Dogma bites back – the evidence for branched actin

The review about actin filament branching from Victor Small describes an elegant and compelling set of results that offer a valuable counterpoint to the status quo. However, the assertion that ‘the validity of the dendritic nucleation model rests on the assumption that the electron microscope images…faithfully reflect the structure of lamellipodia actin networks’ is not correct. As is clearly stated in the original description of the dendritic nucleation model, referenced by Small, the dendritic model is based on biochemistry and electron microscopy of purified material. Pure Arp2/3 complex does not nucleate new actin filaments on its own – it requires the presence of existing actin filaments as a template and. Total internal reflection fluorescence (TIRF) videomicroscopy of nucleation paints a more detailed picture of the same process and. If this well-supported finding is true, and Small provides nothing to contradict it, it makes some form of dendritic array inevitable. Although the apparent agreement with electron micrographs of intact lamellipodia was felicitous, it is not essential to the model. A dendritic network of consistent size that is extending at a constant rate will have, on average, one branch per complete actin filament. Small and colleagues observe extremely long actin filaments (some >1 μm). This means that branches, should they exist, will be widely spaced. Such long filaments must also have taken a significant time to assemble. Therefore, if there is any significant rate of dissociation of branches (‘debranching’) during cell movement (or even during fixation and sample preparation), then observable branches will be rarer still. I suggest that Small and colleagues’ excellent electron micrographs and tomograms imply that lamellipodial actin forms in a dendritic network, but one that is less dense than was implied by earlier work, and which undergoes significant debranching as it evolves. This explanation is more parsimonious – and contradicts far less of the literature – than complete rejection of dendritic growth