Fig 1.
TEM visualization of the EVs produced by ExPEC strain.
(A) The purified FY26 EVs obtained from fractions F1–F8 were visualized with TEM. Scale bars: 200 nm. (B) DGU-purified EVs of different ExPEC strains (FY26, CBE59, and CFT073) were observed with TEM. The size distributions of the three purified EVs were detected with Dynamic light scattering (DLS). Scale bars: 200 nm.
Fig 2.
Comparative analysis of total proteins from ExPEC EVs.
(A) Density-gradient ultracentrifugation (DGU)-purified EVs of ExPEC strains FY26, CBE59, and CFT073 were analyzed with SDS-PAGE. Total protein (5 μg) from EVs or 15 μg of total protein from whole-cell lysates (WCLs) was loaded into the lanes. Lane M: protein marker. (B) ExPEC EVs from different density gradient fractions were analyzed with SDS-PAGE. Fractions (15 μL loaded onto the gel) are numbered according to increasing density. The images show one representative experiment. M: Protein marker; WCL: 5 μg loaded into each well of the gel. (C) Principal components analysis (PCA) of the EV proteomes of the ExPEC strains. The graph shows three cluster patterns with overlapping features for CFT073 (red cluster), CBE59 (green cluster), and FY26 (blue cluster). (D) Comparison of EV protein profiles in different ExPEC strains with a Venn diagram. Venn diagram indicates the numbers of identified proteins in the EVs of the three ExPEC strains. (E) Classification of the EV proteins identified in different ST strains by subcellular location. Subcellular localization of the identified EV proteins was predicted with CELLO (http://cello.life.nctu.edu.tw). (F) Functional classification of EV unigenes identified in the three ExPEC strains was determined with Gene Ontology (GO). The GO enrichment analysis identified three categories: cellular component, molecular function, and biological process.
Fig 3.
Quantification and bioinformatic analysis of the proteins enriched in ExPEC EVs.
(A) Abundances of the 300 most-abundant proteins of FY26 EVs relative to the total protein. The horizontal axis indicates 50 proteins sorted by abundance from most to least abundant relative to the total protein abundance, and the vertical axis indicates the protein abundance relative to the total protein. (B) Subcellular localization classification of the 300 most-abundant proteins identified in EVs from different ExPEC strains (FY26, CBE59, and CFT073). Subcellular localization of the identified proteins was determined with CELLO (http://cello.life.nctu.edu.tw). The numbers of the proteins are displayed below the pie chart, and the percentage abundances are shown in the pie chart. (C) The 300 most-abundant proteins isolated from the FY26 EV fractions were compared with their abundances in the total FY26 cellular fractions. Subcellular localization is shown as follows: outer-membrane proteins (light blue), periplasmic proteins (purple), inner-membrane proteins (red), cytoplasmic proteins (deep blue), moonlighting proteins (green), and ribosomal subunit proteins (orange). Proteins enriched in the FY26 EVs are shown below and to the right of the dashed line, and proteins depleted in the EVs are shown above and to the left of the line. The data for this figure can be found in S1 Table.
Fig 4.
Detection of membrane and cytoplasmic proteins in density gradient ultracentrifugation (DGU)-purified ExPEC EVs.
(A) Expression of membrane and cytoplasmic proteins in EVs and EV-free extracellular medium of ExPEC strains was determined with western blotting. Total protein (1 μg) was loaded into the OmpA and Gapdh lanes, and 5 μg of total protein was loaded into the other lanes. Molecular weight markers are shown on the right. (B) Expression of membrane and cytoplasmic proteins in DGU-purified EVs (F1–F10) was determined with western blotting. Nonfractionated EVs were used as the positive controls, and whole-cell lysates (WCLs) were used as the loading controls. Total protein (1 μg) was loaded into the OmpA and Gapdh lanes, and 5 μg of total protein was loaded into the other lanes. (C) Dissociation assays confirmed that the protein cargoes were tightly associated with the EVs of the ExPEC strains. OptiPrep-purified EVs were treated with HEPES buffer containing the indicated chemical agents or with HEPES buffer only. The pellets (P; containing EVs) and extracellular media (S; containing proteins released from EVs) were collected by ultracentrifugation, and the samples were analyzed with western blotting. Total protein (1 μg) was loaded into the OmpA and Gapdh lanes, and 5 μg of total protein was loaded into the other lanes. (D) Immunoblots of proteinase K (PK)-untreated (PK-) and PK-treated (PK+) ExPEC EVs either intact (EDTA+) or lysed with 0.1 M EDTA (EDTA+) with the indicated antibodies.
Fig 5.
ExPEC EVs contain multiple types of vesicles.
(A) Green fluorescent protein (GFP) in ExPEC strains was visualized with fluorescence microscopy. Scale bars: 20 μm. (B) Illustration of experiment to detect cytoplasm-carrying membrane vesicles in ExPEC EVs. (C) Immunofluorescent images of EVs produced by ExPEC strains. The EVs were stained with TRITC-labeled primary mouse anti-EV antibody. The coexistence ratio of GFP-carrying vesicles (green fluorescence) among the total EVs (red fluorescence) represents the percentage of cytoplasm-carrying membrane vesicles, which is indicated on the right. Data were obtained from at least three independent experiments. Graphs show a representative image. Scale bars: 10 μm. (D) GFP and OmpA proteins in EVs were determined with western blotting. (E) The outer–inner-membrane vesicles (OIMVs) of ExPEC were observed with transmission electron microscopy (TEM). ExPEC EVs displayed a double bilayered structure (red square), or single bilayered structure (orange square). Scale bars: 200 nm. (F) The outer–inner-membrane vesicles (OIMVs) of ExPEC were observed with Cryo-TEM. ExPEC EVs displayed a double bilayered structure (red square), or single bilayered structure (orange square). Scale bars: 200 nm.
Fig 6.
Deletion of pal gene affects the production of ExPEC outer-membrane vesicles (OMVs).
(A) The concentrations of purified EVs produced by FY26 and FY26ΔPal were determined with a nanoparticle tracking analysis (NTA). (B) Immunofluorescent detection of EVs produced by FY26Δpal. The percentage of GFP-carrying vesicles (green fluorescence) is indicated on the right. Data were obtained from at least three independent experiments, with three replicates. Graph shows a representative image. Scale bars: 10 μm. (C) GFP and OmpA proteins in EVs were determined with western blotting. (D) The EVs produced by FY26ΔPal were observed with transmission electron microscopy (TEM), and the size distributions of the purified EVs were detected with Dynamic light scattering (DLS). Scale bars: 200 nm. (E) Levels of membrane and cytoplasmic proteins in EVs from FY26 and FY26Δpal were determined with western blotting. (F) Levels of membrane and cytoplasmic proteins in EV-free extracellular medium of FY26 and FY26Δpal were determined with western blotting.
Fig 7.
Prophage endolysins induce the formation of ExPEC cytoplasm-carrying vesicles.
(A) Transcription levels of endolysin genes in ExPEC strains in different growth phases (4, 6, 8, 10, and 12 h) were determined with RT–qPCR.–h, putative holin genes; -e, putative endolysin genes. Data shown are the means ± SEM of three independent experiments relative to the housekeeping gene dnaE. Statistical significance was evaluated with one-way ANOVA (**P < 0.01). (B) Protein levels of endolysins (Epel1 and Epel2 variants) in whole-cell lysates of ExPEC strains were determined with western blotting. (C) Protein levels of endolysins (Epel1 and Epel2 variants) in EVs of ExPEC strains (CBE59, FY26 and CFT073) at 12 h were determined with western blotting. (D) The concentrations of purified EVs produced by several mutants were determined with a nanoparticle tracking analysis (NTA). (E) Protein levels of Epel1 in complemented strains were determined with western blotting. (F) Protein levels of Epel1 and the Epel2 variants in the EVs produced by the complemented strains at 12 h were determined with western blotting. (G) The concentrations of purified EVs produced by FY26Cepel1, FY26Cepel2.1, and FY26Cepel2.2 were determined with NTA. (H) Immunofluorescent detection of EVs produced by several mutants and complemented strains. (I) GFP and OmpA proteins in EVs were determined with western blotting. (J) The EVs produced by FY26Cepel1, FY26Cepel2.1, and FY26Cepel2.2 were observed with transmission electron microscopy (TEM). The size distributions of the three purified EVs were detected with Dynamic light scattering (DLS). Scale bars: 200 nm.
Fig 8.
SOS response promotes the formation of ExPEC cytoplasm-carrying vesicles.
(A) The concentration of purified EVs produced by FY26ΔlexA were determined with a nanoparticle tracking analysis (NTA). (B) Immunofluorescent staining of EVs produced by FY26ΔlexA. (C) The EVs produced by FY26ΔlexA were observed with transmission electron microscopy (TEM). Scale bars: 200 nm. (D) Schematic diagram of transcriptional promoter regions of epel1, epel2.1, and epel2.2 operons. The LexA-binding box is marked in red frame; the corresponding −10 and −35 boxes, the transcription initiation sites, and the start codons of epel1, epel2.1, and epel2.2 are underlined. (E) Direct binding of LexA to the promoters of epel1 (a), epel2.1 (b), and epel2.2 (c) was detected with EMSA. DNA fragment (200 bp) of each endolysin gene promoter (epel1, epel2.1, and epel2.2) containing the LexA-binding site, and the negative control (200 bp) containing the pal gene were amplified. Each DNA probe was mixed with an increasing amount of LexA protein for the EMSA analysis. (F) The concentrations of purified EVs produced by wild-type (WT) FY26 and mutant FY26ΔrecA were determined with NTA. The strains were exposed to sublethal concentrations of H2O2 or cultured under routine conditions. (G) Immunofluorescent staining of the EVs produced by FY26ΔrecA. (H) Immunofluorescent detection of EVs produced by FY26 and FY26ΔrecA. The strains were exposed to sublethal concentrations of H2O2. (I) GFP and OmpA proteins in EVs were determined with western blotting. (J) The EVs produced by FY26 and FY26ΔrecA were observed with transmission electron microscopy (TEM). The strains were exposed to sublethal concentrations of H2O2. Scale bars: 200 nm.
Fig 9.
Deletion of ftsK gene increased the production of ExPEC cytoplasm-carrying vesicles.
(A) The concentrations of EVs purified from FY26ΔftsK, FY26CftsK and FY26ΔftsK/recA were determined with a nanoparticle tracking analysis (NTA). (B) Immunofluorescent staining of EVs produced by FY26ΔftsK, FY26CftsK and FY26ΔftsK/recA. Percentage of cytoplasm-carrying vesicles (green fluorescence) among the total EVs (red fluorescence) is indicated on the right. Graph shows a representative image. Scale bars: 10 μm. (C) GFP and OmpA proteins in EVs were determined with western blotting. (D) The EVs produced by FY26ΔftsK and FY26ΔftsK/recA were observed with transmission electron microscopy (TEM). Scale bars: 200 nm. (E) Levels of membrane and cytoplasmic proteins in EVs (a) and EV-free extracellular medium (b) from FY26ΔftsK and FY26CftsK were determined with western blotting. (F) Protein levels of endolysins Epel1 and Epel2 variants in EVs of FY26ΔftsK and FY26ΔftsK/recA were determined with western blotting.
Fig 10.
Absence of t6A synthesis genes in ExPEC enhanced the production of cytoplasm-carrying vesicles.
(A) The concentrations of EVs purified from FY26Δt6A, FY26Ct6A and FY26Δt6A/recA were determined with a nanoparticle tracking analysis (NTA). (B) Immunofluorescent staining of EVs produced by FY26Δt6A, FY26Ct6A and FY26Δt6A/recA. Percentage of cytoplasm-carrying vesicles (green fluorescence) among the total EVs (red fluorescence) is indicated on the right. Graph shows a representative image. Scale bars: 10 μm. (C) GFP and OmpA proteins in EVs were determined with western blotting. (D) The EVs produced by FY26Δt6A and FY26Δt6A/recA were observed with transmission electron microscopy (TEM). Scale bars: 200 nm. (E) Levels of membrane and cytoplasmic proteins in EVs (a) and EV-free extracellular medium (b) from FY26Δt6A and FY26Ct6A were determined with western blotting. (F) Protein levels of endolysins Epel1 and Epel2 variants in EVs of FY26Δt6A and FY26Δt6A/recA were determined with western blotting.
Fig 11.
Proportion of cytoplasm-carrying vesicles affects the DNA content carried by ExPEC EVs.
(A) Total DNA in EVs was measured with microplate reader. Total DNA was extracted from equivalent numbers of EVs (7.1 × 1011 vesicles) from FY26, FY26Δpal, FY26ΔftsK, FY26CftsK, FY26Δt6A, and FY26Ct6A were measured with microplate reader. Data were obtained from at least three independent experiments, with three replicates. Statistical significance was evaluated with one-way ANOVA (**P < 0.01). (B) Total DNA in the EVs (FY26, FY26Δpal, FY26ΔftsK, FY26CftsK, FY26Δt6A, and FY26Ct6A) was visualized with nondenaturing polyacrylamide gel electrophoresis. Total DNA was isolated from equivalent numbers (7.1 × 1011) of PK/DNase-treated and untreated EVs from wild-type (WT) strain FY26 and several mutants. ‘+’ indicates that samples were treated with PK and DNaseI, and ‘−’ indicates that samples were not treated with PK or DNaseI. Naked DNA (pET-32a) was used as a control. (C) Total DNA in EVs was measured with microplate reader. Total DNA was extracted from equivalent numbers of EVs (7.1 × 1011 vesicles) from WT FY26, triple-deleted mutant FY26Δepel1/2.1/2.2, and endolysin-overexpressing complemented strains FY26Cepel1, FY26Cepel2.1, and FY26Cepel2.2, and measured with microplate reader. (D) Total DNA in EVs was visualized with nondenaturing polyacrylamide gel electrophoresis. Total DNA was extracted from equivalent numbers of EVs (7.1 × 1011 vesicles) from WT FY26, triple-deleted mutant FY26Δepel1/2.1/2.2, and endolysin-overexpressing complemented strains FY26Cepel1, FY26Cepel2.1, and FY26Cepel2.2, and analyzed with nondenaturing polyacrylamide gel electrophoresis. Naked DNA (pET-32a) was used as a control. (E) Circular maps of the complete plasmid sequence carried by FY26 EVs. FY26 contained one large plasmid pFY26-1. Circles from outer to inner show: pFY26-1 plasmid sequence (ring 1, red), homologous alignment (ring 2, green), DNA in FY26 EV (ring 3, orange), genes on the forward strand, colored according to COG classification (ring 4), and genes on the reverse strand, colored according to COG classification (ring 5).
Fig 12.
Antibiotics enhance the production of ExPEC EVs.
(A) The concentrations of purified EVs in FY26 cultured with sublethal concentrations of antibiotics were determined with a nanoparticle tracking analysis (NTA). FY26 strain was treated with sublethal doses of seven antibiotics (ampicillin, 2 μg/mL; ceftazidime, 0.3 μg/mL; imipenem, 2.5 μg/mL; ciprofloxacin, 0.3 μg/mL; chloramphenicol, 2 μg/mL; colistin, 0.3 μg/mL, sulfamethoxazole, 2 μg/mL). (B) Immunofluorescent staining of EVs produced by FY26 strains treated with sublethal concentrations of antibiotics. Proportion of cytoplasm-carrying vesicles is indicated on the right. (C) GFP and OmpA proteins in EVs were determined with western blotting. (D) The purified EVs in FY26 cultured with sublethal concentrations of antibiotics were observed with transmission electron microscopy (TEM). Scale bars: 200 nm. (E) Protein levels of endolysins Epel1 and the Epel2 variants in whole-cell lysates (WCLs) and EVs of FY26 strain cultured with sublethal concentrations of antibiotics were determined with western blotting.
Fig 13.
(A) Confocal microscopic visualization of the internalization of ExPEC EVs into HD11 macrophages. Infected HD11 cells were incubated with anti-ExPEC antibody, and stained with FITC-conjugated goat anti-mouse IgG antibody (green), phalloidin (red), and DAPI (blue). Images show a representative of at least three independent experiments. Scale bar: 10 μm. (B) To investigate the cytotoxic effects of ExPEC EVs on THP-1 and HD11 macrophages, THP-1 or HD11 cells were incubated with EVs (50 μg/mL) for different periods (2, 4, 8, 16, or 24 h). Cell viability was measured with a CCK-8 kit. The results are presented as means ± SEM of at least three independent experiments. One-way ANOVA was used to evaluate statistical significance (*P < 0.01). (C) To investigate the cytotoxic effects of ExPEC EVs on THP-1 and HD11 cells, EVs were isolated from mutant strains FY26ΔftsK, FY26Δt6A, and FY26ΔlexA, the endolysin-overexpressing strain FY26Cepel1, and antibiotic-treated strain FY26. (D) The cytokines (IL-1β, IL-6, IL-8, IL-12β, and TNF-α) from the cell-free supernatant of THP-1 macrophages was measured using the commercial cytokine ELISA kits. The results are presented as means ± SEM of at least three independent experiments. One-way ANOVA was used to evaluate statistical significance (*P < 0.01).
Fig 14.
Increased proportion of cytoplasmic vesicles in ExPEC EVs caused more-severe mitochondrial dysfunction and apoptosis in macrophages.
(A) Mitochondrial membrane potential (ΔΨm) was determined with JC-1 dye and flow cytometry. Bone-marrow-derived macrophages (BMDMs) were treated with FY26 EVs (50 μg/mL) or PBS for 0, 2, 4, 8, 12 or 24 h and labeled with JC-1. The percentage of normal cells (JC-1 aggregates) in BMDMs is indicated on the right. (B) Mitochondrial membrane potential (ΔΨm) was determined with JC-1 dye and flow cytometry. Bone-marrow-derived macrophages (BMDMs) were treated with purified ExPEC EVs (50 μg/mL) or PBS for 12 h and labeled with JC-1. The percentage of normal cells (JC-1 aggregates) in BMDMs is indicated on the right. (C) Cytosolic and mitochondrial fractions of BMDMs were analyzed for cytochrome c with western blotting. BMDMs were treated with purified ExPEC EVs (50 μg/mL), staurosporine (STS), or PBS for 24 h, anti-tubulin and anti-COX4 were used as the loading control. Molecular weight markers are shown on the left. (D) Levels of caspase 3 (17 kDa) in the BMDMs were determined with western blotting. BMDMs were treated with purified ExPEC EVs (50 μg/mL) or PBS for 24 h; anti-tubulin was used as the loading control. (E) Levels of MCL-1 and BCL-XL in the BMDMs were determined with western blotting. BMDMs were treated with purified ExPEC EVs (50 μg/mL) or PBS for 2, 4, 6, or 24h; anti-tubulin was used as the loading control.
Fig 15.
Illustration of the generation mechanisms of EVs in ExPEC.
(A) The biogenesis mechanisms for different types ExPEC EVs. ExPEC EVs can be considered a mixture of classical outer-membrane vesicles (OMVs) and cytoplasm-carrying vesicles: explosive outer-membrane vesicles (EOMVs) and outer–inner-membrane vesicles (OIMVs). Therefore, typical OMVs, formed through the blebbing of the bacterial outer membrane from the envelope, have a single-layered membrane, which originated from the unbalanced biosynthesis of the cell envelope. A defect in the crosslinking between peptidoglycan and the outer membrane leads to the production of OMVs. The “explosive cell lysis” model is a possible mechanism for the formation of ExPEC EVs, in which the bacterial peptidoglycan is degraded by phage-derived endolysins to induce explosive cell lysis, and the broken membrane fragments gather and self-assemble to form EOMVs (single-layered membrane) or OIMVs (bilayer membrane). (B) Model of the generation mechanism of ExPEC cytoplasm-carrying vesicles. The unbalanced cell division caused by the deletion of the ftsK gene and t6A synthesis defects or by the toxicity caused by exposure to H2O2 reduces ExPEC viability, thus increasing the production of cytoplasm-carrying vesicles through the RecA/LexA-dependent SOS response. Two regulatory proteins (RecA and LexA) control the expression of the SOS response genes in E. coli. This study demonstrated that the repressor LexA directly suppresses the expression of endolysins (Epel1, Epel2.1, and Epel2.2) by binding to the SOS boxes in the endolysin promoter regions. In response to DNA damage, the binding of RecA to single-stranded RNA activates RecA to stimulate the autocatalytic cleavage of the LexA repressor. The expression of prophage-associated endolysins is then activated, which triggers cell lysis and increases the production of ExPEC cytoplasm-carrying vesicles. Antibiotic treatment also reduces bacterial viability, thus increasing the production of ExPEC cytoplasm-carrying vesicles through the RecA/LexA-dependent SOS response. In contrast, the deletion of pal only causes a peptidoglycan crosslinking defect, which promotes the formation of classic OMVs.