Final Report Summary - MEMPART (Membrane partitioning of homologous proteins)
The clustering and segregation of lipids and proteins in distinct domains is important for the spatial patterning of biological membranes. Goal of MEMPART was to elucidate how members of the SNARE protein family and other structurally and functionally homologous membrane proteins form such distinct domains in biological membranes of diverse compositions. To reach this goal, we characterized domains in the plasma membrane of neuroendocrine PC12 cells and in the phagosomal membrane of dendritic cells enriched in homologous SNARE proteins by a combination of super-resolution STED microscopy, TIRF microscopy and biochemical approaches. Understanding how these SNARE proteins segregate into distinct membrane clusters is essential for understanding their distinct roles in regulated and constitutive exocytosis and in endo/phagosomal trafficking. By combining molecular imaging experiments with perturbation experiments, both in cells and in reconstituted artificial membranes, we were able to show that hydrophobic mismatch between the length of transmembrane domains and the thickness of the lipid membranes induces the clustering of SNARE proteins. Even although the transmembrane helices of the SNAREs syntaxin-1 and syntaxin-4 differ in length by only a single residue, hydrophobic mismatch can segregate these structurally closely homologous membrane proteins in distinct membrane domains. Studies of SNAREs in immune cells showed that the segregation of membrane domains by hydrophobic mismatch is a general principle. In neuroendocrine PC12 cells, we showed that the domain formation by hydrophobic mismatch is fine-tuned by interactions with the polyphosphoinositide phosphatidylinositol 4,5-bisphosphate (PI(4,5)P2). Although PI(4,5)P2 is a minor component of total plasma membrane lipids, it has an essential role in the regulation of many cellular functions, including SNARE mediated exocytosis. Our results show that PI(4,5)P2 colocalizes in domains containing syntaxin-1 as well as syntaxin-4, and that these domains form recognition sites for vesicle docking. This colocalisation of PI(4,5)P2 with the syntaxin proteins is driven by electrostatic interactions between the polyanionic lipid with conserved juxtamembrane polybasic domains of both syntaxin-1 and syntaxin-4. We also demonstrated that calcium promotes syntaxin-1 clustering in the plasma membrane by acting as a charge bridge that specifically and reversibly connects multiple syntaxin/PI(4,5)P2 complexes into larger mesoscale domains. This transient reorganization of the plasma membrane by physiological calcium concentrations is likely to be important for SNARE function. Overall, MEMPART revealed how structurally homologous membrane proteins are clustered and segregated into distinct membrane domains by a combination of hydrophobic mistmatch and electrostatic interactions with phosphoinositides and bivalent cations, and this is important for our understanding how SNAREs and cargo molecules are sorted into trafficking vesicles and for the intracellular organization of organellar trafficking.