page 429 by Benink and Bement.
A wounded cell, such as the frog oocyte system used by the authors, rapidly repairs its broken membrane by an onslaught of exocytosis. After sealing the hole, the cell must rebuild the actin cytoskeleton underneath the new membrane. This is partly accomplished by stretching the undamaged surrounding cytoskeleton inward over the wound, which requires actomyosin-based contraction.
The authors now see that this inward motion of actin is coordinated by ring-like patterns of the active form of two rho GTPase family members known to regulate actin dynamics. The appearance of these rings—an inner loop of RhoA-GTP circumscribed by a halo of Cdc42-GTP—preceded actin accumulation and occurred independently of actin assembly. Inward movement of the GTPases as the wound was repaired, however, depended on the F-actin array.
On the ring's inside, RhoA turned on contraction by activating myosin-2 light chain phosphorylation. At the outer edges, Cdc42 promoted actin turnover, probably via WASP, and possibly also deactivated myosin-2. The resulting relaxation of the actin array on the outskirts may make actin's inward motion easier. In fact, if the authors used constitutively active RhoA to prevent this relaxation, the actin array fractured as it tried to pull forward against high tension.
The upstream signals that set up these GTPase patterns are not known. Exocytosis may be the early spark, as extracellular calcium, whose entry triggers the exocytic patching of wounded membrane, was also needed to establish the GTPase rings. RhoA negatively regulated Cdc42 activity and so is at least partly responsible for the separation of the two rings. Microtubules were also needed for ring segregation, but how they are contributing is not yet clear.
Actomyosin-based contraction also controls cytokinesis and multicellular migration during wound healing in tissues. The authors expect that similar segregated patterns of active rho GTPases coordinate these processes as well.