Table 1.
E. coli strains and plasmids used in this study.
Figure 1.
Generalized LPS structures in a series of E. coli LPS mutants used in this study.
A silver-stained polyacrylamide gel after SDS-PAGE is shown in the upper panel. In the lower panel, the drawing of LPS structures especially highlights the core OS portion of LPS. Structures of the major glycoforms of core OS are based on a structural analysis of a K-12 strain, W3100 [59] and a waaP mutant of strain R1 [60]. The LPS core OS composition of the hldE strain (RN102) is the same as that of the waaC strain (RN101), probably because the lack of HldE protein results in arrest of ADP-Hep biosynthesis. It has been reported that in the E. coli R1 strain, inactivation of the waaP gene results not only in the loss of all phosphate groups on HepI and HepII, but also loss of HepIII [60]. We could not detect any difference in size of LPS between RN107 and BW25113 or RN105, although the loss of Gal (shown by an asterisk) was expected theoretically. Therefore, the LPS structure of RN107 was described based on the information of the SDS-PAGE gel in this study. A more precise analysis is required to fully understand the LPS structure of RN107. Each sugar or amino sugar of core OS is shown by a black (Hep), blue (Glc), red (Gal), green (Kdo), or gray (GlcN) circle. Phosphate groups (Phosphate or pyrophosphorylethanolamine) modified on sugar are shown by yellow boxes (P/PPEtN). Hep: L-glycero-D-manno-heptose, Kdo: 3-deoxy-D-manno-oct-2-ulosonic acid, GlcN: N-acetylglucosamine.
Figure 2.
Biofilm formation and growth of a series of LPS mutants.
(A) Biofilm formation by a series of core OS LPS mutants when compared to the parental strain, BW25113. The mean ± SD of results from 3 independent experiments are shown. Statistical analysis was performed using ANOVA. *P<0.05, against biofilm formation level of strain BW25113 (B) Growth of BW25113 and the LPS mutants. Each strain was grown in LB broth under shaking conditions at 37°C. Absorbance at OD600 was measured at different time points.
Table 2.
Function and phenotype expressed by mutation of the genes selected in this study.
Figure 3.
Autoaggregation phenotype by LPS mutants.
Each strain standardized at OD600 = 1.0 in PBS was used for autoaggregation assay. The value at OD600 after an18-hour incubation is shown as the mean ± SD of results from three independent experiments. Statistical analysis was performed using ANOVA. *P<0.05, against autoaggregation level of strain BW25113.
Figure 4.
Surface hydrophobicity of LPS mutants.
(A) Hydrophobicity assay using hexadecane. Each strain standardized at OD600 = 1.0 in PBS was used for a hexadecane hydrophobicity assay. Absorbance at OD600 in the water phase was quantified before and after a 10-min incubation. The percent hydrophobicity was calculated by the formula: % hydrophobicity = [1− (OD600 after vortex/OD600 before vortex)] ×100. The mean ± SD of results from 4 independent experiments are shown. Statistical analysis was performed using ANOVA. *P<0.05, against hydrophobicity of strain BW25113. (B) XPS data showing the percent of total carbon at the surface that is present in the form of aliphatic (non polar) carbon.
Figure 5.
Fourty eight-hour-cultured biofilms were collected and analyzed by TEM, and by Western blot for FliC. (A–E) TEM images of the bacterial cells and the cell appendages are shown for strains (A) BW25113, (B) RN102, (C) BW25113/pNTR-SD, (D) RN102/pNTR-SD, and (E) RN102/pNT3(hldE). Flagella found in figures (A), (C), and (E) are shown by arrowheads. Representative electron-microphotographs of each strain are shown. A 1-μm-long bar is shown in the lower left corner. (F) Supernatants were collected from 48-hour bacterial cultures. Result of Western blot using anti-FliC antiserum is shown. Lanes; 1, BW25113; 2, RN102; 3, BW25113/pNTR-SD; 4, RN102/pNTR-SD; 5, RN102/pNT3(hldE); 6, RN110.
Figure 6.
Contribution of eDNA to biofilm structure formed by RN102.
(A–E) CLSM images of biofilms formed by strains: (A) BW25113 (B) RN102, (C) BW25113/pNTR-SD, (D) RN102/pNTR-SD, and (D) RN102/pNT3(hldE). Images of biofilms stained with acrydine orange are shown as digital CLSM images. In each strain, a section which has the largest sum of signals in the defined area (127.3 μm by 127.3 μm) among all X–Y sections is shown in the upper row (X–Y). The overview of biofilms in the same area of each X–Y section is shown as 3D image in the lower row (3D). The volume of each 3D image (μm3) in the area of the X–Y planes was quantified and the mean ± SD obtained from 3 different areas chosen at random are denoted in the upper-right corners. The data shown are representative microphotographs of two independent experiments. (F) Quantification of eDNA from BW25113 and RN102 strains. The bars represent the ratio of extracellular DNA to intracellular DNA (eDNA/iDNA). Results are shown as the mean ± SD from 3 independent experiments. *P<0.05. Stastical analysis was performed using Mann-Whitney's U-test. (G and H) Effect of DNase I on biofilm formation by the RN102 in a clear test tube (G) and as quantified in a 96-well plate (H). The RN102 was grown in presence of different concentrations of DNase I or in presence of pre-heated DNase I or without DNase I for 48 hours under static conditions at 37°C. The mean ± SD of results from 3 independent experiments are shown. Statistical analysis was performed using ANOVA. *P<0.05, against the biofilm formation by RN102 without DNase I treatment.
Figure 7.
Loss of outer membrane integrity in strain RN102.
Bacterial samples were collected from 48-hour-cultured biofilms for TEM analysis. TEM images of the bacterial cells and the cell appendages are shown for strains: (A) BW25113, (B) RN102, (C) BW25113/pNTR-SD, (D) RN102/pNTR-SD, and (E) RN102/pNT3(hldE). The outer membranes are indicated by arrows. Representative electron-microphotographs of each strain are shown. A 500-nm-long bar is shown in the lower left corner of each eclectron-micrograph. (F) Western blot analysis of supernatants from BW25113 and RN102. Supernatants were harvested by centrifugation from bacterial liquid culture grown for 48 hours under static conditions. Results of Western blot using anti-Crp, anti-DsbA, anti-OmpC, and anti-OmpA antisera are shown. (G) Supernatants from bacterial liquid cultures of BW25113 or RN102 grown for 48 hours under static conditions were serially diluted with TE. The diluted samples were used as template DNA for PCR using E. coli atoS gene-specific primer pairs. Lanes: 1, without dilution; 2, 10−1 dilution; 3, 10−2 dilution; 4, 10−3 dilution; 5, 10−4 dilution; 6, 10−5 dilution; 7, 10−6 dilution.
Figure 8.
(A) The OMVs were isolated from the supernatant of a 80-ml bacterial liquid culture incubated 14 hours under shaking conditions. Finally, each OMV preparation was resuspended with 200 μl of 20 mM Tris-Cl (pH 8.0). Five μl of each OMV sample per well was run on 12% PAGE and subjected to silver staining. Total protein amounts (mg) of OMVs from a 80-ml bacterial culture of each strain were determined by Bradford assay and are presented below each lane. Lanes; 1, BW25113; 2, RN102; 3, BW25113/pNTR-SD; 4, RN102/pNTR-SD; 5, RN102/pNT3(hldE). (B and C) AFM images of OMVs prepared from BW25113 and RN102 on 1 μm2 surfaces were shown.