NCAM2

NCAM2 is a brain-enriched immunoglobulin superfamily cell adhesion molecule that mediates homophilic trans-interactions, organizes dendritic and axonal cytoskeletal architecture, regulates synapse maturation, and controls the intracellular trafficking and amyloidogenic processing of APP and BACE1. Its Ig1 domain undergoes beta-strand swapping to form homophilic dimers (PMID:18706912), while its rigid FnIII domains bind FGFR and activate Ras-MAPK signaling to promote neurite outgrowth (PMID:29895898). NCAM2 complexes with MAP2 and 14-3-3 proteins to stabilize dendritic microtubules and support neuronal polarization and cortical migration (PMID:32043120), and triggers c-Src-dependent propagation of L-type Ca²⁺ channel-mediated submembrane Ca²⁺ spikes that regulate dendritic protrusion dynamics and synaptic vesicle maturation (PMID:29522129). NCAM2 binds APP extracellularly and Rab11-FIP5 intracellularly to direct APP into BACE1-containing Rab11-positive recycling endosomes for amyloidogenic processing, while BACE1-generated NCAM2 C-terminal fragments reciprocally promote BACE1 targeting to those endosomes; Aβ oligomers in turn bind NCAM2 and induce its cleavage and removal from synapses, driving disassembly of GluR1-containing glutamatergic synapses (PMID:40721030, PMID:36251052, PMID:26611261).

1. 1997 High

   Cloning of NCAM2 established it as a brain-enriched Ig-superfamily member structurally related to NCAM1, defining the domain architecture (five Ig, two FnIII, TM, cytoplasmic tail) that subsequent work would parse for discrete functions.

   Evidence cDNA cloning and sequence analysis from human fetal brain library

   PMID:9226371

   Open questions at the time

   • No functional data at this stage
   • Expression pattern limited to Northern blot survey
2. 2008 High

   Crystal structure of the Ig1 domain revealed that homophilic binding is mediated by reciprocal beta-strand swapping, providing the first molecular mechanism for NCAM2 trans-adhesion.

   Evidence X-ray crystallography at 2.7 Å resolution with gel-filtration confirmation of Ig1 dimerization

   PMID:18706912

   Open questions at the time

   • Full-length extracellular domain structure not determined
   • Contribution of other Ig domains to binding affinity unknown
   • No cell-based validation of the swapping interface
3. 2015 High

   Discovery that Aβ binds NCAM2 at the neuronal surface, induces its proteolytic cleavage and synaptic removal, and thereby triggers GluR1-positive glutamatergic synapse disassembly provided the first link between NCAM2 and Alzheimer's disease-related synaptotoxicity.

   Evidence Surface binding assays, shRNA knockdown, cleavage-resistant mutant rescue in cultured hippocampal neurons; confirmation in AD patient hippocampus

   PMID:26611261

   Open questions at the time

   • Identity of the protease cleaving NCAM2 in this context was not established
   • Whether NCAM2 cleavage is necessary or sufficient for synapse loss in vivo was not resolved
4. 2018 High

   Structural and functional dissection of the FnIII domains showed they form a rigid unit that binds FGFR and activates Ras-MAPK signaling to drive neurite outgrowth, assigning a signaling output to a specific extracellular region distinct from the homophilic Ig1 domain.

   Evidence NMR and SAXS structural analysis; in vitro FGFR binding; neurite outgrowth with FGFR and MAPK inhibitors

   PMID:29895898

   Open questions at the time

   • Direct FnIII–FGFR binding interface not mapped at residue level
   • Relevance to in vivo neurite outgrowth not tested
5. 2019 High

   NCAM2 overexpression was shown to induce L-type Ca²⁺ channel-dependent submembrane Ca²⁺ spikes in dendritic protrusions, propagated by c-Src, that increase protrusion instability and impair spine maturation, linking NCAM2 levels to calcium-dependent control of synapse development.

   Evidence Live Ca²⁺ imaging in mouse cortical neurons with pharmacological dissection (L-type Ca²⁺ channel blockers, c-Src inhibitors)

   PMID:29522129

   Open questions at the time

   • Whether endogenous NCAM2 loss-of-function has reciprocal effects on Ca²⁺ dynamics not shown
   • Direct physical interaction between NCAM2 and L-type channels or c-Src not demonstrated
6. 2020 High

   Identification of a NCAM2–MAP2–14-3-3γ/ζ complex and demonstration that NCAM2 knockdown destabilizes microtubules, shortens dendrites, and impairs neuronal polarization and cortical migration established a cytoskeletal organizing function for NCAM2.

   Evidence Co-immunoprecipitation with proteomic validation; shRNA knockdown in hippocampal neurons and in vivo cortical electroporation

   PMID:32043120

   Open questions at the time

   • Whether NCAM2 directly binds MAP2 or the interaction is bridged by 14-3-3 unknown
   • Phosphorylation-dependence of complex formation not functionally tested
7. 2021 Medium

   A broader interactome (neurofilaments, CaMKIIα, actin, Nogo) was identified for NCAM2 in cerebral cortex, expanding the scope of cytoskeletal and signaling partners beyond the MAP2/14-3-3 complex.

   Evidence Mass spectrometry immunoprecipitation with co-IP validation from mouse cortex

   PMID:34299022

   Open questions at the time

   • Single-lab MS interactome; independent validation in other systems needed
   • Functional significance of individual interactions not tested
   • In silico phosphorylation site predictions not experimentally confirmed
8. 2022 High

   BACE1 was identified as the protease that cleaves NCAM2, and the resulting C-terminal fragment was shown to associate with BACE1 and promote its targeting to Rab11-positive recycling endosomes, revealing a reciprocal trafficking relationship.

   Evidence Co-IP, confocal co-localization, gain- and loss-of-function in neurons and CHO cells; NCAM2-KO mice show increased surface BACE1 and altered Sez6 cleavage

   PMID:36251052

   Open questions at the time

   • Exact BACE1 cleavage site on NCAM2 not mapped
   • Whether NCAM2 CTF has signaling activity beyond BACE1 trafficking unknown
9. 2023 Medium

   In vivo loss of NCAM2 reduces axonal BACE1 in hippocampal mossy fibers, shortens the infrapyramidal bundle, and causes behavioral deficits in spatial memory and cognitive flexibility, demonstrating a physiological role for the NCAM2–BACE1 trafficking axis in circuit function.

   Evidence NCAM2-deficient mouse model with immunofluorescence quantification and behavioral testing

   PMID:37522285

   Open questions at the time

   • Single-lab KO model; replication in independent cohorts needed
   • Whether behavioral phenotypes are BACE1-dependent or reflect other NCAM2 functions not dissected
10. 2023 Medium

    NCAM2 overexpression in the adult dentate gyrus arrests radial glial progenitors in an undifferentiated state, revealing a role in regulating adult neurogenesis progression beyond its known synaptogenic and cytoskeletal functions.

    Evidence In vivo viral overexpression in adult mice with immunofluorescence characterization of neurogenic stages

    PMID:37724425

    Open questions at the time

    • Mechanism by which NCAM2 maintains RGP-like state not identified
    • Whether NCAM2 loss-of-function accelerates RGP differentiation not tested
11. 2025 High

    A complete trafficking model was established: NCAM2 binds APP extracellularly and Rab11-FIP5 intracellularly, internalizes APP from the surface into BACE1-containing Rab11+ recycling endosomes for amyloidogenic processing, with Aβ oligomers and synaptic NMDA receptor activation enhancing NCAM2–APP association, creating a potential pathogenic feed-forward loop.

    Evidence Domain-specific co-IP, confocal co-localization, gain- and loss-of-function in CHO cells and neurons, Aβ oligomer and NMDA receptor stimulation

    PMID:40721030

    Open questions at the time

    • Whether the NCAM2–APP–BACE1 axis operates identically in vivo in AD models not shown
    • Stoichiometry and structural basis of the NCAM2–APP and NCAM2–Rab11-FIP5 interactions not determined

• Key unresolved questions include: the structural basis of full-length NCAM2 extracellular domain interactions (only Ig1 and FnIII structures are known), the identity and regulation of proteases other than BACE1 that cleave NCAM2 in synaptic contexts, whether the NCAM2–APP–BACE1 trafficking axis is a viable therapeutic target in Alzheimer's disease in vivo, and how the cytoskeletal, signaling, and trafficking functions of NCAM2 are coordinated through its phosphorylation state.
• Full-length ectodomain structure unavailable
• Non-BACE1 proteases responsible for Aβ-induced cleavage unidentified
• In vivo AD model validation of the trafficking axis lacking
• Phosphorylation-dependent regulation of NCAM2 complexes unexplored