Pseudomonas aeruginosa Effector ExoS Inhibits ROS Production in Human Neutrophils

Abstract

Neutrophils are the first line of defense against bacterial infections, and the generation of reactive oxygen species is a key part of their arsenal. Pathogens use detoxification systems to avoid the bactericidal effects of reactive oxygen species. Here we demonstrate that the Gram-negative pathogen Pseudomonas aeruginosa is susceptible to reactive oxygen species but actively blocks the reactive oxygen species burst using two type III secreted effector proteins, ExoS and ExoT. ExoS ADP-ribosylates Ras and prevents it from interacting with and activating phosphoinositol-3-kinase (PI3K), which is required to stimulate the phagocytic NADPH-oxidase that generates reactive oxygen species. ExoT also affects PI3K signaling via its ADP-ribosyltransferase activity but does not act directly on Ras. A non-ribosylatable version of Ras restores reactive oxygen species production and results in increased bacterial killing. These findings demonstrate that subversion of the host innate immune response requires ExoS-mediated ADP-ribosylation of Ras in neutrophils.

Keywords: T3SS; effector function; keratitis; type III secretion.

Figure 1. NADPH oxidase mediates ROS production by neutrophils and facilitates clearance of P. aeruginosa during bacterial keratitis

(A) Representative images of corneal opacification 24 h post-infection of C57BL/6 and gp91^phox −/− (CGD) mice infected with 1x10⁵ CFU PAO1 (WT) or with the ΔpscD (T3SS null) mutant strain. (B) Quantification of corneal opacity by determining average pixel intensity of corneas described previously (Sun et al., 2012)(n = 5 mice). (C) Colony forming units (CFU) recovered from infected corneas 24 h post-infection (n = 9 mice). (D) Corneal sections were stained with hematoxylin and eosin, or (E) 4’,6-Diamidino-2-Phenylindole dye (DAPI, Blue) and an antibody to Ly6G (NIMP-R14, FITC, green) (Epi: epithelium, Str: stroma, End: corneal endothelium, AC: anterior chamber). B, C: Data points represent individual corneas. Median and interquartile range are indicated. Significance was calculated using the Kruskal-Wallis test, with Dunn’s multiple comparison correction. * p<0.05, **p<0.01, ***p<0.001, or n.s., not significant. See also Figure S1.

Figure 2. ExoS and ExoT ADPRT activities inhibit ROS production in human neutrophils

ROS production was measured using a chemiluminescent substrate (relative light units, RLU). Neutrophils were infected with: wild type (PAO1), a T3SS null mutant (ΔpscD), as well as: (A) a strain lacking the translocation apparatus (ΔpopBD), a strain lacking all 3 effectors (Δ3TOX), and a strain lacking exoS and exoT effector genes (ΔexoST). (B) a strain lacking exoS (ΔexoS), a strain lacking exoT(ΔexoT), or (C) strains with chromosomal point mutations inactivating the Rho-GAP (G-) or ADP-ribosyltransferase activities (A-) of ExoS and/or ExoT. A time course representative of at least three independent experiments is shown. See also Figure S2.

Figure 3. ExoS and ExoT ADP-ribosyltransferase activities interfere with PI3K signaling in neutrophils

(A) Cell lysates from uninfected human neutrophils, or from neutrophils infected with wild type PAO1, or a ΔpscD mutant strain were probed by Western blot for Akt, P-Akt (Thr308), p40^phox, P-p40^phox (Thr154), and Ras. The experiments were repeated 3 times with similar results. (B) Cell lysates from uninfected human neutrophils, or neutrophils infected for 30 minutes with PAO1, ΔpscD, or with strains in which the ADP-ribosyltransferase activity was inactivated in ExoS only (exoS(A-)), ExoT only (exoT(A-)), or both (exoST(A-)). P-Akt (Thr308), Akt, P-p40^phox (Thr154), p40^phox, total Ras, GTP-bound Ras, and Grb2 (loading control) were detected by western blot. The experiments were repeated 5 times with similar results. (C) Model of Ras (gray) bound to the Ras-binding domain (light blue) of PI3K (dark blue) based on the structure PDB:1he8 (Pacold et al., 2000). Residue Arg 41 of Ras is highlighted red. (D) Purified ExoS was used to ADP-ribosylate HA-tagged versions of Ras, or Ras(R41K), in vitro and subsequently mixed with purified PI3Kγ. The interaction between Ras and PI3Kγ was probed by immunoprecipitating Ras using an anti-HA-tag antibody. PI3Kγ, as well as unmodified and ADP-ribosylated Ras were detected by western blot. The experiments were repeated 3 times with similar results. Input and output levels of PI3Kγ were determined by densitometry. The input/output ratio for the untreated control sample was set to 100% and compared to the corresponding ExoS-treatment condition (mean and standard deviation of three independent replicates are noted below each lane). Results were compared by 1-way ANOVA with Bonferroni correction (**** p<0.0001). See also Figures S3.

Figure 4. Tat-Ras(R41K) rescues ROS production in human neutrophils, resulting in increased killing of P. aeruginosa.

(A) Human neutrophils were treated with increasing concentrations of Tat-Ras(R41K) for 30 minutes, and extracellular protein was degraded by proteinase K. Western blots of cell lysates were probed with antibodies to Ras (which detect endogenous and Tat-Ras(R41K)) or the HA tag (which detects only Tat-Ras(R41K)). (B) ROS production by human neutrophils infected with PAO1, exoT(A-), or exoST(A-) P. aeruginosa was measured by chemiluminescence. Neutrophils were incubated with increasing amounts of Tat-Ras (R41K) thirty minutes prior to infection with an exoT(A-) strain. The experiment was repeated 4 times with similar results. (C) Human neutrophils were incubated with Tat-Ras(R41K) (red) or Tat-Ras (blue) (3 μM final concentration) for 30 minutes prior to infection with PAO1, ΔpscD, exoS(A-), or exoST(A) 15 min (MOI30). Extracellular bacteria were killed with gentamicin for 30 minutes. Each point represents an individual human donor (n = 4 donors). Statistical significance was measured by 1-way ANOVA with Bonferroni correction. **p<0.01, ***p<0.001, n.s. not significant. See also Figure S4.