Focal adhesions come unstuck

Focal adhesions come unstuck C rawling cells pick up after themselves before moving on. Ezratty et al. (1) show that cells use the endocytic protein clathrin to reabsorb the integrin receptors that attach them to surfaces, instead of simply leaving the molecules behind. Cells on the go gain traction by forming focal adhesions (FAs), temporary attachments to the surface. But if they can't break these connections, cells get stuck. Scientists know much more about the assembly of FAs than about their disassembly. Microtu-bules appear to spur FA breakdown, possibly by shipping in an unidentifi ed " relaxing factor " (2). Also unclear is what happens to the integrins that bunch up at FAs and fasten to the extracellular matrix. Cells might just crawl away from their integrins (3), leaving them behind like the pitons mountaineers leave in a rock face. But the leading hypothesis holds that cells absorb integrins through endocytosis, possibly allowing them to be reused later. Teasing out the molecular mechanism of FA disassembly has proven diffi cult because cells simultaneously create and break the connections. Four years ago, the researchers devised a way to sort out the two processes (4). They collapsed the cell's micro-tubules with nocodazole, triggering a fl urry of FA formation. Removing the drug prompts the adhesions to break down as the tubules reassemble. Using this same approach, Ezratty et al. tested whether clath-rin—a protein that encapsulates vesicles and promotes endocytosis— takes part in FA breakdown. Not only did clathrin amass at FAs, two of the adaptor molecules that attach clathrin to its targets, Dab2 and ARH, also accumulated there. In cells lacking clathrin, FA disassembly decreased by up to 80%. The process also faltered if the adaptor proteins were absent. And cells missing clathrin or the adaptors stalled; their front ends advanced while their tails remained attached to the surface, stretching the cells out. The researchers observed that in-tegrins abandoned the cell surface upon FA disassembly. Surface levels of the ␣5 integrin subunit plunged 50% after nocodazole removal, for instance. These absconding integrins showed up in vesi-cles carrying the small GTPases Rab5 and Rab11, evidence that the receptors had been absorbed through the endocytic pathway. And the cell surface levels of integrins that were not part of cell attachments didn't change, showing that endocytosis was specifi c to FAs. Clathrin's role as a shepherd for integrins became clear when the team observed the …