CHMP6

CHMP6 (also known as human Vps20) is an N-myristoylated ESCRT-III subunit that bridges ESCRT-II to ESCRT-III on endosomal membranes, nucleating Snf7/CHMP4b filament assembly to drive intralumenal vesicle formation, cargo sorting, and cytokinetic abscission. CHMP6 is recruited to membranes via N-terminal myristoylation and directly engages ESCRT-II through both the EAP20/Vps25 subunit and the Vps28 C-terminal domain, with ESCRT-II–Vps20 binding enhanced on curved membranes to provide spatial regulation of ESCRT-III polymerization (PMID:15511219, PMID:16749904, PMID:21835927). At the intercellular bridge during cytokinesis, CHMP6 and ESCRT-II form ordered assemblies whose disruption by dominant-negative CHMP6 fragments blocks abscission, and on endosomes CHMP6 additionally binds H-Ras and N-Ras to facilitate their recycling to the plasma membrane (PMID:25232011, PMID:22231449). The Bro1/ALIX–Vps20 interaction modulates Doa4 deubiquitinase access to Snf7, establishing a regulatory checkpoint that coordinates ubiquitin recycling with ESCRT-III disassembly (PMID:23444383, PMID:34908216).

1. 2004 Medium

   Establishing that human Vps20/CHMP6 localizes to endosomes and functionally cooperates with Snf7-1 answered the basic question of where this ESCRT-III subunit acts and provided the first evidence it participates in post-endosomal sorting.

   Evidence Immunofluorescence and co-expression experiments in mammalian cells showing endosomal staining and cholesterol sorting defects

   PMID:14583093

   Open questions at the time

   • No direct binding partners identified
   • Mechanism of cholesterol sorting defect unexplored
2. 2005 High

   Identifying CHMP6 as N-myristoylated and showing it directly bridges ESCRT-II (via EAP20) to ESCRT-III (via CHMP4b) through its N-terminal basic domain defined CHMP6 as the critical adaptor linking ESCRT-II to ESCRT-III on endosomal membranes.

   Evidence Metabolic [3H]myristate labelling, co-immunoprecipitation, in vitro pull-down with recombinant proteins, and cargo sorting assays in mammalian cells

   PMID:15511219

   Open questions at the time

   • Structural basis of the EAP20–CHMP6 interface not resolved
   • Contribution of myristoylation to membrane targeting versus protein–protein interaction not separated
3. 2006 High

   Crystallography of the Vps28 C-terminal domain and mutagenesis revealed a second, ESCRT-I–mediated route for Vps20/CHMP6 recruitment, and functional assays showed this interface is required for ESCRT-dependent viral budding.

   Evidence Crystal structure of Vps28-CTD, mutagenesis abolishing Vps20 binding in vitro, and EIAV Gag late-domain budding rescue assay

   PMID:16749904

   Open questions at the time

   • Relative contributions of ESCRT-I versus ESCRT-II recruitment of CHMP6 in physiological MVB biogenesis not determined
4. 2011 High

   Reconstitution on liposomes demonstrated that ESCRT-II–Vps20 binding is membrane-curvature dependent and that this complex nucleates single-filament Snf7 polymers, establishing the mechanistic basis for spatially restricted ESCRT-III assembly.

   Evidence Liposome co-flotation, fluorescence-based binding, and high-resolution atomic force microscopy with purified yeast proteins

   PMID:21835927

   Open questions at the time

   • Filament architecture at endogenous protein concentrations unknown
   • Whether mammalian CHMP6–CHMP4b polymerization follows identical curvature dependence not tested
5. 2012 Medium

   Discovery that CHMP6 selectively binds GTP-loaded, ubiquitylated H-Ras and N-Ras on endosomes and that CHMP6 knockdown blocks Ras and EGFR recycling to the plasma membrane expanded CHMP6 function beyond ILV formation to signaling-relevant receptor and GTPase recycling.

   Evidence cDNA library screen, direct binding assay, FRAP, cell fractionation, RNAi with transformation assay

   PMID:22231449

   Open questions at the time

   • Structural basis of CHMP6–Ras interaction not defined
   • Whether CHMP6–Ras binding is direct on native endosomes or requires additional adaptors not resolved
   • Single-lab finding
6. 2013 High

   Identification of a MIM1-like motif in Vps20 that binds the Doa4 MIT domain, restricting Doa4's non-catalytic ILV-promoting function, revealed an unexpected regulatory role for Vps20 beyond filament nucleation.

   Evidence Recombinant protein binding assay and yeast genetic epistasis (bro1Δ rescue) with ILV biogenesis readout

   PMID:23444383

   Open questions at the time

   • Whether mammalian CHMP6 similarly regulates a deubiquitinase untested
   • Molecular details of Doa4's non-catalytic ILV function unclear
7. 2014 High

   Demonstrating that CHMP6 and ESCRT-II form ordered structures at the intercellular bridge and that a dominant-negative CHMP6 N-terminal fragment blocks abscission in an ESCRT-II-binding-dependent manner established a direct role for CHMP6 in cytokinesis.

   Evidence High-resolution imaging, truncation and point mutagenesis of CHMP6-N (first 52 aa), live-cell abscission assay

   PMID:25232011

   Open questions at the time

   • Mechanism by which the first 10 residues of CHMP6 contribute to abscission function beyond bridge localization not elucidated
8. 2015 Medium

   Biophysical analysis of C. elegans VPS-20 showed it adopts an open, extended conformation in solution — unlike VPS-24 — indicating CHMP6/Vps20 is constitutively competent for ESCRT-II engagement without requiring membrane-dependent activation.

   Evidence Recombinant protein binding and biophysical conformational analysis of C. elegans VPS-20

   PMID:25588614

   Open questions at the time

   • Whether mammalian CHMP6 shares this open conformation not directly tested
   • Single-lab study on ortholog
9. 2022 Medium

   Showing that Bro1 binds Vps20 and relieves Vps20-mediated suppression of Doa4–Snf7 interaction completed a regulatory circuit in which Bro1 coordinates deubiquitinase engagement with ESCRT-III disassembly.

   Evidence Recombinant protein binding assay, yeast genetics with constitutively Doa4-binding Vps20 mutant

   PMID:34908216

   Open questions at the time

   • Kinetics of Bro1-mediated relief in vivo not measured
   • Mammalian ALIX–CHMP6 regulatory axis not validated

• Key unresolved questions include the high-resolution structure of the CHMP6–ESCRT-II interface in human, whether CHMP6's Ras-binding function is mechanistically coupled to ESCRT-III filament dynamics, and the physiological relevance of CHMP6 in extracellular vesicle biogenesis beyond overexpression systems.
• No atomic-resolution structure of human CHMP6 in complex with ESCRT-II
• CHMP6–Ras interaction awaits structural and independent validation
• Role in EV biogenesis demonstrated only by overexpression of myristoylation peptide