The Pseudomonas aeruginosa exoenzyme Y impairs endothelial cell proliferation and vascular repair following lung injury

Abstract

Exoenzyme Y (ExoY) is a Pseudomonas aeruginosa toxin that is introduced into host cells through the type 3 secretion system (T3SS). Once inside the host cell cytoplasm, ExoY generates cyclic nucleotides that cause tau phosphorylation and microtubule breakdown. Microtubule breakdown causes interendothelial cell gap formation and tissue edema. Although ExoY transiently induces interendothelial cell gap formation, it remains unclear whether ExoY prevents repair of the endothelial cell barrier. Here, we test the hypothesis that ExoY intoxication impairs recovery of the endothelial cell barrier following gap formation, decreasing migration, proliferation, and lung repair. Pulmonary microvascular endothelial cells (PMVECs) were infected with P. aeruginosa strains for 6 h, including one possessing an active ExoY (PA103 exoUexoT::Tc pUCPexoY; ExoY(+)), one with an inactive ExoY (PA103ΔexoUexoT::Tc pUCPexoY(K81M); ExoY(K81M)), and one that lacks PcrV required for a functional T3SS (ΔPcrV). ExoY(+) induced interendothelial cell gaps, whereas ExoY(K81M) and ΔPcrV did not promote gap formation. Following gap formation, bacteria were removed and endothelial cell repair was examined. PMVECs were unable to repair gaps even 3-5 days after infection. Serum-stimulated growth was greatly diminished following ExoY intoxication. Intratracheal inoculation of ExoY(+) and ExoY(K81M) caused severe pneumonia and acute lung injury. However, whereas the pulmonary endothelial cell barrier was functionally improved 1 wk following ExoY(K81M) infection, pulmonary endothelium was unable to restrict the hyperpermeability response to elevated hydrostatic pressure following ExoY(+) infection. In conclusion, ExoY is an edema factor that chronically impairs endothelial cell barrier integrity following lung injury.

Keywords: cyclase; microtubules; permeability; pulmonary edema; tau.

Fig. 1.

Exoenzyme Y (ExoY) induces interendothelial cell gaps and impairs barrier restoration. Pulmonary microvascular endothelial cells (PMVECs) were inoculated with ExoY⁺ and ExoY^K81M bacterial strains for 6 h at a 20:1 MOI. A: ExoY⁺-treated cells developed large gaps by 6 h, whereas ExoY^K81M-treated cells did not. B: after 6-h bacterial inoculations, medium containing bacteria was removed and replaced with fresh medium containing serum (10%) and antibiotics. Cells were maintained in the incubator for 72 h and images were captured. Interendothelial cells gaps remained visible in ExoY⁺-treated cells. C: after 6-h bacterial inoculations, cells were trypsinized, counted, and reseeded in 6-well plates at a density of 5 × 10⁵ cells per well with fresh medium containing serum (10%) and antibiotics. Cells were maintained in the incubator for 72 h and images were captured. ExoY⁺-treated cells failed to reach confluence with 72 h, whereas ExoY^K81M-treated cells grew to confluence over the same time period. Images in A, B, and C are each representative of 10 separate experiments. Arrowheads denote interendothelial cell gaps.

Fig. 2.

ExoY decreases endothelial cell proliferation. PMVECs were infected with ExoY⁺, ExoY^K81M, or ΔPcrV bacterial strains for 6 h, medium was removed, and adherent cells were trypsinized and resuspended in the presence of serum (10%) and antibiotics into 6-well plates for a standard growth curve. A: pictures were taken every 24 h for 6 days to visualize growth. B: cells were trypsinized and counted at 24-h intervals for a week. Whereas both ΔPcrV- and ExoY^K81M-treated cells grew to confluence, ExoY⁺-treated cells did not enter log phase growth and did not grow to confluence. Data represent means ± SE taken from 5 separate infections and growth curves. *P < 0.05 vs. ExoY⁺-treated, ^P < 0.05 vs. day 0 of the growth curve.

Fig. 3.

ExoY⁺-infected PMVECs do not display chronic elevations in cyclic nucleotide concentrations. Cells were infected with ExoY⁺, ExoY^K81M, or ΔPcrV for 6 h, medium was removed, and the adherent cells were trypsinized and reseeded onto 6-well plates in the presence of serum (10%) and antibiotics. After 3 and 4 days, cells were lysed and prepared for radioimmunoassay. PMVEC cAMP concentrations were not different among treatment groups at either 3 or 4 days. P = not significant (ns).

Fig. 4.

ExoY⁺ causes pneumonia and acute lung injury and hinders pulmonary vascular repair. ExoY⁺ and ExoY^K81M were delivered through the trachea, and lungs were fixed 24 and 72 h postinfection. A: ExoY⁺-treated lungs developed alveolar flooding (*) with large perivascular cuffs (^) within 24 h, accompanied by profound alveolar neutrophil infiltration (∇), hemorrhage (◇), and areas of consolidation (∫) and atelectasis (top). By 72 h, considerable lung repair was evident, with limited persistence of edema (bottom). Neutrophils were largely cleared from the airways. Red blood cells remained visible in some alveolar segments. B: ExoY^K81M-treated lungs displayed significant alveolar neutrophil infiltration and areas of consolidation and atelectasis. However, extensive alveolar fluid and perivascular cuffs were not seen, and alveolar hemorrhage was not prominent. By 72 h, considerable lung repair was also evident.

Fig. 5.

ExoY prevents normal repair of the endothelial cell barrier. ExoY⁺ and ExoY^K81M were delivered through the trachea. Infected animals were allowed to recover for 7 days, at which time isolated perfused lung studies were performed. A: once baseline perfusion and pressures were established, flow (cardiac output) was increased sequentially in 5-min intervals. Pulmonary artery (top left; P_A), vein (top right; P_V) and double occlusion (bottom left; P_DO) pressures were similar in uninfected controls (n = 3) and ExoY⁺- (n = 6) and ExoY^K81M (n = 5)-infected animals over the whole range of flows. However, ExoY⁺-treated animals gained lung weight at relatively low cardiac outputs (bottom right), with significant increases in these responders evident by flows of 36 ml/min (P < 0.05 vs. uninfected control and ExoY^K81M). B: lungs from ExoY⁺-treated animals displayed clear evidence of edema on gross inspection following the isolated perfused lung experiment (arrow), whereas in ExoY^K81M-treated lungs edema was not characteristically seen.

Fig. 6.

ExoY⁺ causes a sustained increase in endothelial permeability, as determined by fluid filtration coefficient (K_f). ExoY⁺ and ExoY^K81M were delivered through the trachea. Infected animals were allowed to recover for 7 days, at which time isolated perfused lung studies were performed. A: lungs from control animals displayed low K_f values, whereas lungs from ExoY^K81M-treated animals possessed variable permeability responses that were on average not different from controls (top left; P = ns). In contrast, K_f was significantly increased in the lungs of ExoY⁺-treated animals (P < 0.05). The response of individual K_f measurements is shown along with the average ± SE. Pulmonary artery (top right), double occlusion (bottom left), and venous (bottom right) pressures were not different among groups during the isolated perfused lung experiments (P = ns). B: albumin-conjugated Evans blue retained in isolated perfused lungs reflected K_f measurements. Little Evans blue was seen in control lungs. In ExoY^K81M-treated animals, lungs that demonstrated low K_f also had little Evans blue dye retained in the lung, whereas those lungs that displayed high K_f showed marked Evans blue dye accumulation. ExoY⁺ treated lungs uniformly displayed extensive Evans blue dye staining. C: lungs were formalin fixed for hematoxylin and eosin staining following the K_f experiment. Histology sections revealed open airways in control animals, with visible accumulation of fluid in perivascular cuffs. Lungs of ExoY^K81M-treated animals were remarkable for their open airways but displayed interstitial cellular infiltration along with perivascular cuffs. Lungs from ExoY⁺-treated animals also had significant cellular infiltration, with evidence of diffuse alveolar damage (arrow).