Abstract
How the actin cytoskeleton is organized in a cell is critical for regulation of cell migration, maintenance of cell shape, and endocytosis. In a migrating cell, actin exists in two organizational domains: a highly dynamic and branched network in the lamellipodia, extending 1-3 μm in from the cell edge; and a stable, bundled actin network in the lamella that exists further back from the lamellipodia, towards the nucleus. Bundled actin in the lamella associates with non-muscle myosin II to form an actomyosin network, which consists of three types of structures: transverse arcs, dorsal stress fibers and ventral stress fibers. Proper organization of the actomyosin network regulates rearward flow of transverse arcs towards the nucleus – termed actomyosin retrograde flow. Actomyosin network organization can also facilitate engagement with and maturation of focal adhesions, multi-protein complexes linking actin network and the underlying substratum.
An overarching goal has been to elucidate the relationship between actin networks in the lamellipodia and lamella and to clearly understand how the lamellar actomyosin networks are generated. One idea postulates that actin in the lamellipodia disassembles completely and lamellar networks are then synthesized de novo. Other evidence however argues that a component of the lamellipodial actin contributes to generation of the lamella. An outstanding goal of this research is to understand the mechanisms by which lamellipodial actin filaments are efficiently remodeled from their branched organization within the lamellipodia to one of parallel filament bundles in the lamella and to elucidate the proteins involved in this process.
Previous work in our lab and by others had uncovered an unconventional actin-regulatory protein, dynamin2. In vitro studies demonstrated a role for dynamin2 in bundling and reorganization of, or remodeling, of F-actin and studies in fixed cells implicated dynamin2 in actomyosin organization. However, to elucidate how dynamin2 regulates the dynamic process of actomyosin assembly, I used live-cell imaging and siRNA-mediated depletion of endogenous dynamin2 to investigate a function for dynamin2 in remodeling lamellipodial and lamellar actin networks in vivo. Nascent myosin puncta appeared at an increased rate during actomyosin assembly in dynamin2-depleted U2-OS cells and organized poorly at the lamellipodium-lamellum boundary. Also, the rate of myosin retrograde flow was high compared with that of control siRNA-treated U2-OS cells. These indicated a poorly organized actomyosin network. However, dynamin2 did not localize to the actomyosin network. Instead, GFP-dynamin2 localized to the distal lamellipodial edge in live cells and regulated the spatiotemporal distribution of the actin cross-linker α-actinin at the lamellipodia. Therefore, I postulated a function for dynamin2 in reorganizing lamellipodial F-actin to set up an actin architecture that creates an ideal template for actomyosin formation within the lamellum.
In order to elucidate a mechanism for how dynamin2 influenced lamellipodial F-actin, various fluorescently-tagged dynamin2 mutants were assessed for rescue of the defects in lamellar actomyosin and lamellipodial α-actinin. Direct interaction of dynamin2 with F-actin and C-terminal proline rich domain (PRD) and the rate of dynamin2-dependent GTP hydrolysis were critical for regulating the spatiotemporal distribution of lamellipodial α-actinin and for rescuing high actomyosin retrograde flow. Importantly, to rule out secondary effects on actin networks via defects in endocytosis, I demonstrated that dynamin2 depletion did not affect internalization of integrin β1 and transferrin receptor. This work establishes dynamin2 as a critical regulator of global actin networks. This novel function for dynamin2 can have far-reaching implications in regulating the actin cytoskeleton for important functions like cytokinesis, maintaining cellular junctions and membrane modulations during endocytosis
