Communicating the age of the actin cytoskeleton through conformational changes
The active remodeling of the actin cytoskeleton is one of the most dynamic processes in the cell, which is directly coupled to the nucleotide state of the actin filaments. The Raunser lab, for the first time, solved six high-resolution cryo-structures of F-actin in all its nucleotide states. The structures reveal how the state links to conformational dynamics and how this could determine interactions with actin binding proteins.
Actin is the most abundant protein in eukaryotic cells participating in a plethora of protein-protein interactions. It is a key player in many essential cellular functions as cellular motility or cell shape maintenance, e.g. the contraction of cells during cell division. These functions depend strongly on the dynamics of the actin cytoskeleton and its ability to be rapidly remodeled by assembly and disassembly.
Before filament formation, the nucleotide binding site of monomeric globular actin (G-actin) is occupied by ATP. The subsequent polymerization to filamentous actin (F-actin) increases ATP hydrolysis but the resulting phosphate is not released directly. Thus, F-actin passes three nucleotide states during its lifetime: the initial ATP state, the intermediate ADP-Pi state and the final ADP state - the least stable state of the filament. The nucleotide state thereby marks the “age” of filamentous actin and provides a readout for different actin binding proteins that regulate the remodeling and the organization of the actin cytoskeleton. Biochemical and structural studies suggest that the readout process could work by conformational changes at the surface of actin. However, the mechanisms behind these processes are still unclear.
The group of Prof. Dr. Stefan Raunser now revealed the so far missing structural details providing important insights into these mechanisms. Applying electron cryo-microscopy the scientists solved the structures of six different nucleotide states of F-actin in near-atomic resolution. Although very challenging, a reliable model for most of the side-chains of the active site was developed. The model revealed that increased ATP hydrolysis in F-actin can be attributed to a repositioning of a single amino acid - the catalytic base - closer to the -phosphate. The structures of the different nucleotide states demonstrate conformational changes in a loop – the D-loop - at the actin periphery, known to be important for nucleotide-state sensing. The initial and intermediate triphosphate states show a mixture of open and closed D-loops, whereas the final ADP state is locked in the closed confirmation. Adding the actin stabilizing small molecule Jasplakinolide before polymerizing fromADP-actin, locks the filaments in the open state. On the basis of these data it was shown in collaboration with the Bieling group at the MPI Dortmund that interaction of actin binding proteins with actin is conformation dependent. The change from the closed to the open confirmation through Jasplakinolide allowed the actin binding protein Coronin-1B a similar robust binding to the ADP-F-actin as to ADP-Pi-Actin although binding to ADP-F-actin has been reported 50 times weaker than to ADP-Pi-Actin.
“Open and closed states might always be present and their balance might be regulated by the nucleotide bound at the binding site. The connection between the nucleotide-binding site and the periphery of F-actin allows actin-binding protein to directly sense the nucleotide state of the filament.” suggests Stefan Raunser.
JJ/FM