Protein Structure Effects on Membrane Bending by Protein-Protein Crowding

Abstract

Two major mechanisms of cellular membrane bending during processes such as clathrin-mediated endocytosis have been previously proposed: bending by curved protein scaffolds such as a clathrin coat, and bending by insertion of wedge-like amphipathic helices into the membrane by adaptor proteins such as epsin1. Recently we have reported a third general membrane bending mechanism; bending by protein-protein crowding, where pressure generated by densely bound proteins drives membrane bending(1). Several endocytic adaptor proteins consist of a folded N-terminal membrane binding domain, and an unfolded C-terminal domain that binds clathrin and other proteins. Due to their lack of structure, the unfolded protein domains have much larger hydrodynamic radii than folded protein domains, potentially increasing the effects of their crowding compared to proteins of equal molecular weight. We have investigated the capability of these unfolded portions of adaptor proteins to bend membranes by binding them to giant unilamellar vesicles. using a Foster resonance energy tranfer based assay of protein density, developed in our previous studies, we find that the unstructured epsin C-terminus can bend model membranes at substantially lower densities than the structured epsin N-terminal homology domain, which has traditionally been thought to drive bending. These findings suggest that concentrating unfolded domains of adaptor proteins at endocytic sites may have a previously unappreciated role in promoting membrane bending. We also find that the addition of clathrin can locally increase the concentration of epsin1 on the membrane surface. Our ongoing experiments are investigating how clathrin and adaptor proteins work together to curve membrane surfaces.(1) Stachowiak, J.C. et al., Nat. Cell Biol. 14, 944, (2012).