Cryo-EM structures of both ends of the actin filament explain why the barbed end elongates faster than the pointed end

Abstract

Actin filament ends are the sites of subunit addition during elongation and subunit loss during depolymerization. Prior work established the kinetics and thermodynamics of the assembly reactions at both ends but not the structural basis of their differences. Cryo-EM reconstructions of the barbed end at 3.1 [A] resolution and the pointed end at 3.5 [A] reveal distinct conformations at the two ends. These conformations explain why barbed ends elongate faster than pointed ends and why pointed ends rapidly dissociate the {gamma}-phosphate released from ATP hydrolysis during assembly. The D-loop of the penultimate subunit at the pointed end is folded onto the terminal subunit, precluding its binding incoming actin monomers, and gates on the phosphate release channels of both subunits are wide open. The samples were prepared with FH2 dimers from fission yeast formin Cdc12. The barbed end reconstruction has extra density that may be partial occupancy by the FH2 domains. Significance StatementCells depend cytoplasmic filaments assembled from the protein actin for their physical integrity, as tracks for myosin motor proteins and movements of the whole cell and internal organelles. Actin filaments elongate and shrink at their ends by adding or dissociating single actin molecules. We used cryo-electron microscopy to determine the structures of the two ends of actin filaments at 3.5 [A] resolution for the slowly growing pointed end and 3.1 [A] for the rapidly growing barbed end. These structures reveal why barbed ends grow faster than the pointed ends, why the rate at the pointed end is not diffusion-limited and why the pointed end has a low affinity for the {gamma}-phosphate released from bound ATP inside the filament.