Fig 2: Re-infection by non-motile bacteria is inhibited by Shigella intracellular replication ASchematic representation of the experimental design for re-infection experiments. Cells were infected with Shigella-mCherry (primary infection) or mock-treated. At 3 hpi, cells were re-infected with Shigella or Salmonella expressing GFP (secondary infection). Secondary infection was analyzed at 0.5 hpi (i.e., 1 h after addition of bacteria to cells).B–ERepresentative images (B, D) and cfu quantification (C, E) of re-infection assays with Shigella WT (B, C) or Salmonella WT (D, E) of HeLa cells primarily infected with Shigella WT or ?ipaB/Inv mutant strain, or mock-treated. Secondary infection is shown in bottom images.F, GRepresentative images (F) and cfu quantification (G) of the re-infection assays with Salmonella ?fliC, ?fliC/pFliC, or ?flhC mutant strains in HeLa cells primarily infected with Shigella WT or mock-treated. Secondary infection is shown in bottom images.HSchematic representation of the experimental design for re-infections upon knockdown or inhibitor treatment.I, JRepresentative images (I) and cfu quantification (J) of the secondary infection with Shigella WT, in cells transfected with ASM siRNA or control siRNA.K–MCfu quantification of the secondary infection with Shigella WT in HeLa cells following amitriptyline (K), transfection with MAPK p38 siRNA (L), or SB203580 treatment (M).Data information: Infection was performed at MOI 10 for Shigella WT, MOI 350 for Shigella ?ipaB/Inv, MOI 25 for Salmonella WT (D, E, G), or MOI 50 for Salmonella WT, ?fliC, ?fliC/pFliC, or ?flhC mutant strains (F, G). Results are shown as mean ± s.e.m. of five independent experiments; *P < 0.05, **P < 0.01, ***P < 0.001 (one-way ANOVA for C, E, J-M; two-way ANOVA for G). Scale bars, 50 µm.

Fig 3: Inhibition of Shigella infection upon host cell stress is a consequence of ASM activation A–FRepresentative images (A, D), cfu (B, E), and qRT–PCR (C, F) quantification of intracellular Shigella in HeLa cells pre-treated with arsenite or anisomycin, in the presence or not of amitriptyline, and corresponding controls.G–IRepresentative images (G), cfu (H), and qRT–PCR (I) quantification of intracellular Shigella in HeLa cells transfected with ASM siRNA or control siRNA and pre-treated or not with arsenite prior to infection.J–M3D reconstruction of representative images of ASM (J, K) or ceramide (L, M) staining in HeLa cells treated with arsenite or anisomycin and corresponding controls. ASM, ceramide, and Hoechst staining were surface-converted by voxel distance.Data information: Infection was performed with Shigella WT at MOI 10 and analyzed at 0.5 hpi. Results are shown as mean ± s.e.m. of 5 (H, I) or 7 (B–F) independent experiments; *P < 0.05, **P < 0.01, ***P < 0.001 (one-way ANOVA). Scale bars, 50 µm (A, D, G) and 5 µm (J–M).

Fig 4: Intracellular Shigella replication induces remodeling of the host cell plasma membrane in vitro and in vivo A, B3D reconstruction of representative images of ASM (A) or ceramide (B) staining in HeLa cells infected with Shigella WT or mock-treated, analyzed at 3 hpi. ASM, ceramide, Hoechst, and Shigella staining were surface-converted by voxel distance.C, DASM enzymatic activity (C) and ceramide (D) measurement in HeLa cells infected with Shigella WT or mock-treated, analyzed at 3 hpi. The ASM enzymatic activity was determined in the membrane fraction corresponding to 3.0 × 105 cells per condition. Ceramide is shown normalized to mock-treated cells.E–HRepresentative spinning-disk confocal images of colon tissue sections of non-infected (E and F) and Shigella-infected (G and H) guinea pigs, collected at 4 hpi. Panels (E and G) are montages of multiple image fields; dashed boxes in (E and G) are shown enlarged in (F and H), respectively.IMean fluorescence intensity of ceramide signal in surface epithelial cells (outermost tissue layer) of non-infected or Shigella-infected guinea pig colon tissue sections.JModel depicting the effect of stress-induced host plasma membrane remodeling on infection by bacterial pathogens. Different stressors, including TNF-a stimulation and Shigella intracellular replication, induce the relocalization and activation of acid sphingomyelinase (ASM) to the host plasma membrane, leading to the conversion of sphingomyelin into ceramide. As a consequence of membrane remodeling, the binding of non-motile bacteria, such as Shigella, to host cells is inhibited. Motile bacteria, such as Salmonella, are able to scan the host membrane for the remaining permissive binding sites, and thus circumvent this defense mechanism.Data information: Shigella infection was performed at MOI 100 (A, B). Results are shown as mean ± s.e.m. of 9 (C) or 6 (D) independent experiments; whiskers in the box-plot (I) correspond to min/max from 10 (non-infected) or 13 (Shigella) tissue sections. *P < 0.05, **P < 0.01, ***P < 0.001 (one-way ANOVA for C, D; unpaired t-test for I). Scale bars, 5 µm (A, B), 50 µm (E, G), and 12.5 µm (F, H).

Fig 5: Neutral sphingomyelinase (NSM) inhibition does not affect Shigella infection upon host cell stress A–CRepresentative images (A), cfu (B), and qRT–PCR (C) quantification of intracellular Shigella in HeLa cells, pre-treated with arsenite, with arsenite plus the NSM inhibitor GW4869, or control.D–FRepresentative images (D), cfu (E), and qRT–PCR (F) quantification of intracellular Shigella in HeLa cells, pre-treated with anisomycin, with anisomycin plus GW4869, or DMSO (control).G, HCfu (G) and qRT–PCR (H) quantification of intracellular bacteria in HeLa cells treated or not with amitriptyline and infected with Shigella WT.I, JCfu (I) and qRT–PCR (J) quantification of intracellular bacteria in HeLa cells transfected with ASM or control siRNA.KASM enzymatic activity quantification in HeLa cells, treated with arsenite, with arsenite plus the ASM inhibitor amitriptyline, or control. The ASM enzymatic activity was determined in the membrane fraction corresponding to 3.0 × 105 cells per condition.L, MCeramide quantification in HeLa cells, treated with arsenite (L) or anisomycin (M), with the stressors plus amitriptyline, or control. Ceramide levels are shown normalized to mock-treated cells.NAnalysis of smpd1 (ASM) expression determined by qRT–PCR in HeLa cells transfected with ASM or control siRNA. Results are shown normalized to cells transfected with control siRNA.OWestern blot analysis of ASM levels in HeLa cells transfected with ASM siRNA or control siRNA; ß-actin was used as loading control.PASM enzymatic activity quantification in HeLa cells transfected with ASM siRNA or control siRNA, and treated or not with arsenite. The ASM enzymatic activity was determined in the membrane fraction corresponding to 3.0 × 105 cells per condition.Data information: Shigella infection was performed at MOI 10 and analyzed at 0.5 hpi. Results are shown as mean ± s.e.m. of 5 (panels J, K–M, and P), 6 (panel I), or 7 (panels B, C, E–H, and N) independent experiments; *P < 0.05, **P < 0.01, ***P < 0.001 (one-way ANOVA for panels B, C, E, F, K–M; paired t-test for panels G–J and N; two-way ANOVA for panel P). Scale bars, 50 µm.Source data are available online for this figure.