Towards a structural understanding of the remodeling of the actin cytoskeleton

Figures

• 

  Fig. 1. Structural transition upon actin polymerization. (A) Illustration of the actin domain architecture consisting of four subdomains (SD) and a central nucleotide binding cleft. (B) During polymerization actin undergoes a transition from its monomeric (G-actin) to filamentous (F-actin) form. In F-actin SD1 and SD2 (yellow) are rotated ∼20˚ relative to the other domains (gray) closing the central nucleotide-binding cleft. (C) Close-up view into the nucleotide-binding cleft of Mg-ATP-G-actin (PDB: 1NM1) and F-actin bound to either Mg-AppNHp (PDB:5OOE), Mg-ADP-BeFx (PDB: 5OOF) or Mg-ADP-Pi (PDB: 6FHL). The crystal structure of G-actin (top left panel) includes two water molecules (red balls) which are essential for nucleotide hydrolysis. Wnuc is responsible for the nucleophilic attack and Wbridge bridges the catalytic base with Wnuc. Upon the transition to F-actin H161 rotates taking the position of Wbridge and Q137 moves slightly towards the nucleotide. Waters and H161 of ATP G-actin are shown transparently in the other tiles to emphasize the structural transition. As BeFx is not included into the atomic model, its corresponding density (EMDB: 3839) is shown instead (gray mesh). Actin…

• 

  Fig. 2. Schematic overview of actin polymerization, filament ageing and associated helical symmetry. (A) Actin polymerizes into a helix that can be either described as a right-handed, double-stranded (long pitch) or a left-handed single start (short pitch) filament. The colors were chosen to highlight individual strands of F-actin while the numbering represents consecutive subunits along the short pitch helix. (B) ATP-G-actin adds to the barbed (+) end of the filament undergoing a structural transition to F-actin which includes flattening of its structure and repositioning of side chains essential for catalysis, especially H161 (also see Fig. 1). F-actin has a characteristic nucleotide gradient along the filament, marking the local age of the filament. The age of the filament is transmitted to an essential interaction site at the surface which is made up by the d-loop (shades of yellow) of one subunit and the C-terminal tail of an adjacent subunit (gray). While both the open and the closed d-loop conformation coexist before and after hydrolysis, the occupancy of the closed state increases with age until phosphate release eventually locks the filament in the closed d-loop. (C) Cryo-EM…

• 

  Fig. 3. Schematic overview of the dendritic nucleation model of actin. The dynamic remodeling of the actin cytoskeleton requires a tight spatial and temporal organization by ABPs. Proteins like capping protein (red) and the Arp2/3 complex (purple) remodel the cytoskeleton by either capping the barbed end or by introducing a branch point from which a new filament can grow. Other proteins like ADF/cofilin (green) and profilin (blue) depolymerize and sever actin or maintain a constant pool of polymerizable actin, respectively. The ATPase activity of actin itself is also essential for the regulation as it results in a nucleotide gradient along the filament which can be sensed by accessory proteins (also see Fig. 2). Actin is depicted in shades of yellow depending on its nucleotide state. The color gradient represents the nucleotide gradient along the filament from the darkest shade (ATP) to brightest shade (ADP). The figure is based on [112].

• 

  Fig. 4. Inhibitors and depolymerizers of F-actin. (A) Profilin (blue, PDB:2BTF) binds and blocks the barbed end of monomeric actin (G-actin). At the same time, profilin promotes the conversion from ADP-G-actin to ATP-G-actin to build a pool of polymerizable actin. (B) Capping protein (red, PDB: 5ADX) binds to the barbed end of the actin filament. It consists of two subunits, α and β, which both have C-terminal tentacles that block the barbed end and stop elongation of actin filaments. (C) Cofilin (green, PDB: 5YU8) preferentially binds to ADP-bound F-actin, where it locates at the interface of two actin subunits within one strand. It changes the helical twist and weakens the interactions within F-actin, eventually depolymerizing and severing actin. As the d-loop of the original model was fragmented, the lower actin subunit in the right panel was replaced by one from PDB: 5ONV to illustrate the proximity of the N-terminus of cofilin (residues missing in shown model are represented schematically as a dashed green line) and the d-loop. Actin is depicted in shades of yellow depending on the nucleotide state (see Fig. 3 for color legend).

• 

  Fig. 5. Nucleators of F-actin. (A) The Arp2/3 complex (shades of purple, PDB: 4JD2) binds to the side of actin filaments, creating a branch point which acts as a seed for a branch filament at an angle of ∼70˚. (Left panel) Docking of atomic models of the Arp2/3 complex and F-actin (PDB: 5JLF) into an in situ sub-tomogram average structure (transparent gray, EMDB: 4790). (Middle and right panel) Illustration of domain architecture of the Arp2/3 complex. Arp2 and Arp3 (shades of red) rearrange upon binding to F-actin forming a seed for a new filament. (B) Formins like mDia1 (shades of cyan, model from [34]) bind to the barbed end of actin filaments as a dimer. Their FH2 domains enclose the filament. Their FH1 domains (indicated by dashed lines) are located towards the periphery, where they bind profilin/actin complexes. Actin is depicted in shades of yellow.

• 

  Fig. 6. Architecture and working cycle of myosin. (A) The motor domain of myosin consists of four major subdomains. The upper (dark green) and lower 50 kDa (khaki) subdomains make up the actin-binding interface and form a central cleft that can open and close. The N-terminal domain (olive green) sits at the hinge region and transmits the nucleotide state in concert with the converter domain (brown) along the lever arm of myosin (brown). (B) Simplified working cycle of myosin illustrating the dependence of the lever arm position and interactions with F-actin on the bound nucleotide (light blue). The lever arm position in the ADP-Pi state relative to the ADP state is not known and thus shown transparently. Cryo-EM structures of (C) myosin NM-2C (EMDB: 8165 PDB: 5JLH) and (D) myosin 1B in the rigor state (EMDB: 7331 PDB: 6C1H). The rigor states of different myosin isoforms do not superimpose (lever arm helix of myosin NM-2C in rigor state shown in transparent gray for comparison). (E) Cryo-EM structure of myosin 1B in complex with ADP (EMDB: 7329 PDB: 6C1D). The lever arm in the ADP state does not superimpose with the one in rigor state (transparent gray) suggesting that the power stroke…