Polymerization of actin filaments is necessary for both protrusion of the leading edge of crawling cells and propulsion of certain intracellular pathogens [1] and it is sufficient for generating force for bacterial motility in vitro [2]. Motile intracellular pathogens are associated with actin-rich comet tails containing many of the same molecular components present in lamellipodia [3] suggesting that these two systems use a similar mechanism for motility However, available structural evidence suggests that the organization of comet tails differs from that of lamellipodia. Actin filaments in lamellipodia form branched arrays [4], which are thought to arise by dendritic nucleation mediated by the Arp2/3 complex [5, 6]. In contrast, comet tails have been variously described as consisting of short, randomly oriented filaments [7], with a higher degree of alignment at the periphery [8], or as containing long, straight axial filaments with a small number of oblique filaments [9]. Because the assembly of pathogen-associated comet tails has been used as a model system for lamellipodial protrusion, it is important to resolve this apparent discrepancy. Here, using a platinum replica approach, we show that actin filament arrays in comet tails in fact have a dendritic organization with the Arp2/3 complex localizing to Y-junctions as in lamellipodia [10]. Thus, comet tails and lamellipodia appear to share a common dendritic nucleation mechanism for protrusive motility. However, comet tails differ from lamellipodia in that their actin filaments are usually twisted and appear to be under significant torsional stress.
