EPS8L2 (EPS8 signaling adaptor L2) is a cytoplasmic actin-binding adaptor protein of the EPS8 family, with an N-terminal PTB domain, a central SH3 domain and a C-terminal effector region that binds actin filaments directly. In mammalian fibroblasts EPS8L2 joins ABI1 and the Ras/Rac exchange factor SOS1 in a complex that activates SOS1 guanine nucleotide exchange toward Rac, and it is recruited to growth factor-induced, F-actin-rich membrane ruffles, where it supports Rac-dependent ruffle formation. In the inner ear, mouse Eps8L2 sits at the tips of hair-cell stereocilia and is needed to maintain the staircase architecture of the adult hair bundle; loss of EPS8L2 causes progressive hearing loss in mice and autosomal recessive deafness (DFNB106) in humans. The donkey protein (681 aa) is a near full-length ortholog of human EPS8L2 in the same PANTHER subfamily (PTHR12287:SF20), 84% identical over 679 aligned residues, with all three domains present.
Summary: EPS8L2 binds actin filaments directly through its conserved C-terminal effector region (F-actin cosedimentation of the purified EPS8L2 C-terminal fragment). The donkey protein retains this region.
Reason: Actin binding is correct and is the defining activity of the EPS8L2 subfamily, but the experimental evidence is more specific - the EPS8L2 C-terminus cosediments with polymerized actin (F-actin), which is actin filament binding (GO:0051015, an is_a child of GO:0003779; mouse Eps8l2 already carries it from the same paper). The donkey protein sits in the same PANTHER subfamily (PTHR12287:SF20) as human and mouse EPS8L2 and the effector region is conserved (84% identity end to end), so the TreeGrafter transfer is sound; only the term should be refined.
Consistent with the immunoprecitation results, eps8, eps8L1, and L2, but not eps8L3 could directly associate to F-actin, albeit with different stoichiometry, pointing to the existence of a novel, conserved F-actin–binding surface present in the C-terminus of eps8Ls.
Summary: EPS8L2 is a soluble cytoplasmic adaptor that is recruited to cortical F-actin structures; cytoplasmic localization is experimentally recorded for the human ortholog.
Reason: True and as precise as the general localization evidence allows (soluble protein, no signal peptide or transmembrane segment), but the functionally relevant locations are the F-actin-rich ruffle membrane and the stereocilium tip, which are captured separately. Note that the donkey SubCell line itself is ARBA-derived, so support comes from the human experimental record.
Summary: EPS8L2 is a peripheral actin-associated protein at specialized plasma membrane regions (growth factor-induced ruffles; stereocilia tips), not an integral or general plasma membrane protein.
Reason: The term is true but too general. The evidence places EPS8L2 at F-actin-rich membrane ruffles, and ruffle membrane (GO:0032587, which is part of the plasma membrane via plasma membrane region) is already annotated to this gene and accepted below. The broad plasma membrane term (an is_active_in IBA at the family node PTN000958212 for human and mouse) adds nothing beyond that.
Summary: EPS8L2 assembles with ABI1 and SOS1 into a complex that activates SOS1 Rac-GEF activity, a step in signaling from receptor tyrosine kinases to the Rho-family GTPase Rac.
Reason: Human EPS8L2 interacts with ABI1 and SOS1 and activates the Rac-GEF activity of SOS1 about as efficiently as EPS8 itself (in vitro GEF assays on immunoprecipitated complexes). EPS8L2 is therefore a component of the GEF-activating module in Rac signaling. PAINT placed this term at the EPS8-family node PTN000958212, and the donkey protein lies inside that clade (subfamily SF20).
Similarly to eps8, eps8Ls interact with Abi1 and Sos-1; however, only eps8L1 and eps8L2 activate the Rac-GEF activity of Sos-1, and bind to actin in vivo.
Summary: The only experimental basis for vesicle localization of any EPS8L2 is detection of human EPS8L2 in mass-spectrometry inventories of secreted exosome-like vesicles.
Reason: Human EPS8L2 is annotated to vesicle from a proteomic survey of exosome-like vesicles released by airway epithelium (and to extracellular exosome from similar HDA studies). Cytoskeletal proteins are routinely captured in such preparations, and no study places EPS8L2 function at a vesicle. The human and mouse IBA sets do not include a vesicle term, so this TreeGrafter row appears to propagate a proteomics-only location across the subfamily. Not contradicted, but over-annotated.
Summary: EPS8L2 is enriched in PDGF-induced, F-actin-rich membrane ruffles, where it acts in Rac-dependent actin remodeling.
Reason: Experimentally shown for EPS8L2 expressed in fibroblasts and recorded for human EPS8L2 by IDA; PAINT placed it at node PTN000264860. The localization depends on the conserved F-actin-binding C-terminus, which the donkey protein retains.
Eps8L1 and eps8L2, similarly to eps8, were found enriched in PDGF-induced, F-actin–rich ruffles
GO:0035023 regulation of Rho protein signal transduction
IEA GO_REF:0000118
ACCEPT
Summary: By activating the Rac-GEF activity of SOS1 within the ABI1-SOS1 complex, EPS8L2 regulates Rac (Rho-family) signaling.
Reason: This is the best-characterized signaling function of EPS8L2 (human IDA, family-node IBA at PTN000958212). EPS8L2 itself contributes the GEF-stimulating activity of the complex, so this is participation, not an indirect effect.
Thus, the members of the eps8 family display overlapping functional properties, which can be ranked by their ability to elicit Rac-GEF activity and to substitute for eps8 (eps8 = eps8L2 > eps8L1 ≫ eps8L3), and generate functional redundancy in pathway connecting RTK to Rac.
Summary: EPS8L2 restores growth factor-induced ruffle formation in Eps8-null fibroblasts, as efficiently as EPS8.
Reason: Rescue of PDGF-induced ruffling in Eps8-/- fibroblasts directly links EPS8L2 activity (Rac-GEF activation plus F-actin binding) to ruffle assembly; human IGI and family-node IBA (PTN000264860) agree.
Ectopic expression of eps8, eps8L1, or eps8L2, but not of eps8L3, restored ruffling formation, in response to growth factor stimulation, in eps8 null MEFs
ISS PMID:23918390 Progressive hearing loss and gradual deterioration of sensor...
NEW
Summary: Mouse Eps8L2 is a hair bundle protein localized at the tips of stereocilia of cochlear and vestibular hair cells.
Reason: Mouse Eps8l2 (Q99K30) has a direct experimental annotation to stereocilium tip from this paper, and human EPS8L2 carries it by ISS. Mouse, human and donkey EPS8L2 all belong to PANTHER subfamily PTHR12287:SF20, so this is an ortholog-to-ortholog transfer within one subfamily, not a family-level inference; the donkey protein retains the F-actin-binding C-terminus that underlies tip targeting of EPS8-family proteins. The PAINT family nodes (PTN000958212, PTN000264860) span EPS8 and all EPS8L paralogs and carry no stereocilium term, which is consistent with this being a subfamily-level location rather than a deliberate exclusion. Only the location is proposed; the hearing and stereocilium-maintenance phenotypes of the knockout are not added as process terms.
Localizes at the tips of the stereocilia of the inner and outer hair cells.
Core Functions
Actin filament-binding adaptor that, together with ABI1 and SOS1, activates SOS1 Rac-GEF activity and so promotes Rac-dependent membrane ruffling and actin remodeling at F-actin-rich ruffles.
Consistently, eps8L1 and eps8L2, but not eps8L3, localize to PDGF-induced, F-actin–rich ruffles and restore receptor tyrosine kinase (RTK)-mediated actin remodeling when expressed in eps8 –/– fibroblasts.
Actin filament-binding protein at the tips of hair-cell stereocilia, where it is needed to keep the staircase architecture of the mature hair bundle (shown for the mouse ortholog).
Q: Does donkey EPS8L2 cap or bundle actin filaments at stereocilia tips as mouse Eps8l2 and Eps8 do, and is it required for hearing in equids?
Q: Human EPS8L2 carries contributes_to guanyl-nucleotide exchange factor activity (GO:0005085) from PMID:14565974, but PAINT did not propagate that term to the family nodes; is the SOS1-stimulating activity better captured as a GEF-complex contribution or as a molecular adaptor activity for this family?
External Prediction Reviews
These computational predictions are reviewed separately from the GOA annotation set used for this review. The assessments below are from this project and do not constitute official GO annotations or endorsement by GO/UniProt. They are not included in the existing annotation review above.
Assessment of the single ProtNLM2 GO prediction for donkey EPS8L2 (A0A9L0KBR4), an EPS8-family actin-binding signaling adaptor that is 84% identical to human EPS8L2. The predicted process phosphorylation (GO:0016310) is judged incorrect (NPI); its ProtNLM2 evidence key is the UniProt Phosphoprotein keyword (KW-0597), and being a phosphorylated protein does not make EPS8L2 a participant that carries out phosphorylation. ProtNLM2 also emitted a subcellular location of Cytoplasm (evidence key KW-0963), which agrees with the experimentally supported cytoplasmic localization of human EPS8L2 and with the donkey GOA cytoplasm row; no function text was emitted. Non-GO outputs are noted, not scored.
Prediction method: ProtNLM2 · Version: UniProt API snapshot 2026-10-14
Review rationale: GO:0016310 phosphorylation is the process of introducing a phosphate group, so it applies to gene products that catalyze or directly carry out phosphate transfer. EPS8L2 has no kinase domain; its domains are PTB, SH3 and the EPS8 actin-binding effector region, and its characterized activities are F-actin association and stimulation of SOS1 exchange activity toward Rac. The ProtNLM2 evidence for this prediction is the Phosphoprotein keyword (KW-0597), which describes EPS8L2 as a substrate of phosphorylation rather than as an agent of it. The term is absent from donkey GOA, which instead carries actin binding, Rho protein signal transduction and ruffle-assembly terms, and the merged gene review describes EPS8L2 as an actin filament-binding adaptor that activates SOS1 Rac-GEF activity. The prediction is therefore refuted for EPS8L2.
Supporting Evidence:
PMID:14565974: "only eps8L1 and eps8L2 activate the Rac-GEF activity of Sos-1, and bind to actin in vivo"