| 2025 |
Cryo-EM structures of Ca2+-permeable GluA3 homomers reveal a unique architecture where the N-terminal domain (NTD) and ligand-binding domain (LBD) tiers are closely coupled throughout gating states. A stacking interaction between two Arg163 residues in the NTD dimer interface traps a unique NTD dimer conformation enabling close NTD-LBD contacts. Rupture of the Arg163 stack alters extracellular region structure/dynamics and increases GluA3 heteromer trafficking to synapses. A mammalian-specific GluA3 trafficking checkpoint determines LBD tier conformational stability. |
Cryo-electron microscopy with functional validation (mutagenesis, trafficking assays) |
Nature |
High |
40592473
|
| 2017 |
GluA2/3-containing AMPARs are in a low-conductance state under basal conditions and contribute little to synaptic currents despite being present at synapses. When intracellular cAMP levels rise (e.g., via β-adrenergic receptor activation), GluA2/3 channels shift to a high-conductance state, producing synaptic potentiation. This cAMP-driven potentiation requires both PKA and the GTPase Ras activation. |
Electrophysiology in mouse hippocampal neurons (CA1), pharmacological manipulation of cAMP/PKA/Ras pathways, GluA3 knockout comparison |
eLife |
High |
28762944
|
| 2017 |
Cerebellar LTP at the parallel-fiber-to-Purkinje-cell synapse and vestibulo-ocular reflex adaptation depend on GluA3-containing AMPARs, not GluA1-containing AMPARs. This LTP does not require GluA1 AMPAR trafficking but instead requires changes in open-channel probability of GluA3-AMPARs mediated by cAMP signaling and Epac activation. |
GluA3 and GluA1 knockout mice, electrophysiology, VOR behavioral assay, pharmacological cAMP/Epac manipulation |
Neuron |
High |
28103481
|
| 2016 |
Aβ oligomer-mediated synaptic depression, spine loss, and blockade of LTP all require the presence of GluA3-containing AMPARs. Hippocampal neurons lacking GluA3 are resistant to Aβ-mediated synaptic depression and spine loss. Aβ-overproducing mice show memory impairment that is absent in GluA3-deficient congenics. |
GluA3 knockout mice crossed with Aβ-overproducing AD mouse model; electrophysiology, spine imaging, behavioral memory tests |
Proceedings of the National Academy of Sciences of the United States of America |
High |
27708157
|
| 2025 |
Aβ oligomers trigger endocytosis of GluA3-containing AMPARs and promote their translocation to endolysosomal compartments for degradation. These effects critically depend on the PDZ-binding motif of GluA3; a single point mutation in the GluA3 PDZ-binding motif prevents Aβ-driven endocytosis and renders synapses fully resistant to Aβ. Proteomics of APP/PS1 transgenic mouse synaptosomes confirmed selective early reduction of GluA3. |
Electrophysiology, AMPAR imaging, site-directed mutagenesis of PDZ-binding motif, live-cell endocytosis assays, synaptosome proteomics from APP/PS1 mice |
The Journal of neuroscience |
High |
39779375
|
| 2010 |
Single-channel analysis of homomeric GluA3 receptors shows activation by glutamate and the partial agonist fluorowillardiine to the same three open conductance levels but with different open probabilities. Five modes of channel activity were identified, analyzable with a kinetic model of three closed states and two open states per conductance level. |
Single-channel electrophysiology (patch clamp) with agonist concentration series and X-means algorithm sorting |
Biophysical journal |
High |
20816055
|
| 2010 |
Crystal structures of the flop-selective allosteric modulator PEPA bound to the ligand-binding domains of GluA2 and GluA3 flop isoforms reveal that flop selectivity is conferred by bidentate hydrogen bonding between PEPA and N754 (asparagine in flop; serine in flip). Five subsites on the binding surface contribute to stoichiometry, orientation, and functional outcome of modulator binding. |
X-ray crystallography of GluA2 and GluA3 LBD-PEPA complexes |
Biochemistry |
High |
20199107
|
| 2018 |
Druggability simulations identified a novel ligand-binding site specific to the GluA3 AMPAR N-terminal domain (NTD), arising from its unique conformational flexibility. Crystal structures of GluA3 NTD were trapped in vastly different conformational states, revealing pharmacophoric features not present in other AMPAR subunits. |
Molecular dynamics druggability simulations plus X-ray crystallography of GluA3 NTD in multiple conformations |
Structure |
High |
30528594
|
| 2016 |
Poor surface expression of homomeric GluA3 receptors is caused by nonproductive assembly and aggregation, driven by residues Tyr-454 and Arg-461 in the ligand-binding domain. Co-assembly with GluA2 (but not stargazin) rescues surface expression. GluA3 homomers show a Tyr-454/Arg-461-dependent tendency to aggregate detectable by detergent solubility and density-gradient centrifugation. |
Biochemical fractionation (nonionic detergent solubility, density gradient centrifugation, native gel electrophoresis), surface expression assays, site-directed mutagenesis of LBD residues |
The Journal of biological chemistry |
High |
26912664
|
| 2017 |
ABHD6 suppresses membrane delivery and glutamate-induced currents of GluA2- and GluA3-containing AMPA receptors. Pull-down experiments show ABHD6 binds directly to GluA1-3 via their C-terminal domains; deletion of the GluA C-terminus abolishes both ABHD6 binding and its inhibitory effects on surface expression and currents. |
HEK293T overexpression, electrophysiology, surface biotinylation assays, pull-down (ABHD6 binding to GluA C-termini), C-terminal deletion constructs |
Frontiers in molecular neuroscience |
Medium |
28303090
|
| 2022 |
Interaction proteomics on hippocampi from wildtype vs. Gria3 knockout mice shows that GluA2/3-containing AMPARs preferentially co-purify CNIH-2, TARP-γ2, and Noelin1 (Olfactomedin-1), whereas GluA1/2 receptors preferentially co-purify TARP-γ8, SynDIG4, and CNIH-2. TARP-γ8 and SynDIG4 interact directly and co-assemble into an AMPAR subcomplex at synaptic sites. |
Affinity purification mass spectrometry (interaction proteomics) from Gria1- or Gria3-KO mouse hippocampi; co-immunoprecipitation and fractionation for TARP-γ8/SynDIG4 interaction |
Cells |
Medium |
36429079
|
| 2017 |
In cochlear nucleus, GluA3 subunits are concentrated at the center of auditory nerve-bushy cell (AN-BC) synapses at higher density than GluA4, whereas GluA4 is more abundant at AN-fusiform cell synapses. In GluA3-knockout mice, the central intrasynaptic distribution of AMPARs is lost and gold particles distribute evenly, demonstrating GluA3 specifically organizes the central clustering of AMPARs at AN-BC synapses. |
Quantitative freeze-fracture replica immunogold labeling (FRIL) of cochlear nucleus synapses in wildtype and GluA3-KO mice |
Brain structure & function |
High |
28397107
|
| 2020 |
GluA3 determines the ultrafast kinetics of endbulb-bushy cell synapse glutamatergic currents by promoting insertion of postsynaptic AMPARs containing fast-desensitizing flop subunits. GluA3 is also required for normal presynaptic terminal function, structure, and development: GluA3 KO mice show altered short-term depression, increased vesicle release probability, impaired vesicle replenishment, and reduced readily-releasable pool. GluA3 makes the speed of synaptic depression rate-invariant. |
Patch-clamp electrophysiology at endbulb synapses in wildtype vs. GluA3 KO mice, electron microscopy of presynaptic terminals |
The Journal of neuroscience |
High |
32051325
|
| 2016 |
Deletion of GluA3 leads to impaired auditory signaling (decreased ABR peak amplitudes, increased peak 2 latency, early hearing loss), reduced number and size of postsynaptic densities at AN-BC synapses, and hampered ABR threshold recovery after transient ear plugging, demonstrating GluA3 is required for normal auditory processing and experience-dependent synaptic plasticity. |
Auditory brainstem responses (ABR) and electron microscopy of AN-BC synapses in wildtype vs. GluA3-KO mice, with transient monaural earplugging paradigm |
Hearing research |
Medium |
28011083
|
| 2017 |
A missense variant A653T in the GRIA3 ion conduction pore (transmembrane domain) stabilizes the channel in a closed conformation in vitro, opposite to the Lurcher mutation effect. Knock-in of the orthologous mutation in mice by CRISPR-Cas9 causes significantly altered sleep/activity structure with fewer brief bouts and enhanced period lengthening under constant light. |
In vitro electrophysiology of A653T mutant channels; CRISPR-Cas9 knock-in mouse model with actigraphy/sleep analysis |
Human molecular genetics |
High |
29016847
|
| 2021 |
A gain-of-function variant R660T in GRIA3 causes slower desensitization and deactivation kinetics with substantial non-desensitized steady-state currents in homomeric GluA3 and in GluA2/GluA3 heteromers expressed in HEK cells. In cultured cerebellar granule neurons and hippocampal CA1 neurons expressing R660T, mEPSCs and evoked EPSCs are significantly slower than wildtype. |
Two-electrode voltage clamp in Xenopus oocytes, patch clamp in HEK cells and primary neurons, in utero electroporation into hippocampal CA1 neurons |
PLoS genetics |
High |
34161333
|
| 2024 |
Electrophysiological assays of 43 GRIA3 missense variants found in NDD patients showed 31 alter receptor function: some are loss-of-function (reduced currents) and some are gain-of-function (increased/prolonged currents). GOF variants associate with earlier seizure onset, hypertonia, and movement disorders; LOF variants associate with later seizure onset, hypotonia, and sleep disturbances, establishing distinct NDD phenotypes for each functional class. |
Electrophysiological assays (patch clamp/voltage clamp) of 44 variants expressed in heterologous cells; clinical correlation across 25 patients |
Brain |
High |
38038360
|
| 2020 |
The GRIA3 missense variant Gly826Asp (c.2477G>A) in transmembrane domain 4 causes decreased current response to agonists (kainic acid and glutamate) and reduced cell-surface expression, consistent with a loss-of-function mechanism, as shown by two-electrode voltage clamp in Xenopus oocytes and a β-lactamase reporter assay in HEK293 cells. |
Two-electrode voltage clamp in Xenopus laevis oocytes; β-lactamase reporter assay in HEK293 cells for surface expression |
Movement disorders |
Medium |
32369665
|
| 2022 |
Two GRIA3 missense variants, G630R and E787G, found in male patients with aggressive behavior, completely abolish GluA3 ion channel function. In GluA3 KO mice, excitatory neurotransmission and neuronal activity in the medial prefrontal cortex (mPFC) are impaired, and viral re-expression of GluA3 in the mPFC alleviates aggressive behavior. |
Patch-clamp electrophysiology of mutant channels in heterologous cells; GluA3 KO mice with viral rescue (AAV-mediated GluA3 re-expression in mPFC); aggression behavioral assays |
Molecular psychiatry |
High |
35697757
|
| 2022 |
A gain-of-function GRIA3 variant (p.Ala615Val) causes slower desensitization and deactivation kinetics in patch-clamp recordings. In a Drosophila model, expression of a double mutant (Ala615Val + Lurcher, which makes channels leaky) caused developmental defects, while either single variant alone did not, confirming partial GOF. The patient's seizures and hypertonia were ameliorated by carbamazepine, which inhibits glutamate release from presynapses. |
Patch-clamp recordings of mutant GluA3 in heterologous cells; Drosophila genetic model with epistasis; clinical pharmacological intervention |
Human genetics |
Medium |
35031858
|
| 2025 |
α2δ-1 (but not α2δ-2 or α2δ-3) co-expression diminishes GluA3 AMPAR currents and protein levels via ubiquitin-proteasome-mediated degradation. K861 at the GluA3 C-terminus is identified as a key ubiquitination site. Nerve injury increases GluA3 ubiquitination and reduces GluA3 and GluA2/GluA3 heteromer levels in spinal cord. Intrathecal Gria3 gene delivery reverses nerve injury-induced nociceptive hypersensitivity and restores GluA2/GluA3 heteromers. |
Co-expression in heterologous cells, proteasome inhibition, ubiquitination assays, site-directed mutagenesis of K861, nerve injury mouse model with intrathecal Gria3 gene delivery, electrophysiology |
The Journal of clinical investigation |
High |
41129242
|
| 2010 |
GRIA3 is a downstream transcriptional target of CUX1 in pancreatic cancer cells. Knockdown of GRIA3 reduces proliferation and migration and enhances apoptosis; overexpression has the opposite effects. GRIA3 knockdown reduces xenograft tumor growth in vivo. AMPAR antagonists (GYKI52466, SYM2206) decrease pancreatic cancer cell survival. |
RNAi loss-of-function screen, GRIA3 overexpression/knockdown in vitro, xenograft mouse model, AMPAR antagonist treatment |
Neoplasia |
Medium |
20689760
|
| 2019 |
Anti-GluA3 autoantibodies in FTD patients decrease glutamate release and alter levels of GluA3-containing AMPAR, accompanied by changes in scaffolding proteins involved in receptor synaptic retention/internalization, as shown in FTD patient brain specimens and confirmed by TMS measures of glutamatergic neurotransmission. |
Molecular/neurochemical analyses on patient brain specimens (FTLD-tau neuropathology); transcranial magnetic stimulation; CSF glutamate/serine measurements |
Neurobiology of aging |
Medium |
31784278
|
| 2017 |
CSF containing anti-GluA3 antibodies from FTD patients decreases GluA3 subunit synaptic localization of AMPARs and causes loss of dendritic spines in rat hippocampal primary cultures and hiPSC-derived neurons. Reduced GluA3 in the postsynaptic fraction is accompanied by increased neuronal Tau levels. |
Rat hippocampal primary neuronal cultures and hiPSC-derived neurons incubated with patient CSF; immunofluorescence for GluA3 synaptic localization; dendritic spine counting |
Scientific reports |
Medium |
28751743
|
| 2025 |
Long-term exposure to human anti-GluA3 IgGs causes delocalization of GluA3-containing AMPARs to extrasynaptic sites, dendritic arbor reorganization, increased dendritic spine number with altered morphology, enhancement of NMDA receptor-mediated postsynaptic Ca2+ currents, and increased nuclear phospho-CREB levels in primary rat hippocampal neurons. |
Primary rat hippocampal neurons, immunofluorescence for GluA3 localization, dendritic morphology analysis, Ca2+ imaging, phospho-CREB immunostaining |
Neurobiology of disease |
Medium |
39954743
|
| 2011 |
After 7 days of monaural conductive hearing loss (earplugging), immunolabeling for GluA2/3 (but not GluA2 alone) increases on bushy cells and fusiform cells of both ipsilateral and contralateral cochlear nuclei, while GlyRα1 is downregulated in the same cells, indicating activity-dependent regulation of GluA3-containing AMPAR composition at auditory synapses. |
Monaural earplugging in rats, quantitative immunohistochemistry and biochemistry with subunit-specific antibodies |
Neuroscience |
Medium |
22044924
|
| 2024 |
Loss-of-function of Gria3 in mice profoundly alters synapse protein composition, and transcriptomic analysis shows activity-regulated genes are downregulated in cortical regions while immune/glia-related pathways show brain-region-specific changes, distinct from Grin2a mutant mice despite both encoding glutamate receptor subunits associated with schizophrenia risk. |
Gria3 protein-truncating knock-in mouse model; transcriptomics (RNA-seq); synaptosome proteomics |
bioRxivpreprint |
Medium |
|
| 2024 |
ABHD6 accelerates the deactivation and desensitization of GluA2(R)/GluA3(R) heteromeric receptors in the presence of TARP γ-2 (but not in its absence), as shown by ultra-fast glutamate application and outside-out patch recordings. |
Ultra-fast agonist application, outside-out patch recordings in HEK cells expressing GluA2/GluA3 heteromers with/without TARP γ-2 and ABHD6; ABHD6-KO neurons |
bioRxivpreprint |
Medium |
|