Figure 5. The sheath amplifies flagellar curvature to enable motility in the spirochete Leptospira spp.

(A) Histogram of wild-type filament curvatures derived from 3D filament trajectories. A minority of filaments presumed to have shed some or all of the sheath (see Figure 5—figure supplement 1), as judged by a smaller measured diameter, were excluded from this analysis. (B) Histogram of fcpB^– mutant filament curvatures. As in A, a minority population of smaller-diameter filaments were excluded. (C) Histogram of fcpA^– mutant filament curvatures, subdivided into distributions for the larger-diameter population (see Figure 5—figure supplement 1B, 3rd panel) and the smaller-diameter population (see Figure 5—figure supplement 1B, 4th panel). (D–E) Model for sheath-enforced curvature in the Leptospira flagellar filament; inherent curvature is amplified due to binding of FcpA and FcpB along the convex side of the core. (F) Model depicting how sheath-enforced flagellar curvature would interact with the coiled body in Leptospira to generate large-scale curved deformations in the body. (G) If flagellar curvature is reduced, the flagellum can pass through the body helix without deforming it, so filament rotation would not directly induce body deformations (except due to rolling and/or sliding friction against the cell cylinder).

Figure 5—figure supplement 1. Shedding of sheath layers observed in wild-type and mutant L. biflexa flagellar filaments.

(A) Cryo-EM images of wild-type and mutant L. biflexa flagellar filaments that exhibit shedding of sheath layers, either on the inner (red triangles) or outer (green triangles) curvature. Shedding was only rarely seen in wild-type and fcpB^– filaments (Table 3) and could be observed on both inner and outer curvatures, sometimes in the same filament. In contrast, fcpA^– filaments usually lost sheath layers from the inner curvature and concurrent shedding on both inner and outer curvatures in this mutant was extremely rare (Table 3). (B) 2D class averaging of filament subtomogram segments reveals distinct filament diameters and/or curvature for wild-type vs. mutant flagella. Wild-type flagella subtomogram 2D classes mainly yielded images with a diameter of ~250 Å. Subtomogram 2D classes from fcpB^– samples have a smaller diameter (~95% of wild-type for the dominant class), while classes from fcpA^– samples are narrower still (66%–87% of wild-type). Multiple classes of different radii were observed in all samples wild-type and mutant; only the most common ones are shown. Curvature is most evident in the wild-type and fcpB^– samples, while fcpA^– filament class averages are straighter; the narrowest class (bottom panel), which likely represents the core FlaB assembly absent a sheath, shows little or no evidence of curvature. Classes similar to the bottom panel were also observed in the wild-type and fcpB^– samples.

Figure 5—figure supplement 2. Shedding of sheath layers in wild-type L. biflexa flagellar filaments coincides with loss of curvature.

Shown are three examples where a transition (marked by green and red triangles, as in Figure 5—figure supplement 1A) from thick (blue dashed lines) to thin (yellow dashed lines) filament diameter co-localizes with a transition to a less tightly curved supercoil.

Figure 5—figure supplement 3. Supercoiling parameters of wild-type and fcpB^– mutant flagellar filaments estimated from subtomogram average structures.

(A) Superhelical pitch and diameter estimated for wild-type filament (see Materials and methods). 3D rendering of a full superhelical turn was generated from a composite of ~30 copies of the subtomogram average. (B) Superhelical pitch and diameter of the fcpB^- mutant, similar to A.

Figure 5—figure supplement 4. Sinusoidal supercoil forms observed in the fcpA^- sample.

(A, B) Two examples of sinusoidally oscillating filament supercoils imaged by cryo-EM for the fcpA^- sample. Sinusoidal oscillations were rare or nonexistent in the wild-type and fcpB^- samples, but were occasionally observed in fcpA^- samples. To approximate the superhelical pitch and supercoil diameter, each of the imaged filaments was extrapolated (right) to a full wavelength representation by manual superposition of rotated and mirror-image copies of itself. The resulting, supercoil parameter estimates fall within the range of values observed for polymorphic flagellar forms in exoflagellates (|superhelical pitch| ~ 0–2.3 µm; |supercoil diameter| ~ 0–2.7 µm) (Kamiya and Asakura, 1976). The handedness of the fcpA^- supercoils was not determined.