Fig. 5: Fascin cross-bridge structural flexibility mediates interfilament rotations.

a, Histogram of amplitudes along the first multibody refinement eigenvector for all particles (n = 113,800), which follows a unimodal Gaussian distribution (R² = 0.9986). Red lines indicate cutoffs for partitioning particles into three bins. Rotational phase differences between filaments from each reconstructed bin are depicted. b, Side (left) and end-on (right) views of atomic models from eigenvalue binned reconstructions superimposed on F-actin 1. Predominant structural transitions explaining the relative rotation of F-actin 2 are indicated by double-headed arrows. c,d, Structural comparison of eigen_left and eigen_middle atomic models when superimposed either on F-actin 1 (c) or F-actin 2 (d). Rigid-body repositioning of fascin subdomains are indicated by black arrows. e,f, Structural comparison of eigen_middle and eigen_right reconstructions when superimposed either on F-actin 1 (e) or F-actin 2 (f). g, Views of a representative simulated hexagonal bundle element. Actin is shown in gray and fascins are colored by normalized energy scores. h, Contact maps of experimental (n = 5) and simulated (n = 10) hexagonal bundle elements. Models are aligned on the central filament. Each filament contains 16 protomers (A-P) indexed alphabetically from the barbed end to the pointed end. The peripheral filaments are numbered sequentially from 1 to 6 (as arranged in Fig. 4b), whereas the central filament is assigned index 7. Each matrix element depicts a potential fascin cross-linker position in the bundle, whereas color corresponds to the observed number of fascin cross-bridges. a.u., arbitrary units.

Source data