Promoter Cleavage: A TopoIIβ and PARP-1 Collaboration

dynamic adhesions within the cell. The data from Gupton and Waterman-Storer (2006) suggest that low fibronectin is suboptimal for migration because there are too few adhesions even though the adhesions are dynamic. High fibronectin is suboptimal because the adhesions are too stable and do not signal. Cells on intermediate fibronectin have a high number of small, dynamic adhesions of the kind that are known to signal to Rac and Cdc42. Thus, the overall rates of both integrin ligation and adhesion assembly should be maximal, leading to high Rac and Cdc42 activity. These simple dynamics may be the key factor driving lamellipodial actin polymerization and flow. The differential effects on Rho activity can also be understood if we consider integrin activation of Rho, which is temporally biphasic (Ren et al., 1999). Following new integrin ligation, Rho activity first decreases, followed by an increase at later times. Thus, young adhesions appear to inhibit Rho, whereas more mature adhesions activate it. As adhesion lifetimes increase in direct proportion to the concentration of fibronectin, these results may explain why Rho activity correlates with fibronectin. This study has revealed some major insights and represents a large step in our understanding of cell migration. However, there are more chapters to be written. The actinintegrin linkage is a critical site that integrates adhesion, signaling, and protrusion; its regulation remains to be parsed. The “clutch” mechanisms by which myosin II-mediated tension regulates both adhesion assembly and disassembly may depend on the state of the adhesion, which remains to be defined. Finally, some highly motile cells do not show the highly organized adhesions and actin filaments seen in fibroblasts or epithelial cells migrating on fibronectin yet can show optima in migration speed that are dependent on substrate concentration. The factors that produce this relationship in these cells remain to be established. RefeRences