page 409, address this question with a detailed kinetic and structural analysis, presenting a novel model that should help to guide future work in the field.
Using GFP-tagged versions of the Arp2/3, VASP, and fascin proteins and innovative electron microscopy techniques, the authors examined the molecular dynamics of filopodium initiation in mouse melanoma cells. Rather than forming a distinct nucleation complex, Arp2/3 seems to lay the groundwork for filopodium birth by producing a normal lamellipodial actin array. Conical structures, termed Λ-precursors, then rise up from this array. VASP gradually coalesces from a broad distribution along the lamellipodial edge into defined spots at the tips of the Λ-precursors. Fascin then abruptly appears at the Λ-precursor vertex before filopodia grow from these structures.
The data suggest that, in a normal dendritic actin network in the lamellipodium, the elongating barbed ends of certain actin filaments bind to a set of molecules that protect them from capping and mediate their association with other elongating barbed ends. When these privileged filaments collide, they merge to form the conical Λ-precursors, and their tip complexes then recruit or activate the bundling protein fascin, allowing filament cross-linking and the growth of long, strong actin bundles characteristic of filopodia. This model also implies that microspikes, and retraction fibers—filopodium-like structures seen in some migrating cells—arise by a similar mechanism and are interconvertible with filopodia.
In separate work, the authors have developed an in vitro system that mimics this convergent elongation process. They are now trying to identify proteins in the filopodial tip complex that regulate initiation. ▪