page 515) show that β1 and β3 integrins direct different types of motility, despite binding fibronectin with similar efficiencies. Cells expressing β1 integrin move in a random manner, whereas cells with β3 integrin expression move in a persistent fashion. The proteins control cell migration by altering the activities of the Rho GTPases RhoA and Rac, which in turn control three aspects of motility: lamellipodia formation; cytoskeletal polarization; and the dynamics of cell matrix adhesions.
The types of integrins expressed and their relative abundance change during dynamic processes such as wound healing and development. To understand the function of these shifts, Danen et al. studied GE11 cells that lack β1 and express trace amounts of β3. Previous work demonstrated these cells don't migrate well in a wound-healing assay.
When GE11 cells were transduced with β3 (GEβ3), they formed lamellipodia, had active cofilin, and migrated persistently. Cells transduced with β1 integrin (GEβ1) migrated in a wound healing assay but contained many actin stress fibers, phosphorylated inactive cofilin, and multiple random protrusions rather than a well-organized lamellipod. Inhibition of RhoA—which in its active form promotes stress fiber formation—activated cofilin in GEβ1 cells, restoring persistent migration. By contrast, overexpression of Rac, which supports lamellipodia formation, was not sufficient to promote persistent migration in GEβ1 cells. Thus the problem in GEβ1 cells seems to be too many stress fibers rather than too little lamellipodia formation.
Just how the different β integrins connect to distinct Rho GTPases to control migration is not clear. Switching the intracellular tails between integrins doesn't change their effect on Rho GTPases (Danen et al. 2002. J. Cell Biol. 159:1071–1086), so the researchers think that the signaling differences result from different integrins working with different coreceptors. Syndecans may fit that role—they interact functionally with integrins and they regulate Rho GTPases.