Polarised apical-like intracellular sorting and trafficking regulates invadopodia formation and degradation of the extracellular matrix in cancer cells

Abstract

Invadopodia are proteolytically active protrusions formed by invasive tumoral cells when grown on an extracellular matrix (ECM) substratum. A current challenge is to understand how proteolytic activity is so precisely localised at discrete sites of the plasma membrane to produce focalised ECM degradation at invadopodia. Indeed, a number of components including metalloproteases need to be directed to invadopodia to ensure proper segregation of proteolytic activities. We recently found invadopodia to feature the properties of cholesterol-rich membrane domains (a.k.a. lipid drafts) and that ECM degradation depends on the tight control of cholesterol homeostasis. Since apically directed polarised sorting and transport in epithelial cells relies on segregation of proteins into lipid rafts at the Golgi complex, we hypothesised that invadopodia-dependent ECM degradation might also rely on lipid raft-dependent polarised transport routes. To investigate this issue we undertook a three-pronged approach. First, we found that microtubule depolymerisation, which is known to disrupt polarised transport in polarised cells, strongly inhibited invadopodia formation, while not affecting overall protein transport. In the second approach we found that glycosylphosphatidylinositol-anchored green fluorescent protein (an apical model protein), but not vesicular stomatitis virus G-protein or influenza virus hemagglutinin (both model basolateral model cargoes), was transported to sites of ECM degradation. Finally, RNAi-mediated knock-down of proteins known to specifically regulate polarised apical or basolateral transport in epithelial cells, such as caveolin 1 and annexin XIIIB or clathrin, respectively, demonstrated that the selective inhibition of the apical, but not the basolateral, transport route impairs invadopodia formation and ECM degradation. Taken together, our findings suggest that invadopodia are apical-like membrane domains, where signal transduction and local membrane remodelling events might be temporally and spatially confined via selective raft-dependent apical transport routes.