Cell crawling two decades after Abercrombie

In response to extracellular signals, cells remodel actin networks. Monomeric actin subunits at the cell's leading edge assemble into linear polymers that are cross-linked by accessory proteins into three-dimensional structures that are contracted by myosins to generate hydraulic force; elsewhere in the cell, actin networks dismantle. Actin subunit sequestering proteins prevent spontaneous actin nucleation, but not the growth of actin sub-units on to fast-growing filament ('barbed') ends, and at least half of the actin in most cells is filamentous. Therefore regulation of cellular actin assembly also requires proteins that block ('cap') actin filament barbed ends. Members of the capping protein gelsolin family also sever actin filaments mechanically. Calcium and protons activate gelsolin for severing and capping. Phosphoinositides reverse such capping, and a pathway has been defined in which receptor perturbation operates through GTP-Rac1 to stimulate the synthesis of endogenous phosphoinositides that uncap actin filaments. Other GTPases (and other signalling pathways) target phosphoinositide synthesis where other protrusions (e.g. filopodia) emerge. Cells maintain adequate, albeit compromised, locomotion in the absence of some, but not all, important machine parts. For example, gelsolin-null fibroblasts crawl using predominantly filopodia rather than lamellae. However, ABP-280 (actin-binding protein of 280 kDa), which promotes orthogonal branching of short actin filaments, seems to be necessary for membrane stability and translational locomotion. ABP-null cells hardly crawl at all, although they are viable and engage in surface movements.