CNTN3

CNTN3 (BIG-1/PANG) is a GPI-anchored immunoglobulin superfamily cell adhesion molecule built from six Ig-like domains and four fibronectin type III repeats that functions as an axon-associated adhesion molecule promoting neurite outgrowth (PMID:8060619). Its expression is restricted to specific neuronal subpopulations, including dentate gyrus granule cells, where it overlaps with but is distinct from related TAG-1/F3-subgroup members, consistent with roles in the formation of particular neuronal networks (PMID:8108413, PMID:8586965). Mechanistically, the second fibronectin type III domain of CNTN3 directly binds the copper-binding domain of APP, with alanine-scanning mutagenesis defining the contact residues on one face of each protein (PMID:31318883). Beyond its neuronal adhesion role, CNTN3 is a transcriptional target of PRRX2 in colorectal cancer, where its elevated expression drives proliferation, suppresses apoptosis, and promotes epithelial-to-mesenchymal transition, migration, and invasion (PMID:41882765).

1. 1994 Medium

   Establishing CNTN3 as a molecular entity answered whether the brain expresses a distinct GPI-anchored Ig-superfamily adhesion molecule capable of supporting axon growth.

   Evidence PCR-based cloning and recombinant protein neurite outgrowth assay (BIG-1); independent cDNA cloning and domain analysis (PANG)

   PMID:8060619 PMID:8108413

   Open questions at the time

   • No identified binding partner or receptor for the neurite-promoting activity at the time
   • GPI anchoring inferred from domain structure, not directly demonstrated functionally
   • Mechanism of outgrowth promotion (homophilic vs heterophilic) unresolved
2. 1995 Medium

   Mapping CNTN3 mRNA to defined neuronal subtypes addressed where this adhesion molecule could act, distinguishing it from related contactins.

   Evidence In situ hybridization of BIG-1, BIG-2, TAG-1, and F3 probes in developing and adult rat brain

   PMID:8586965

   Open questions at the time

   • Expression pattern does not establish functional necessity in those neurons
   • No loss-of-function phenotype assessed
3. 1996 Medium

   Chromosomal mapping placed CNTN3 in defined mouse and human loci, providing a genetic anchor for the gene.

   Evidence Somatic cell hybrid analysis and Southern blot mapping to mouse chr 6 and human 3p26

   PMID:8661054

   Open questions at the time

   • No disease association established at the mapped locus
   • Purely positional, no functional consequence
4. 2019 High

   Identifying APP as a direct binding partner answered which molecular interaction underlies CNTN3 function, defining a precise protein-protein interface.

   Evidence APP-AP and CNTN-Fc fusion protein binding assays with systematic alanine substitution mutagenesis

   PMID:31318883

   Open questions at the time

   • Functional consequence of the CNTN3-APP interaction in vivo not established
   • No structural model of the complex
   • Whether the interaction modulates neurite outgrowth or APP processing untested
5. 2026 Medium

   Linking CNTN3 to PRRX2-driven transcription in colorectal cancer answered whether this neuronal adhesion molecule has a pathological role outside the nervous system.

   Evidence qPCR, Western blot, IHC, ChIP, dual-luciferase reporter, knockdown/rescue functional assays, and xenograft tumor model in CRC

   PMID:41882765

   Open questions at the time

   • Downstream effectors of CNTN3 in EMT not identified
   • Whether the APP interaction contributes to the cancer phenotype unknown
   • Single-lab finding without independent confirmation

• The biochemical mechanism linking CNTN3 adhesion/APP binding to its neuronal network role and its oncogenic EMT function remains unresolved.
• No signaling pathway downstream of CNTN3 defined
• No in vivo neuronal loss-of-function phenotype
• Functional integration of APP binding with cell adhesion and cancer roles unknown