Featured Article - August 2012
Short description: An integration of several structural approaches has generated a model for how vesicle budding may be coordinated by ESCRT complexes.
Multivesicular bodies (MVBs) are crucial for trafficking ubiquitinated proteins to the lysosome. MVBs are created by the ESCRT complexes. ESCRT-0 sequesters ubiquitinated cargo proteins and brings them to the downstream ESCRTs. ESCRT-I and -II drive membrane bud formation by stabilizing the bud neck as it invaginates into the lumen of the endosome. ESCRT-III stimulates membrane scission. The ESCRT-I-II supercomplex connects ESCRT-0 cargo with ESCRT-III.
ESCRT-I and -II are themselves multisubunit assemblies. While the structures of the core assembly components are known, these ESCRTs are also comprised of several flexible, disordered regions. The size and flexibility of the supercomplex has made it difficult to study by traditional solution NMR or X-ray crystallographic methods. To tackle this, Hummer, Hurley and colleagues have instead used a combination of solution methods, including small-angle X-ray scattering (SAXS), double electron-electron resonance (DEER), and single-molecule Förster resonance energy transfer (smFRET), to obtain an improved model of the full-length yeast ESCRT-II complex and of the yeast ESCRT-I-II supercomplex.
The SAXS and smFRET data indicate that on its own, ESCRT-II exists in a rather compact globular state with the exception of the membrane-targeting GLUE domain, which adopts several different conformations with respect to the ESCRT-II core. ESCRT-II does not significantly change conformation in the presence of ESCRT-I, but the former has effects on the conformational dynamics of the latter. Specifically, DEER data collected from spin-labeled ESCRT-I indicate that when alone, the complex exists in a mixture of closed and open states that is biased to the open state when ESCRT-II is present. Molecular dynamics simulations, guided by the data collected from the solution studies, provide a conformational ensemble of the supercomplex. In the ensemble, a majority of the supercomplexes form an extended crescent-shaped assemblage; a small minority have the complexes folded back on one another.
What does this tell us about the supercomplex's role in MVB formation? The authors suggest that its crescent-like shape is aptly suited for interactions with the vesicle bud neck, and is consistent with a “spoke” model, in which multiple copies of ESCRT-II complexes are positioned at the bud pore, with ESCRT-III binding regions facing the center of the pore. ESCRT-I and ESCRT-0 would project outside of the pore. The observed conformational flexibility of the supercomplex suggests a mechanism for transfer of cargo from ESCRT-0 to the neck of the nascent bud. In all, the model is supported by a wealth of previous biological data and provides a structural framework that can be further tested.
E. Boura et al. Solution Structure of the ESCRT-I and -II Supercomplex: Implications for Membrane Budding and Scission.
Structure 20, 874-886 (2012). doi:10.1016/j.str.2012.03.008