Bent out of shape

The bending of cellular membranes is central to the generation of most organelles and dynamic cellular structures such as endocytic pits. A favored model for protein-induced membrane bending is the insertion of amphipathic helices into one leaflet of the lipid bilayer to create membrane-deforming wedges. However, continuum models and molecular dynamics simulations predict that to induce sufficient bending and produce endocytic structures, epsin1—an amphipathic helix-containing protein involved in clathrin-mediated endocytosis—would have to cover >100% of the membrane. Stachowiak, Schmid, Fletcher, Hayden and colleagues set out to test the hypothesis that epsin1 bends membranes by helix insertion, but instead demonstrate that protein crowding can drive bending in the absence of helix insertion. The authors removed the epsin1 amphipathic helix and added a His₆ tag to the ENTH domain of epsin1. They then used vesicles containing modified lipids that can bind histidine tags. At similar surface density, membrane curvature is induced by epsin1's ENTH domain, whether or not a helix is present. The same observations were made with a similarly membrane-targeted GFP, suggesting that protein density per se is capable of inducing bending. It is not yet clear how or even if the ability of protein crowding to bend membranes is at play in vivo. However, the results suggest the interesting possibility that the directionality of membrane curvature could be influenced by the relative density of protein on either side of the membrane. (Nat. Cell Biol. 14, 944–949, 2012)