SH3GL3

SH3GL3 is a BAR- and SH3-domain-containing endophilin-family protein that participates in membrane trafficking at nerve terminals and functions as a tumor suppressor in multiple cancer contexts. Its SH3 domain directly binds the proline-rich regions of dynamin I and synaptojanin, concentrating these endocytic effectors at synaptic terminals to facilitate synaptic vesicle recycling (PMID:9238017); the same SH3 domain mediates interaction with polyglutamine-expanded huntingtin, promoting aggregate formation relevant to Huntington disease pathology (PMID:9809064). In cancer cells, SH3GL3 suppresses proliferation, migration, stemness, and cell-cycle progression through SH3-domain-dependent mechanisms that include physical interaction with STAT3 to restrain its nuclear translocation (PMID:33524871, PMID:33168185) and modulation of FAK/PI3K signaling (PMID:27683032), while its expression is epigenetically silenced by DNMT1-mediated promoter methylation recruited by the lncRNA MIR210HG (PMID:37916731).

1. 1997 High

   Establishing SH3GL3 as a synaptic vesicle recycling component resolved the question of whether all three endophilin-family SH3 domains engage dynamin I and synaptojanin at nerve terminals.

   Evidence Yeast two-hybrid, co-immunoprecipitation from rat brain extracts, and immunofluorescence showing nerve-terminal enrichment

   PMID:9238017

   Open questions at the time

   • No direct demonstration that SH3GL3 loss impairs synaptic vesicle endocytosis in neurons
   • Relative contribution of SH3GL3 versus SH3GL1/SH3GL2 at the synapse is undefined
2. 1998 High

   Demonstrating that SH3GL3 selectively binds pathological-length polyglutamine huntingtin and promotes its aggregation identified a novel molecular partner in Huntington disease aggregate formation.

   Evidence Co-immunoprecipitation from HD human brain, domain-deletion mutagenesis, and filter retardation aggregation assay in COS cells

   PMID:9809064

   Open questions at the time

   • Whether SH3GL3–huntingtin interaction contributes to neurodegeneration in vivo is untested
   • Stoichiometry and structural basis of SH3-domain/polyproline interaction with huntingtin are unresolved
3. 2009 Medium

   Identifying DYDC1 as a BAR-domain interactor and linking SH3GL3 to clathrin-coated vesicle traffic during acrosome biogenesis extended its membrane-trafficking role beyond neurons.

   Evidence Yeast two-hybrid with BAR domain bait, RNAi knockdown in germ cells, and germ cell transplantation

   PMID:19545932

   Open questions at the time

   • Interaction awaits validation by reciprocal co-immunoprecipitation from endogenous testis extracts
   • Whether BAR-domain–mediated membrane tubulation is required for acrosome formation is unknown
4. 2012 Medium

   Showing that SH3GL3 knockdown inhibits glioma cell invasion provided the first evidence that SH3GL3 has tumor-suppressive activity, opening its characterization in cancer biology.

   Evidence shRNA knockdown with Matrigel transwell invasion assay in glioma cells

   PMID:21722156

   Open questions at the time

   • Mechanism by which SH3GL3 restrains invasion was not identified
   • No in vivo tumor model used
5. 2016 Medium

   Linking SH3GL3 overexpression to FAK/PI3K activation in myeloma identified a context where SH3GL3 paradoxically promotes migration and stemness, contrasting with its tumor-suppressive role in glioma.

   Evidence Gain- and loss-of-function in myeloma cells with FAK/PI3K pharmacological inhibitor rescue and Western blot

   PMID:27683032

   Open questions at the time

   • Direct physical interaction between SH3GL3 and FAK or PI3K was not demonstrated
   • Context-dependent pro- versus anti-tumorigenic roles remain mechanistically unexplained
6. 2020 Medium

   Demonstrating SH3-domain dependence for SH3GL3's anti-proliferative, pro-apoptotic, and p21-inducing activities in lung cancer established that its tumor-suppressive function requires the same domain used in endocytic partner binding.

   Evidence Overexpression of full-length versus SH3-deletion mutants in lung cancer cells with proliferation, flow cytometry, and qRT-PCR readouts

   PMID:33168185

   Open questions at the time

   • The SH3-domain binding partner responsible for p21 upregulation is unidentified
   • Transcriptional mechanism linking SH3GL3 to p21 promoter activation is unknown
7. 2021 Medium

   Identifying STAT3 as a direct SH3GL3-interacting protein whose nuclear translocation is blocked by SH3GL3 provided a mechanistic explanation for its tumor-suppressive effects in glioblastoma.

   Evidence Co-immunoprecipitation, immunofluorescence for STAT3 localization, and epistasis rescue with constitutively active STAT3-C in glioblastoma cells

   PMID:33524871

   Open questions at the time

   • Domain on SH3GL3 that mediates STAT3 binding is not mapped
   • Whether STAT3 sequestration explains the lung cancer phenotype or p21 induction is untested
8. 2023 Medium

   Revealing that the lncRNA MIR210HG recruits DNMT1 to methylate the SH3GL3 promoter explained how SH3GL3 expression is epigenetically silenced in lung cancer.

   Evidence ChIP for DNMT1 at SH3GL3 promoter, RNA immunoprecipitation for MIR210HG–DNMT1, methylation-specific PCR, and functional rescue in lung cancer cells

   PMID:37916731

   Open questions at the time

   • Whether MIR210HG-mediated silencing operates in other cancer types is unknown
   • No in vivo demonstration that restoring SH3GL3 expression reverses tumor growth

• The structural basis for SH3GL3's context-dependent pro- versus anti-tumorigenic activities, the identity of the SH3-domain partner responsible for p21 transcriptional induction, and the in vivo physiological consequence of SH3GL3 loss in neurons remain unresolved.
• No knockout or conditional deletion mouse model phenotype reported
• No structural determination of SH3GL3 BAR or SH3 domains
• Relationship between endocytic function and tumor-suppressive role is mechanistically unconnected