ESCRT-III assembly and budding – ESCRTbudding

Eukaryotic cells maintain and adjust their cell surface protein composition in part by removing transmembrane proteins from the plasma membrane by endocytosis and delivering them to the lysosome for degradation. The proteins are sorted into vesicles, which bud off into the lumen of the endosome, giving rise to the multivesicular-body (MVB) compartment. The endosomal sorting system is conserved from yeast to mammals and comprises a network of proteins that includes ESCRT (Endosomal Sorting Complexes Required for Transport) complexes 0, I, II and III and associated proteins. Part or all of the ESCRT machinery have been additionally implicated in other biological processes ranging from enveloped virus budding, cytokinesis, autophagy to neuron specific abnormalities. MVB vesicle sorting, enveloped virus budding and cytokinesis constitute topologically similar processes since they can be seen as budding processes away from the cytosol. All three processes require at least ESCRT-III complexes and the function of VPS4, a AAA-type ATPase, since dominant negative forms of either ESCRT-III or VPS4 inhibit MVB formation, enveloped virus budding and cytokinesis. Thus a minimal common ESCRT-III function for all three processes could be membrane abscission to separate two membrane enveloped structures after completion of budding. ESCRT-III is composed of six members (CHMP1-6) that assemble into a complex on membranes to catalyze late steps in budding and is disassembled by VPS4. Furthermore, mutations in CHMP2B have been associated with neuronal pathologies. The crystal structure of ESCRT-III CHMP3 revealed membrane targeting as well as potential dimerization interfaces to from a polymer structure. The CHMP3 structure represents the activated form of CHMP3, consistent with regulated association of cytsolic CHMP proteins with cellular membranes. The neuron specific effect of the CHMP2B mutation may in part depend on the constitutive activation of CHMP2B (Sadoul lab). Since there are six CHMP proteins, we (Weissenhorn in collaboration with Schoehn lab) set out to understand the assembly propensities of CHMP proteins in vitro. Our structural data show that CHMP2A and CHMP3 alone or together with CHMP4B assemble into large helical tubular structures that are disassembled by VPS4. The polymeric structures contain the VPS4 binding site on the inside and expose the membrane targeting surface on their outside, suggesting that the polymers assemble on the inside of a newly formed bud. We show further that CHMP2A and CHMP3 assemble on negatively charged phospholipids into a salt resistant complex and that the presence of lipid bilayers affects the assembly propensities of CHMP2A, CHMP3 and CHMP4B. Open questions are still (i) how the other three CHMP members participate in ESCRT-III assembly, (ii) how such assemblies affect a membrane structure and (iii) whether ESCRT-III and VPS4 constitute the minimal membrane fission machinery. Our proposal will address these issues by solving the structures of ESCRT substructures by X-ray crystallography and combine it with low-resolution structural models of ESCRT-III assemblies obtained by electron microscopy. We will further develop in vitro assays to test membrane deformation or membrane fission activity of ESCRT complexes. ESCRT-III members interact with specific binding partners that regulate and position budding complexes on cellular membranes. We have shown that CHMP4B interacts with CC2D1A/B proteins, both homologues of the Drosophila protein LgD, which is implicated in endosomal trafficking. Furthermore mutant CC2D1 proteins are associated with neuronal pathologies, similar to ESCRT-III CHMP2B. We have thus initiated structural studies by X-ray crystallography of CC2D1 proteins in complex with CHMP4B to understand its role in normal and pathological processes. Our in vivo studies will focus on ESCRT-III function in neurons, since little is known about neuronal abnormalities induced by either CHMP2B or CC2D1. Thus our combined effort will give important insight into a conserved protein machine that regulates topologically similar processes as well as cell type specific activities, that may distinguish neuronal ESCRT-III from ESCRT-III acting at the endosome, at the midbody during cytokinesis or at sites of enveloped virus egress.