Disassembly of focal adhesions — often a key step in cell migration — has proved difficult to dissect, in part because it is not so easy an event to study in isolation. In the June issue of Nature Cell Biology, Gregg Gundersen and colleagues report the development of a new model system for watching this process and find that adhesions do not simply retrace their steps when they break down but instead use a unique set of factors, one of which is dynamin.

Focal adhesions are dynamic protein complexes containing clustered integrins, which link the extracellular matrix and the internal actin cytoskeleton. In many cell types, they provide traction during migration and so their turnover must be tightly regulated. Microtubule targeting of focal adhesions is known to trigger their disassembly, and it has been proposed that this might occur through inhibition of the GTPase RhoA, whose activation is a key step in adhesion assembly. Gundersen and colleagues have found that this is not the case.

They first devised an elegant assay for watching adhesion disassembly, clearly separated from assembly. They allowed focal adhesions to assemble by depolymerizing microtubules with nocodazole and then initiated microtubule regrowth by washing the nocodazole out. This produced a population of focal adhesions synchronously undergoing disassembly in an ordered and reversible way. Using TIRF microscopy, they confirmed that disassembly correlated with the targeting of microtubules to these sites.

Surprisingly, disassembly was not affected by the activity of RhoA, suggesting that the process is not simply 'assembly in reverse'. Rather, the authors found that dynamin — a large GTPase essential for endocytosis — delivered the fatal blow. Disrupting dynamin, through expression of a dominant-negative construct or siRNA, inhibited disassembly and also reduced cell migration in a wound healing assay.

So how might dynamin lead to the demise of adhesions? Gundersen and colleagues found that, consistent with previous work, focal adhesion kinase (FAK) is important for microtubule-induced adhesion breakdown and that it is an important cohort of dynamin. Endogenous FAK coimmunoprecipitates with dynamin during the early stages of adhesion breakdown and in FAK^-/- fibroblasts, dynamin is no longer found in focal adhesions. They went on to implicate the adaptor Grb2 as a possible bridge between FAK and dynamin.

Given that dynamin is known to regulate endocytosis, it is tempting to speculate that this is how it induces adhesion breakdown. This would also be consistent with previous observations that integrins are endocytosed in migrating cells, and similar to the mechanism by which cadherin-based adherens junctions are removed. Alternatively, dynamin could affect adhesions by altering actin assembly, one of its known talents. With this new assay to hand, it should now be feasible to address this, and thus unravel the intricacies of dynamin's destructive power over adhesions.

Author: Alison Schuldt