The modulation of ICAM-1 depended on the direct effect of the ExoU PLA(2) activity and involved the cyclooxygenase (COX) pathway. No differences in mICAM-1 and sICAM-1 mRNA levels were observed when cultures were infected with the ExoU-producing PA103 strain or the mutant PA103DeltaexoU, suggesting that ExoU may proteolytically cleave mICAM-1, producing sICAM-1 in a COX-dependent pathway.

ExoU-induced vascular hyperpermeability and platelet activation in the course of experimental Pseudomonas aeruginosa pneumosepsis.

To address the question whether ExoU, a Pseudomonas aeruginosa cytotoxin with phospholipase A2 activity, can induce hemostatic abnormalities during the course of pneumosepsis, mice were instilled i.t. with the ExoU-producing PA103 P. aeruginosa or with a mutant obtained by deletion of the exoU gene. Control animals were instilled with sterile vehicle. To assess the role of ExoU in animal survival, mice were evaluated for 72 h. In all the other experiments, animals were studied at 24 h after infection. PA103-infected mice showed significantly higher mortality rate, lower blood leukocyte concentration, and higher platelet concentration and hematocrit than animals infected with the bacterial mutant, as well as evidences of increased vascular permeability and plasma leakage, which were confirmed by our finding of higher protein concentration in bronchoalveolar lavage fluids and by the Evans blue dye assay. Platelets from PA103-infected mice demonstrated features of activation, assessed by the flow cytometric detection of higher percentage of P-selectin expression and of platelet-derived microparticles as well as by the enzyme immunoassay detection of increased thromboxane A2 concentration in animal plasma. Histopathology of lung and kidney sections from PA103-infected mice exhibited evidences of thrombus formation that were not detected in sections of animals from the other groups. Our results demonstrate the ability of ExoU to induce vascular hyperpermeability, platelet activation, and thrombus formation during P. aeruginosa pneumosepsis, and we speculate that this ability may contribute to the reported poor outcome of patients with severe infection by ExoU-producing P. aeruginosa.

Lipid body mobilization in the ExoU-induced release of inflammatory mediators by airway epithelial cells.

This report addressed the question whether ExoU stimulation of airway epithelial cells may contribute to the inflammatory response detected in the course of Pseudomonas aeruginosa respiratory infections. Infection with PA103 P. aeruginosa elicited a potent release of IL-6 and IL-8, as well as of arachidonic acid (AA) and PGE(2) that was reduced by the bacterial treatment with MAFP, a cPLA(2) inhibitor. Airway cells from the BEAS-2B line and in primary culture were shown to be enriched in lipid bodies (LBs), that are cytoplasmic domains implicated in AA transformation into eicosanoids. However, cells infected with PA103 and with a mutant deficient in exoU but complemented with a functional gene exhibited reduced contents of LBs, and this reduction was inhibited by MAFP. FACS analysis showed that the decrease in the LB content correlated with the presence of intracellular PGE(2). Also, in PA103-infected cells, PGE(2) was immunolocalized in LBs, suggesting that the reduction in the cell content of the organelles was due to consumption of their glycerolipids, resulting in local synthesis of the prostanoid. In conclusion, we showed the ExoU ability to induce airway epithelial cells to overproduce PGE(2) and we speculate that LB may represent intracellular loci involved in ExoU-induced eicosanoid synthesis.

Implications of oxidative stress in the cytotoxicity of Pseudomonas aeruginosa ExoU.

ExoU PLA2-like activity has been shown to account for membrane lysis and acute death of infected cells. Translocation of effector proteins by the type III secretion systems depends on close contact between microbial and host cells. Our finding that both the ExoU-producing PA103 Pseudomonas aeruginosa and its mutant obtained by deletion of exoU adhered poorly to endothelial cells (EC) led to the hypothesis that, in some cells, the amount of injected toxin may not be enough to induce cell lysis but cells would suffer from a long-term effect of ExoU intoxication. To address this question, cells were exposed to both bacteria for 1 h and then treated with gentamicin-containing medium, to eliminate infecting microorganisms. After 24 h, the percentage of viable EC in PA103-infected cultures was significantly lower than in cultures exposed to the mutant, as determined by the MTT assay. Cell death was not likely to depend on the ExoU lytic activity since cell labeling with propidium iodide was similar in cultures infected with both bacterial strains. Bacterial cytotoxicity was significantly reduced by MAFP, a specific inhibitor of cPLA2 and iPLA2. Since the PLA2 activity on membrane phospholipids generates free fatty acid, including arachidonic acid (AA), we next compared the bacterial ability to release AA from infected EC. PA103 was shown to induce a potent AA release that was inhibited by MAFP. AA oxidation by oxygenases generates eicosanoids, known to induce both cell death and proliferation. However neither inhibitors of cyclooxygenases (ibuprofen) nor lipoxygenases (NDGA) reduced the ExoU toxicity. Since non-enzymatic oxidation of AA generates reactive radicals, we next investigated the PA103 ability to induce oxidative stress in infected cells. FACS analysis of cell labeling with the C-11 fluor probe and with anti-4-hydroxynonel antibody revealed a significant peroxidation of cell membrane lipids. These results, together with our finding that PA103-infected EC death was significantly attenuated by alpha-tocopherol, led to the conclusion that AA-induced oxidative stress may be another mechanism of cell damage in the course of infection by ExoU-producing P. aeruginosa.

Pseudomonas aeruginosa-induced production of free radicals by IFNgamma plus TNFalpha-activated human endothelial cells: mechanism of host defense or of bacterial pathogenesis?

We have previously shown that human umbilical vein endothelial cells (HUVEC) can be activated by IFNgamma plus TNFalpha to kill intracellular (IC) Pseudomonas aeruginosa through production of reactive oxygen intermediate, but the cumulative effects of cytokine activation and bacterial infection on host cells has not been extensively addressed. In this study we investigated the fate of IFNgamma plus TNFalpha-activated HUVEC that have harboured IC bacteria for up to 24 h. At 10 h, the endothelial cell killing of P. aeruginosa isolates exceeded 90%. IC bacteria enhanced the expression of inducible nitric oxide synthase (iNOS) and induced overproduction of NO and superoxide by infected HUVEC. P. aeruginosa IC infection also induced a slight decrease in the cellular level of reduced glutathione (GSH). Overproduction of NO correlated with a marked peroxidation of plasma membrane lipids and decline in HUVEC viability. Treatment of cells with the antioxidant alpha-lipoic acid significantly increased the survival of infected cells. Our data suggest that with the failure of adequate scavenger mechanisms, oxidant radicals overproduced in response to bacterial infection were highly toxic to host cells. Therefore, instead of contributing to defence against infectious agents, the upregulation of free radicals production by endothelial cells in response to cytokine activation would be detrimental to the host.

Dynamic interaction between airway epithelial cells and Staphylococcus aureus.

Staphylococcus aureus is a major cause of pulmonary infection, particularly in cystic fibrosis (CF) patients. However, few aspects of the interplay between S. aureus and host airway epithelial cells have been investigated thus far. We investigated by videomicroscopy the time- and bacterial concentration-dependent (10(4), 10(6), and 10(8) CFU/ml) effect of S. aureus on adherence, internalization, and the associated damage of the airway epithelial cells. The balance between the secretion by S. aureus of the alpha-toxin virulence factor and by the airway cells of the antibacterial secretory leukoproteinase inhibitor (SLPI) was also analyzed. After 1 h of interaction, whatever the initial bacterial concentration, a low percentage of S. aureus (<8%) adhered to airway cells, and no airway epithelial cell damage was observed. In contrast, after 24 h of incubation, more bacteria adhered to airway epithelial cells, internalized bacteria were observed, and a bacterial concentration-dependent effect on airway cell damage was observed. At 24 h, most airway cells incubated with bacteria at 10(8) CFU/ml exhibited a necrotic phenotype. The necrosis was preceded by a transient apoptotic process. In parallel, we observed a time- and bacterial concentration-dependent decrease in SLPI and increase in alpha-toxin expression. These results suggest that airway cells can defend against S. aureus in the early stages of infection. However, in later phases, there is a marked imbalance between the bactericidal capacity of host cells and bacterial virulence. These findings reinforce the potential importance of S. aureus in the pathogenicity of airway infections, including those observed early in CF patients.