Pseudomonas aeruginosa ExoS Induces Intrinsic Apoptosis in Target Host Cells in a Manner That is Dependent on its GAP Domain Activity

Abstract

Pseudomonas aeruginosa is a Gram-negative opportunistic pathogen that causes serious infections in immunocompromised individuals and cystic fibrosis patients. ExoS and ExoT are two homologous bifunctional Type III Secretion System (T3SS) virulence factors that induce apoptosis in target host cells. They possess a GTPase Activating Protein (GAP) domain at their N-termini, which share ~76% homology, and an ADP-ribosyltransferase (ADPRT) domain at their C-termini, which target non-overlapping substrates. Both the GAP and the ADPRT domains contribute to ExoT's cytotoxicity in target epithelial cells, whereas, ExoS-induced apoptosis is reported to be primarily due to its ADPRT domain. In this report, we demonstrate that ExoS/GAP domain is both necessary and sufficient to induce mitochondrial apoptosis. Our data demonstrate that intoxication with ExoS/GAP domain leads to enrichment of Bax and Bim into the mitochondrial outer-membrane, disruption of mitochondrial membrane and release of and cytochrome c into the cytosol, which activates initiator caspase-9 and effector caspase-3, that executes cellular death. We posit that the contribution of the GAP domain in ExoS-induced apoptosis was overlooked in prior studies due to its slower kinetics of cytotoxicity as compared to ADPRT. Our data clarify the field and reveal a novel virulence function for ExoS/GAP as an inducer of apoptosis.

Figure 1

ExoS-induced mitochondrial membrane disruption and cytochrome c release into the cytosol is dependent on its GAP domain activity. HeLa cells were infected with ExoS-expressing ∆U∆T/ExoS, ExoS/GAP-expressing ∆U∆T/ExoS (G⁺A⁻), ExoS/ADPRT-expressing ∆U∆T/ExoS (G⁻A⁺) strain, or the T3SS mutant (pscJ) at a MOI of 10. (A) Five hours following infection, cells were fixed and stained with DAPI nuclear dye (blue) and cytochrome c (red) to inspect the impact of the ExoS domains on mitochondria. Representative images are shown. Scale bar represents 10 μm. (B,C) Columbus software was used to determine the number of intact mitochondria (defined as cytochrome c positive punctate structures and represented by different colored circles in these images) per cell. Representative images are shown in (B) and the corresponding data are shown in (C). Results were tabulated from 5 random fields, compared to Mock, and presented as the Mean ± SD (****p < 0.0001, One-way ANOVA). (D) The cytoplasmic fractions of the aforementioned infected HeLa cells were assessed for their cytochrome c contents by Western blotting, following a 5 h infection as described above. GAPDH and Cox IV were used as loading controls for cytoplasmic and mitochondrial fractions respectively (Equal amounts of proteins were loaded on 3 gels and run simultaneously. The gels were then probed with either cytochrome c, GAPDH, or Cox IV. Each experiment was repeated at least 3 times and a representative blot of each is shown).

Figure 2

Intoxication with ExoS/GAP results in upregulation and mobilization of Bcl-2 pro-apoptotic proteins to the mitochondrial membrane. HeLa cells were treated with PBS (Mock) or infected with ExoS/GAP-expressing ∆U∆T/ExoS (G⁺A⁻) or the T3SS mutant (pscJ) at a MOI of 10. (A) 5 h following infection, the cells were fixed and analyzed by IF microscopy after staining for Bax (green), the mitochondrial marker MitoTracker (red), and DAPI nuclear stain (blue). Scale bar represents 20 μm. The arrow points to a Bax globular structure at the mitochondria. (B) Bax expression was assessed by determining the mean fluorescent intensity (MFI) from 10 random fields and are presented as the Mean ± SD (***p < 0.001, One-way ANOVA, ns = not significant). (C) 5 h following infection, levels of pro-apoptotic proteins Bax and Bim, and anti-apoptotic proteins Bcl-2 and Bcl-_XL, in cytosolic and mitochondrial fractions were determined by Western blotting. GAPDH and Cox IV were used as loading controls for cytoplasmic and mitochondrial fractions respectively (Equal amounts of proteins were loaded on different gels. The gels were then probed with either Bax, Bim, Bcl-2, Bcl-XL, GAPDH, or Cox IV. Each experiment was repeated at least 3 times and a representative image is shown).

Figure 3

Intoxication with ExoS/GAP domain results in caspase-9 activation. (A) HeLa cells were treated with PBS (Mock) or infected with ExoS/GAP-expressing ∆U∆T/ExoS (G⁺A⁻) or the T3SS mutant (pscJ) at an MOI of 10. Following a 5 h infection, HeLa cells were harvested and probed for active caspase-9 by Western blotting. GAPDH was used as a loading control (Equal amounts of proteins were loaded on 2 different gels and probed for either caspase-9 or GAPDH. Each experiment was repeated at least 3 times and a representative blot of each is shown). (B) HeLa cells were transfected with pIRES2-EGFP mammalian expression control vector (pGFP) or vectors containing full length and truncated GAP-expressing ExoS (pExoS (G⁺A⁻) or pExoS (G⁺)), or their inactive GAP counterparts (pExoS (G⁻A⁻) or pExoS (G⁻)), all directly fused to GFP at their C-termini. 20 h following transfection, cells were fixed and probed for nucleus (blue), gene of interest (GFP), and active caspase-9 (red) by IF microscopy. Note that infection or transfection with functional ExoS/GAP leads to caspase-9 activation. Scale bar represents 20 μm.

Figure 4

Intoxication with ExoS/GAP domain results in caspase-3 activation. (A) HeLa cells were treated with PBS (Mock) or infected with indicated strains as described in Fig. 3 legend. Following a 5 h infection, HeLa cells were harvested and probed for active caspase-3 by Western blotting. GAPDH was used as a loading control. (Equal amounts of proteins were loaded on 2 different gels and probed for either caspase-3 or GAPDH. Each experiment was repeated at least 3 times and a representative blot of each is shown). (B) HeLa cells were transfected with pIRES2-EGFP mammalian expression control vector (pGFP) or the vectors containing full length and truncated GAP-expressing ExoS (pExoS (G⁺A⁻) or pExoS (G⁺)), or their inactive GAP counterparts (pExoS (G⁻A⁻) or pExoS (G⁻)), all directly fused to GFP at their c-termini. 20 h following transfection, cells were fixed and probed for nucleus (blue), gene of interest (GFP), and active caspase-3 (red) by IF microscopy. Note that infection or transfection with functional ExoS/GAP leads to caspase-3 activation. Scale bar represents 20 μm.

Figure 5

Caspase-9 and caspase-3 activities are required to mediate ExoS/GAP-induced cytotoxicity. HeLa cells were infected with ExoS/GAP-expressing ∆U∆T/ExoS (G⁺A⁻) or the T3SS mutant (pscJ) at a MOI of 10 in the absence or presence of caspase-9 inhibitor (Z-LEHD-FMK) or caspase-3 inhibitor (Z-DEVD-FMK). (A) Cytotoxicity was assessed by time-lapse videomicroscopy using propidium iodide (PI) as a marker for cell death, as described^,. Images were taken every 15 min for 20 h. Representative images from videos are shown. (B) Cytotoxicity was assessed by determining mean fluorescent intensity (MFI) of PI from 3 independent experiments and are presented as the Mean ± SD. Range of significant p-values (as determined by One-way ANOVA), are indicated by arrows.

Figure 6

ExoS/GAP domain is sufficient in inducing apoptosis in target host cells. HeLa cells were transfected with pIRES2-EGFP mammalian expression control vector (pGFP) or vectors containing full length and truncated GAP-expressing ExoS (pExoS (G⁺A⁻) or pExoS (G⁺)), or their inactive GAP counterparts (pExoS (G⁻A⁻) or pExoS (G⁻)), all directly fused at their C-termini to GFP. Cytotoxicity was analyzed by IF time-lapse videomicroscopy using PI uptake as a marker for cell death. Images were taken every 15 min. (A) Representative images from videos are shown. (−1H) timepoint shows the transfected cells 1 h prior to GFP expression and 0 H is the moment when GFP is expressed. Note that these times are different from the actual timestamps in the movies but were used to normalize data with respect to transfection and cell death to show ExoS/GAP kinetics of cytotoxicity. (B) Cytotoxicity was measured as previously described and the data are presented as the Mean ± SD (ns = not significant, *p < 0.01, Student’s t-test).

Figure 7

Dynamics of cytotoxicity associated with ExoS and its ADPRT and GAP domains in PA103 strain genetic background. HeLa cells were infected with ExoS-expressing ∆U∆T/ExoS, ExoS/GAP-expressing ∆U∆T/ExoS (G⁺A⁻), ExoS/ADPRT-expressing ∆U∆T/ExoS (G⁻A⁺) strain; or the T3SS mutant (pscJ) at a MOI of 10. (A) Cytotoxicity was observed by IF time-lapse videomicroscopy using PI uptake as a marker for cell death. Images were taken every 15 min for 20 h. Representative images from videos at indicated times are shown. (B) Cytotoxicity was assessed by determining mean fluorescent intensity (MFI) of PI from 3 independent experiments and are presented as the Mean ± SD (****p ≤ 0.0001, ***p ≤ 0.001, **p ≤ 0.01, *p ≤ 0.05, ns = not significant, determined by One-way ANOVA).

Figure 8

Dynamics of cytotoxicity associated with ExoS and its ADPRT and GAP domains in PAK genetic background. HeLa cells were infected with PAK∆S∆T strains expressing ExoS (∆S∆T/ExoS), ExoS/GAP-expressing ∆S∆T/ExoS (G⁺A⁻), ExoS/ADPRT-expressing ∆S∆T/ExoS (G⁻A⁺) strain; or the T3SS mutant (pscJ) at a MOI of 10. (A) Cytotoxicity was observed by IF time-lapse videomicroscopy using PI uptake as a marker for cell death. Images were taken every 15 min for 20 h. Representative images from videos are shown (Red cells are dead). (B) Cytotoxicity was assessed by determining mean fluorescent intensity (MFI) of PI from 3 independent experiments and are presented as the Mean ± SD (****p < 0.0001, ***p ≤ 0.001, **p ≤ 0.01, *p ≤ 0.05, ns = not significant, determined by by One-way ANOVA). (C) 5 h following infection with indicated strains, cells were fractionated and the cytosolic fractions were evaluated for their cytochrome c contents by Western blotting using an antibody against cytochrome c. GAPDH and Cox IV were used as loading controls for cytoplasmic and mitochondrial fractions respectively (Equal amounts of proteins were loaded on 3 gels and run simultaneously. The gels were then probed with either cytochrome c, GAPDH, or Cox IV. Each experiment was repeated at least 3 times and a representative blot of each is shown).