Fig 1.
(A) Crystal structure of the FliC D2/D3 domains from P. aeruginosa. Marine blue—D2, cyan—D3, dark blue—common beta-strand. Alpha-helices and beta-strands are numbered in order of appearance in D2/D3 domains, not in full-length FliC. (B) Superposition of outer domain structures of P. aeruginosa PAO1 (color code as in A) and PAK (orange) strains. (C) Full-length FliC of PAO1 with schematic representation of FliC protein organization. Colors on the bar correspond to domain colors on the structure. (D) Cryo-electron micrograph of the PAO1 flagellar filaments. The scale bar represents ~400 Å. (E) Cryo-EM reconstruction of the PAO1 flagellar filament with curvature present due to not imposing helical symmetry on the reconstruction. (F) Top-down view of the PAO1 filament model fit into the density map.
Fig 2.
(A) Supercoiled model of the PAO1 flagellar filament showing curvature created by aligning several copies of the filament model as previously described [28]. A single protofilament on the outer curve is colored gold. (B) Cryo-EM density map (grey) and model showing the outer domains for the inner curve protofilaments. Models colored blue and pink have distinct outer domain conformations. (C) Subunits with the two different outer domain conformations observed for 10 PAO1 protofilaments. Conformation 1 (blue) has its outer domain tilted counterclockwise with respect to filament axis while conformation two has its outer domain tilted clockwise. (D) Alignment of conformations 1 and 2 by domains D0 and D1 reveals a displacement of 4–6 Å in domain D2 and 8–15 Å in domain D3 between the two conformations. (E) The outer-most curved protofilaments with a seam protofilament (gold) having subunits in nearly identical conformations. (F) A flagellin subunit from the seam protofilament. (G) Comparison of the outer domain orientations of the conformation 1 and 2 subunits with the seam subunit (conformation 3).
Fig 3.
Three interfaces that D20 and D30 form with surrounding domains (highlighted with circles): D30-D2+11 (equivalent to D3-11-D20), D30-D1+11 and D20-D10.
Fig 4.
Effects of the FliC mutations on P. aeruginosa PAO1 swimming motility and filament formation.
Swimming motility analysis and negative-stain EM images of PAO1-ΔfliC strain complemented with FliC bearing mutations in the D20-D10 interface (A), D30-D2+11 interface (B), and D30-D1+11 interface (C). Positions of the interfaces are labeled red on the FliC dimer cartoon on the left. Area of motile spread for each strain represents the average of ten replicates and it is normalized to that of the full-length wild type complemented strain (FliCWT). Error bars represent standard deviation. Statistical significance was determined by Brown-Forsythe and Welch ANOVA test followed by a Dunnett’s T3 multiple comparison test for data series that passed Shapiro-Wilk normality test, or Kruskal-Wallis test for data series that did not pass normality test (ns—not significant; ** p < 0.01; **** p < 0.0001).
Fig 5.
Effects of FliC domain swaps between P. aeruginosa PAK and PAO1 strains on swimming motility and filament formation.
(A) Protofilament comparison of PAO1 (blue) and PAK (orange) strains. Outer domains in PAK are farther apart and do not engage in contacts. (B) Swimming motility and negative-stain EM images of the wild type PAO1 and PAK strains. (C) Swimming motility and negative-stain EM images of PAO1-ΔfliC strain complemented with PAK-FliC or FliC in which one or two domains are replaced with equivalent domains of PAK strain. (D) Comparison of average swimming speeds of P. aeruginosa strains in liquid media complemented with increasing concentration of Ficoll 400. Statistical significance for each condition is in S4 Table Blue bars represents the motile spread on semi-solid agar. Red bars represent average swimming speed in liquid medium (right vertical axis). Area of motile spread for each strain represents the average of ten replicates and it is normalized to that of the PAO1 wild type strain (A), or full-length wild type complemented strain (FliCWT) (B). Error bars represent standard deviation. Statistical significance was determined by Brown-Forsythe and Welch ANOVA test followed by a Dunnett’s T3 multiple comparison test for data series that passed Shapiro-Wilk normality test, or Kruskal-Wallis test for data series that did not passed normality test (ns—not significant; * p < 0.05; ** p < 0.01; *** p < 0.001; **** p < 0.0001).
Fig 6.
Effects of FliC domain swaps between FliC of P. aeruginosa PAO1 and S. Typhimurium on swimming motility.
(A) Left–Swimming motility analysis and negative-stain EM images of PAO1-ΔfliC strain complemented with FliC in which PAO1 domains (blue) or secondary structural elements are replaced with equivalent regions of S. Typhimurium (green). Right–Secondary structural elements replaced presented on the structure of PAO1-FliC. BH—beta-hairpin. Area of motile spread for each strain represents the average of ten replicates and it is normalized to that of the full-length wild type complemented strain(FliCWT). Error bars represent standard deviation. Statistical significance was determined by Brown-Forsythe and Welch ANOVA test followed by a Dunnett’s T3 multiple comparison test (*** p < 0.001). (B) Cross-section of PAO1 filament showing position of helix 5 in the filament core and sequence alignment of helix 5 from PAO1 and S. Typhimurium. Blue—residues that are identical between PAO1 and S. Typhimurium. Green—conservative substitutions; yellow and red—non-conservative substitutions.
Fig 7.
PAO1-like flagellin fold predicted by AlphaFold in different bacterial species.
(A) Superposition of D2 and D3 domains of C. jejuni flagellin FlaA and PAO1-FliC. (B) Superposition of FliC molecules of Pseudomonas species other than P. aeruginosa. (C) Superpositions of flagellin outer domains of non-Pseudomonas species. Unstructured loop regions were removed for clarity. (D) Phylogenetic trees based on 16S rRNA (top) and flagellin outer-domain protein sequences (bottom). Species are grouped into bacterial phyla according to the currently accepted ICNP nomenclature.