MDGA2

MDGA2 is a cell-surface IgCAM-superfamily protein that functions as a selective suppressor of excitatory synapse development and function in the mammalian brain (PMID:27608760, PMID:38321347). Its canonical mechanism is to block neuroligin-1 (NLGN1) engagement of neurexins, restraining excitatory synaptogenesis: Mdga2 haploinsufficient mice and CA1-restricted conditional knockouts show increased asymmetric synapse density, elevated mEPSCs, enhanced AMPAR currents, and impaired LTP, memory, and social behavior, with inhibitory synapse measures left unchanged (PMID:27608760, PMID:38321347). MDGA2 enforces excitatory restraint through multiple parallel cis interactions: it competes with BDNF for TrkB binding via its MAM domain to limit BDNF/TrkB-driven AMPAR-mediated transmission (PMID:40168357), and it binds EphB2 via its first three Ig domains, competing with Ephrin-B1 and forming a complex with GluN2B-containing NMDARs to suppress NMDAR-mediated postsynaptic responses (PMID:40316130). MDGA2 protein abundance is set by a SORT1-mediated lysosomal degradation pathway that is antagonized by RPS23RG1, both of which bind the MDGA2 extracellular domain (PMID:39816685). Homozygous loss-of-function variants impair MDGA2 membrane trafficking and disrupt its NLGN1 interaction, linking the gene to disease through perturbed excitatory synaptic function, consistent with an ASD-associated V930I mutation that drives aberrant BDNF/TrkB signaling (PMID:40168357, PMID:41570816). Beyond the synaptic role, MDGA2 supports rostral longitudinal growth of post-crossing commissural axons through homophilic MDGA2-MDGA2 interactions (PMID:21542908), and a separate report links it to DMAP1 stabilization and p53/p21-mediated cell cycle arrest in gastric cancer cells (PMID:26206665).

1. 2011 Medium

   Established an early non-synaptic role for MDGA2 in axon guidance, showing it acts through homophilic adhesion to steer post-crossing commissural axons.

   Evidence RNAi knockdown and function-blocking antibodies in chick spinal cord commissural neurons with binding studies

   PMID:21542908

   Open questions at the time

   • Mechanism downstream of homophilic binding unresolved
   • Relationship to later synaptic functions not addressed
   • Single lab, ortholog system
2. 2015 Medium

   Identified an unexpected intracellular tumor-suppressive function, showing MDGA2 stabilizes DMAP1 to activate p53/p21 and trigger cell cycle arrest in gastric cancer.

   Evidence Mass spectrometry, Co-IP, reporter assays, knockdown/re-expression, and xenografts in gastric cancer cells

   PMID:26206665

   Open questions at the time

   • How a cell-surface IgCAM accesses nuclear DMAP1 is unexplained
   • Not connected to neuronal functions
   • Single lab
3. 2016 High

   Defined the core synaptic mechanism, showing MDGA2 selectively suppresses excitatory synapse development by blocking the NLGN1-neurexin interaction.

   Evidence Mdga2+/- mouse model with electrophysiology, synapse density, behavior, and voltage-sensitive dye imaging

   PMID:27608760

   Open questions at the time

   • Domain basis of NLGN1 binding not mapped here
   • Mechanism of excitatory vs inhibitory selectivity unresolved
4. 2023 Medium

   Distinguished MDGA2 from MDGA1 functionally, showing endogenous MDGA2 selectively depresses NMDAR-mediated transmission while enhancing inhibitory transmission.

   Evidence Epitope-tagged knock-in mice, shRNA knockdown, CRISPR knockout, and electrophysiology (preprint)

   PMID:37720016

   Open questions at the time

   • Preprint, not peer-reviewed
   • Molecular partner mediating NMDAR depression not identified in this study
   • Apparent contrast with prior excitatory-suppression data needs reconciliation
5. 2024 High

   Confirmed cell-autonomous excitatory suppression in the mature circuit, showing CA1-specific Mdga2 deletion elevates excitatory transmission and impairs LTP, memory, and social behavior.

   Evidence Cre-lox conditional knockout in CA1 pyramidal neurons with electrophysiology, immunofluorescence, and behavioral assays

   PMID:38321347

   Open questions at the time

   • Molecular effector at the postsynapse not dissected here
   • Adult vs developmental requirement not separated
6. 2025 High

   Identified domain-resolved receptor interactions explaining excitatory suppression, showing MDGA2 competes with BDNF at TrkB via its MAM domain and with Ephrin-B1 at EphB2 via its Ig1-3 domains.

   Evidence Domain-deletion Co-IP, competition binding assays, V930I mutagenesis, AlphaFold modeling, electrophysiology, and pharmacological/peptide rescue in Mdga2+/- mice

   PMID:40168357 PMID:40316130

   Open questions at the time

   • How TrkB and EphB2 pathways are integrated with NLGN1 blockade is unclear
   • Stoichiometry of MDGA2-EphB2-GluN2B complex not resolved
   • Whether one MDGA2 molecule engages multiple receptors unknown
7. 2025 Medium

   Defined how MDGA2 protein levels are controlled, showing SORT1 routes MDGA2 to lysosomal degradation and RPS23RG1 protects it by competitive binding.

   Evidence LC-MS/MS interactome of MDGA2 ectodomain, Co-IP, siRNA knockdown, lysosome fractionation, and in vivo rescue

   PMID:39816685

   Open questions at the time

   • Signals regulating the SORT1/RPS23RG1 balance unknown
   • Whether degradation tunes synaptic strength dynamically not shown
   • Single lab
8. 2025 Medium

   Provided a direct disease link, showing homozygous loss-of-function MDGA2 variants impair membrane trafficking and disrupt NLGN1 interaction and excitatory synaptic function.

   Evidence Expression of nonsense variants in mammalian systems with trafficking assays, Co-IP, and electrophysiology in cultured hippocampal neurons

   PMID:41570816

   Open questions at the time

   • Named clinical syndrome and inheritance breadth not detailed here
   • In vivo consequences of trafficking loss not modeled

• How MDGA2's multiple parallel mechanisms (NLGN1 blockade, TrkB competition, EphB2/NMDAR suppression, homophilic adhesion) are coordinated within a single neuron, and how degradation control gates them, remains unresolved.
• No integrated model linking the distinct receptor interactions
• Relative contribution of each pathway to behavior not quantified
• Connection between the non-neuronal DMAP1/p53 role and synaptic functions unknown