Sub-membrane actin rings compartmentalize the plasma membrane

Abstract

The compartmentalization of the plasma membrane is a fundamental feature of cells. The diffusivity of membrane proteins in the plane of the membrane is significantly lower in cells than observed in artificial membranes. This seems due to the sub-membranous actin cortex, but tractable assays to prove a direct dependence of membrane protein motion on actin filaments have been elusive. We recently showed that a periodic array of sub-membrane actin filaments that surround neuronal axons in ring-like structures in the axonal initial segment (AIS) confines membrane protein motion between them, but the local enrichment of ion channels may offer an alternative explanation for compartmentalization. Here we show using computational modeling that in contrast to actin rings, the dense array of ion channels in the AIS cannot mediate membrane protein confinement. We go on to show that indeed the actin rings are closely apposed to the plasma membrane, and that they confine membrane protein motion between them in a [~] 200 nm periodic pattern in progenitor-derived neuronal cells. We find that this compartmentalization is also detectable for inner-leaflet membrane proteins and multi-spanning receptors. Strikingly, several cell types with actin rings in their protrusions, like progenitor-derived astrocytes and oligodendrocytes, also exhibit lateral confinement of membrane proteins in a periodic pattern consistent with actin ring spacing. Actin ring-mediated membrane compartmentalization is thus not unique to neurons. Finally, we show that acute actin disruption in live progenitor-derived neuronal cells leads to a loss of membrane compartmentalization in areas that were compartmentalized before treatment. Taken together, we here develop a much-needed tractable experimental system for the investigation of membrane compartmentalization and show that actin rings compartmentalize the plasma membrane of cells.