Frealix: model-based refinement of helical filament structures from electron micrographs

Area of Science:

• Structural Biology
• Biophysics
• Cryo-Electron Microscopy

Background:

• Helical protein filaments are often studied using electron microscopy (EM) of vitrified samples.
• The 'single-particle' approach is successful for rigid filaments but struggles with flexible or curved structures like amyloid-β (Aβ) fibrils due to alignment inaccuracies and low contrast.
• Accurate structural determination of flexible biological filaments is crucial for understanding their function and associated diseases.

Purpose of the Study:

• To develop a novel computational method for reconstructing the 3D structures of flexible helical protein filaments from EM data.
• To address the limitations of the single-particle approach in handling curved and low-contrast filament segments.
• To enable detailed characterization of filament deformations and evaluate structural models.

Main Methods:

• Development of new software, Frealix, which aligns arbitrarily short filament segments simultaneously.
• Implementation of spatial constraints to ensure continuity of aligned segments, forming a coherent helical structure.
• Benchmarking Frealix against datasets of amyloid-β (Aβ(1-40)) fibrils and tobacco mosaic virus (TMV).

Main Results:

• Frealix achieves comparable resolution to single-particle analysis for well-ordered TMV.
• For Aβ(1-40) fibrils, Frealix provides reliable alignments and reconstructions at ~8-Å resolution, even from curved filaments.
• The method successfully characterizes three-dimensional filament deformations and allows evaluation of the worm-like chain model.

Conclusions:

• Frealix offers a robust solution for reconstructing the 3D structures of flexible and curved biological filaments from EM data.
• This advancement improves the structural analysis of challenging samples, such as amyloid fibrils, previously difficult to resolve accurately.
• The software facilitates detailed insights into filament dynamics and structural heterogeneity.