The Mechanism of Actin Remodeling by the BAR Domain of ArfGAP ASAP1

Abstract

The actin cytoskeleton plays an important role in cellular processes. Normal processes, such as differentiation, use the actin cytoskeleton as a signaling and transport platform, while aberrant processes, such as cancer, repurpose actin for invasion. Actin remodeling involves assembly of globular actin into actin filaments and further assembly of filaments into bundles, which form higher order structures such as stress fibers, lamellipodia and filopodia. A given higher order actin structure relies on organized, timely and regulated mobilization of actin filaments and accessory actin binding and bundling proteins. For one of such structures, the actin stress fibers (SFs), assembly, maturation and maintenance depends on a carefully orchestrated crosstalk between actin filaments, actin binding proteins and the focal adhesion machinery. As a biological phenomenon, SFs provide a mechanotransduction platform, and as a model system, they provide a prototype of the mechanism and the regulation of actin filament assembly. The complexity and diversity of actin structures is, in part, dictated by the actin regulating proteins. Among these are well established actin remodeling families, such as actinin and fascin protein families. More recently a subset of BAR‐domain proteins, widely studied for inducing and sensing the membrane curvature, have been found to control actin filament assembly. However, the mechanism by which BAR domain proteins regulate of the actin cytoskeleton remains unclear. We recently showed that ASAP1, a BAR‐domain containing ArfGAP, remodels actin filaments. Depletion of ASAP1 led to reduction in overall levels of filamentous actin and the number of ECM‐interacting stress fibers. Our structure‐function studies using in vitro actin sedimentation, fluorescence microscopy, as well as cellular assays revealed that the BAR‐PH region of ASAP1 is sufficient and necessary to induce actin reorganization. During the study we observed that another member of the human ArfGAP family that also contains a BAR domain, ACAP1, does not induce the same phenotype as ASAP1. Our current work thus inquires whether structural differences between the BAR domains of ASAP and ACAP explain their disparate effects on actin organization. Molecular modeling‐mediated dissection of the mechanism of actin binding localizes the actin binding site to the 1st and 2nd helix of the BAR domain of ASAP1. Together these data highlight selectivity of the BAR domain proteins in actin remodeling and provide further insights into the different roles human ArfGAPs play in modulating the actin cytoskeleton.Support or Funding InformationThis work was supported by the Intramural Research Program of the National Cancer Institute, Department of Health and Human Services, project number BC007365