VPS37A

VPS37A (HCRP1) is a subunit of the mammalian ESCRT-I complex that functions in membrane-remodeling reactions underlying both endosomal cargo sorting and autophagosome biogenesis (PMID:15240819, PMID:31519728). Within ESCRT-I it interacts with Tsg101 through its mod(r) domain and with hVps28, cofractionating and colocalizing with these partners on LAMP1-positive endosomes; in this context it is required for lysosomal sorting and degradation of ubiquitinated membrane receptors such as the EGF receptor (PMID:15240819). A genetically and structurally separable N-terminal ubiquitin-E2-variant-like (UEVL) domain mediates a distinct function in autophagy: it recruits VPS37A to the phagophore and is required to bring the ESCRT-I subunit VPS28 and the ESCRT-III subunit CHMP2A to the closing autophagosome, while being dispensable for ESCRT-I assembly and EGFR degradation (PMID:31519728). This autophagosome-closure activity is physiologically essential in vivo: mice carrying a UEVL-domain deletion show impaired autophagic flux, p62/SQSTM1 and ubiquitin accumulation, aberrant accumulation of active TBK1 on phagophores, and neuronal dysfunction (PMID:39607828). By controlling receptor trafficking, VPS37A also governs glucagon receptor localization in hepatocytes, restraining endosomal cAMP/PKA/p-CREB signaling and gluconeogenesis (PMID:36243006). A homozygous VPS37A missense mutation causes autosomal recessive complex hereditary spastic paraparesis, with zebrafish knockdown reproducing the locomotor defect (PMID:22717650).

1. 2004 High

   Established that VPS37A is a bona fide subunit of human ESCRT-I and connects the complex to endosomal receptor degradation, defining its core molecular identity.

   Evidence Reciprocal Co-IP, size exclusion chromatography, colocalization, and siRNA knockdown with EGFR degradation assay

   PMID:15240819

   Open questions at the time

   • Did not resolve the structural basis of complex assembly
   • Did not address roles outside the MVB/endosomal pathway
2. 2004 Medium

   Mapped the Tsg101-binding determinant to the mod(r) domain, defining the interface that anchors VPS37A within ESCRT-I.

   Evidence Co-IP with domain-mapping constructs

   PMID:15240819

   Open questions at the time

   • No structural model of the mod(r)–Tsg101 interface
   • Functional consequence of disrupting only this interface not isolated
3. 2012 Medium

   Linked VPS37A loss-of-function to human upper motor neuron disease, establishing physiological importance beyond cell culture.

   Evidence Whole-genome linkage, candidate gene sequencing, and zebrafish morpholino knockdown with locomotion phenotype

   PMID:22717650

   Open questions at the time

   • Mechanism connecting VPS37A loss to neuronal degeneration not defined at the time
   • Effect of the K382N mutation on specific VPS37A functions untested
4. 2019 High

   Discovered a second, separable function: VPS37A is required for phagophore closure and uses its N-terminal UEVL domain to recruit downstream ESCRT machinery to the autophagosome, distinguishing autophagy from endosomal roles.

   Evidence Genome-wide CRISPR screen, live-cell imaging, domain-deletion complementation, and epistasis with CHMP2A/VPS4

   PMID:31519728

   Open questions at the time

   • Molecular signal recruiting VPS37A to the phagophore not identified
   • How the UEVL domain engages VPS28/CHMP2A structurally is unknown
5. 2022 High

   Showed that VPS37A-controlled receptor trafficking has organ-level metabolic consequences by restraining endosomal glucagon receptor signaling.

   Evidence Hepatocyte-specific in vivo knockdown, agonist trafficking assay, cAMP/PKA/p-CREB readouts, and plasma-membrane rescue

   PMID:36243006

   Open questions at the time

   • Whether Gcgr regulation requires ESCRT-I assembly or autophagy not dissected
   • Direct interaction between VPS37A and Gcgr not demonstrated
6. 2024 High

   Demonstrated in vivo that the UEVL domain is specifically required for autophagosome closure and that its loss drives TBK1/p62 pathology and neuronal dysfunction, confirming the domain-function separation.

   Evidence UEVL-deletion knock-in mouse, autophagic flux assays, LC3 proximity proteomics, and histopathology

   PMID:39607828

   Open questions at the time

   • Mechanism by which TBK1 accumulates on UEVL-mutant phagophores unresolved
   • Cell-type basis of the neuronal phenotype not defined
7. 2025 Medium

   Indicated that VPS37A overexpression can redirect TNFR1 to lysosomal degradation and suppress NF-κB signaling under metabolic stress in colorectal cancer cells.

   Evidence Overexpression constructs, NF-κB reporter assay, lysosomal inhibition rescue, RNA-seq, and xenografts

   PMID:40746890

   Open questions at the time

   • Direct VPS37A–TNFR1 interaction not shown
   • Relative contribution of ESCRT vs autophagy routes unresolved

• How VPS37A is selectively targeted to the phagophore versus the endosome, and how its two domains are coordinated to partition between sorting and autophagy, remains unresolved.
• No identified recruitment signal directing VPS37A to phagophore membranes
• No structural model of the UEVL domain engaging VPS28/CHMP2A
• Cancer-context roles (EMT, TNFR1, MLKL-dependent autophagy) rest on low-confidence or retracted data