Invadopodia are specialized, proteolytically active, plasma membrane protrusions developed by transformed and invasive tumor cells grown on a flat extracellular matrix (ECM) substratum. Invadopodia are actin-rich protrusions with focal ECM degradation activity and are typically enriched in integrins, tyrosine kinase activity, actin-associated proteins and proteases including matrix metalloproteases. We recently found that dynamin 2, a large GTPase involved in intracellular trafficking and actin dynamics, is required for ECM degradation at invadopodia. Human cortactin is encoded by the EMS1 gene on ch11q13, often amplified in certain carcinomas; although the role of cortactin in tumorigenesis remains unclear, cortactin amplification or overexpression are associated with poor prognosis. In vitro, cortactin overexpression increases cell motility and invasion and is involved in invadopodia formation. Cortactin is an important F-actin stabilizer and activator of the master actin nucleator Arp2/3 complex, via interaction with the Wiskott-Aldrich syndrome protein N-WASP. It also couples actin dynamics to intracellular trafficking via its binding partner dynamin 2. Furthermore, cortactin is also a substrate of Erk and Src kinases. A switch on/off mechanism has been proposed whereby Erk phosphorylation on cortactin serines activates N-WASP and thus actin nucleation by Arp2/3 whereas Src phosphorylation on tyrosines terminates activation of N-WASP. We find that cortactin depletion by siRNA blocks invadopodia formation and hence, ECM degradation. Also, the overexpression of deleted or point mutant forms of cortactin that prevent the binding of specific partners such as Arp2/3 complex, F-actin, N-WASP or dynamin 2, also significantly decreased invadopodia formation and ECM degradation. Finally, we investigated the role of cortactin phosphorylation through the use of two cortactin phosphorylation mutants. One, in which the three Src-phosphorylated tyrosines were changed to phenylalanine; the other in which the two Erk-phosphorylated serines were substituted by aspartic acid, thus mimicking Erk-phosphorylated cortactin. Overexpression of tyrosine-phosphorylation-deficient mutant, lead to a considerable reduction in invadopodia formation and ECM degradation. In contrast, overexpression of the Erk-phosphorylated-like cortactin produced a remarkable increase in ECM degradation. We conclude that cortactin is absolutely required for invadopodia formation. In addition, cortactin must be modulated by phosphorylation in tyrosine and serine to constitute a functional ECM degradation machinery.

[Proc Amer Assoc Cancer Res, Volume 47, 2006]