A) CHO cells or B) CHO, HeLa, MEF Atg5^+/+ and T24 cells were infected for 60 min with wild-type S. marcescens and then extracellular bacteria were eliminated with gentamicin. At the indicated times, cells were washed with PBS and lysed with 0.05% Triton X-100. The CFUs were determined on LB agar plates, percentages were calculated relative to CFUs in the inoculum. In the plot shown in (B) fold change in CFUs were calculated relative to the values obtained at 120 min p.i. The average ± S.D. for two independent experiments is shown (* p<0.001).

HeLa, MEFs Atg5^+/+, MEF Atg5^−/− and T24 cells were transiently transfected with pRFP-LC3 (left panels, red fluorescence) and invaded with wild-type S. marcescens transformed with pGFP (right panels, green fluorescence). Cells were fixed at 180 min p.i. and visualized by confocal laser microscopy. Images are representative of at least 300 infected cells monitored in two independent assays. Bars: 10 µm.

A) Inhibition or neutralization of vacuolar acidification on Serratia invasion. CHO-EGFP-LC3 cells were maintained in α-MEM, incubated with DMSO 1% (which equals the final concentration of DMSO used as vehicle for bafilomycin A1), 100 nM bafilomycin A1 or 10 mM NH₄Cl, as indicated in Materials and Methods. Cells were infected with wild-type S. marcescens following the gentamicin protection assay procedure. Invaded cells were lysed with 0.05% Triton X-100 and CFUs were determined on LB agar plates at 120, 240 or 360 min p.i. Percentages were calculated relative to CFUs in the inoculum. The average ± S.D. is shown for three independent experiments, *p<0.001. B) Modulation of autophagy on Serratia invasion. CHO-EGFP-LC3 cells were maintained in α-MEM, or pre-treated in EBSS starvation medium (Starv). 100 nM Wortmannin (Wn) was added 2 h previous to the infection assay (α-MEM + Wn; or Starv + Wn) or at 60 min p.i. (α-MEM + Wn p.i). Cells were infected with wild-type S. marcescens following the gentamicin protection assay procedure. Invaded cells were lysed with 0.05% Triton X-100 and the CFUs were determined on LB agar plates at 120 or 360 min p.i. The average ± S.D. is shown for three independent experiments (% intracellular CFUs are indicated above each bar in the graph; nd: not determined, *p<0.001).

CHO-EGFP-LC3 cells subjected to the invasion assays as described in Fig. 5A were visualized by confocal laser microscopy. a and b panels show representative images of control, non-invaded cells and c-f panels depict invaded cells incubated with α-MEM, α-MEM +100 nM Baf or α-MEM +10 mM NH₄Cl, as indicated. Bacteria were detected with antibodies anti-Serratia coupled with a secondary antibody labeled with Cy3. Representative DIC (c), EGFP-LC3 green fluorescence (d), red fluorescence-labeled bacteria (e) and merged images (f) are shown. At least 300 cells were inspected for each condition in two independent experiments. Bars: 10 µm.

A) Autophagy triggered by Serratia requires Atg5. MEF Atg5^+/+ and MEF Atg5^−/− cells were infected for 60 min with wild-type S. marcescenes. Thereafter, extracellular bacteria were eliminated with gentamicin. At the indicated times, cells were washed with PBS and lysed with 0.05% Triton X-100. The CFUs on LB agar plates were determined. Units were calculated relative to CFU recovered from MEF Atg5^+/+ cells at 120 min p.i. The media ± S.D. is shown for three independent experiments. B) S. marcescens actively induces autophagy from the extracellular media. Overnight S. marcescens cultures were incubated 60 min in LB without antibiotic or supplemented with 20 or 100 µg/ml chloramphenicol (bacteriostatic condition). Subsequently, CHO-EGFP-LC3 cells were infected with bacteria and then extracellular bacteria were eliminated with gentamicin (bacteriostatic treatment was maintained throughout the invasion assay). At 180 min p.i, cells were fixed and bacteria were detected with antibodies anti-Serratia coupled with a secondary antibody labeled with Cy3. Finally, the percentage of colocalization of vesicles containing Serratia with EGFP-LC3, relative to total infected cells, was determined by fluorescence microscopy. The lower panel shows the percentage of cells, infected and non-infected, with an autophagic phenotype. At least 300 infected cells were counted for each condition. The average ± S.D. is shown for three independent experiments, * and ** p<0.001. C) CHO-EGFP-LC3 cells (green fluorescence) infected with S. marcescens (red fluorescence) untreated (upper panel), treated with 20 µg/ml chloramphenicol (middle panel) or treated with 100 µg/ml chloramphenicol (lower panel) are shown. Bars: 10 µm.

A) CHO-EGFP-LC3 cells (green fluorescence) or B) CHO cells transfected with EGFP-Rab7 (green fluorescence), were infected with wild-type S. marcescens and cells were fixed at 120 min p.i (a-d) or 360 minutes p.i (e-h). Bacteria were detected with antibodies anti-Serratia coupled with a secondary antibody labeled with Cy3 (red fluorescence). Representative confocal Z-slices are shown. Inset images show higher magnification of the boxed areas in merged images. Bars: 10 µm. C) CHO-EGFP-LC3 cells or CHO cells transfected with pEGFP-Rab7 were infected with wild-type S. marcescens. Cells were fixed at the indicated time points and bacteria were detected with antibodies anti-Serratia coupled with a secondary antibody labeled with Cy3. The percentage of colocalization of EGFP-LC3 (•) or EGFP-Rab7 (▴) with bacteria was determined by fluorescence microscopy. At least 200 infected cells were counted for each time point. The average ± S.D. for three independent experiments is shown.

A) CHO-EGFP-LC3 cells were infected with wild-type S. marcescens. After 60 min p.i, cells were incubated with 3 µM of LysoTracker to label acidic compartments. Cells were fixed at 240 or 360 min p.i. and intracellular bacteria were detected with antibodies anti-Serratia coupled with a secondary antibody labeled with Alexa Fluor 647. The percentages of colocalization of bacteria with EGFP-LC3 and/or with Lysotracker were determined by fluorescence microscopy. At least 300 infected cells were counted for each condition. The average ± S.D. is shown for two independent experiments, *and **p<0.001. B) Representative confocal laser captured images of CHO-EGFP-LC3 cells (green fluorescence, d) infected with S. marcescens (blue fluorescence, b) and incubated with Lysotracker (red fluorescence, c) at 360 min p.i. are shown. Inset image (f) shows higher magnification of the boxed area in the merged image (e), and highlights a non-acidic, autophagic SeCV. Arrows point at acidic vesicles. Bars: 10 µm. C) CHO-EGFP-LC3 cells were pre-incubated with 10 µg/ml of DQ-BSA for 4 hours to label degradative compartments. Subsequently, cells were infected with wild-type S. marcescens and fixed at 240 or 360 min p.i. Intracellular bacteria were detected with antibodies anti-Serratia coupled with a secondary antibody labeled with Alexa Fluor 647. The percentages of colocalization of bacteria with EGFP-LC3 and/or with DQ-BSA were determined by fluorescence microscopy. At least 300 infected cells were counted for each condition. The average ± S.D. is shown for two independent experiments, *and **p<0.001. D) Representative confocal laser captured images of CHO-EGFP-LC3 cells (green fluorescence, d) pre-incubated with DQ-BSA (red fluorescence,c) and infected with S. marcescens (blue fluorescence, b) at 360 min p.i. are shown. Inset image (f) shows higher magnification of the boxed area in the merged image (e) and highlights a non-degradative, autophagic SeCV. Arrows point at degradative vesicles. Bars: 10 µm.

CHO-EGFP-LC3 cell monolayers subjected to the invasion assays as described in Figure 5 were harvested and resuspended in SDS-PAGE sample buffer. The samples were resolved by SDS-PAGE (30 µg of total protein was loaded per lane), transferred to nitrocellulose and detected with anti-GFP polyclonal antibodies (“invasion”, lower panel blots). Control, non-invaded CHO-EGFP-LC3 cells, that were subjected to the same treatments were also analyzed (“control”, upper panel blots). The positions of LC3-I and LC3-II bands are indicated. Protein loading was controlled according to the α-tubulin content (the blots detected with antibodies anti-α-tubulin are shown). Blots are representative of three independent experiments. The bands intensity was quantified with the ImageJ Programe, and the LC3II/LC3I (II/I) was calculated for the immunodetection experiment shown in the figure. The histogram represents the average ± S.D. of the relative intensities of LC3-I or LC3-II versus α-tubulin of three independent experiments.

CHO-EGFP-LC3 cells subjected to the invasion assays as described in Fig. 5B were visualized by confocal laser microscopy. Bacteria were detected with antibodies anti-Serratia coupled with a secondary antibody labeled with Cy3. Representative DIC (a and e), EGFP-LC3 green fluorescence (b and f), red fluorescence labeled bacteria (c and g) and merged fluorescence images (d and h) are shown. In the assays where wortmannin was used, representative images from both the scarce SeCV observed (a-d) and from barely invaded cell fields (e-h) are shown. The lower panel shows representative images of control, non-invaded cells incubated with α-MEM, in starvation medium or in starvation medium + Wn (green EGFP-LC3 fluorescence, a, b and c, respectively). Bars: 10 µm. The percentage of invaded cells displaying the phenotype shown in the images in the corresponding row is indicated to the right. At least 200 cells were counted for each condition; the media ± S.D. is shown for two independent experiments.

A) CHO cells were infected for 60 min with S. marcescens wild-type, flhD and flhD/pflhD strains (dark gray, light gray or black bars, respectively). Cells were washed with PBS and lysed with 0.05% Triton X-100 (left graphic); or incubated 120 o 240 min p.i with medium supplemented with gentamicin to eliminate extracellular bacteria. Finally, cells were lysed with 0.05% Triton X-100 (right graphic) and the CFUs were determined on LB agar plates. Adherence and intracellular CFUs percentages were calculated relative to initial inoculums. The average ± S.D. is shown for four independent experiments, * p<0.001. B) CHO-EGFP-LC3 cells (green fluorescence) were infected with S. marcescens wild-type (b and f), flhD (c and g) and phlA (d and h) strains. Non infected cells are shown as control (a and e). Then, cells were not treated (a-d) or treated with gentamicin (e-h) for 120 min. Cells were fixed and bacteria were detected with antibodies anti-Serratia coupled with a secondary antibody labeled with Cy3 (red fluorescence). Representative confocal Z-slices are shown. Bars: 10 µm.