AbstractBackgroundAlzheimer’s disease (AD) is a neurodegenerative disease, evolving amyloidosis and tau protein hyperphosphorylation, synapse dysfunction, neuron degeneration and cognitive deficits. Investigations based on the Aβ/tau hypothesis have been performed in transgenic animal models expression familial/mutated AD genes. However, treatments based on this hypothesis have not provided effective clinical therapies. Hyperactivity of the glutamatergic NMDA receptor (NMDAR) is a common pathophysiology of AD and has been viewed as a consequence of Aβ pathology. Among NMDAR subunits, GluN3A is a unique inhibitory subunit that suppresses NMDA currents and Ca2+ influx. Recent evidence from rodent and human studies shows that GluN3A is widely expressed and plays important functional roles in neonatal and adult brains. Ca2+ overload is a trigger of excitotoxicity, while GluN3A plays a regulatory role of Ca2+ homeostasis. The role of GluN3A regulation in AD is unknown.MethodGluN3A knockout mice, gene modifications, immunohistochemistry, Western blotting, MEA recordings, Ca2+ imaging, pathological examinations, functional and behavioral tests.ResultChronic neuronal hyperactivity and Ca2+ dysregulation caused by GluN3A deficiency lead to slowly advanced “neurodegenerative excitotoxicity”, dementia and AD pathology in an age‐dependent manner. Ca2+ imaging and MultiElectrode Array recordings in GluN3A KO brain slices revealed moderate but sustained elevation of cytosolic Ca2+ levels, neuronal hyperactivity, and impaired synaptic plasticity compared to age matched controls. Aging GluN3A KO mice showed increased Ca2+‐dependent signals, chronic inflammation, neuronal loss and disease symptoms including early onset of olfactory dysfunction followed by cognitive deficits. β‐Amyloid deposition and tau protein hyperphosphorylation were identified as a late pathology in the cortex and hippocampus. Expression of GluN3A in the GluN3A KO brain prevented AD progression. Early treatment at the pre‐clinical stage using the NMDAR antagonist memantine prevented symptoms in GluN3A KO mice. Western blot analysis of AD patient’s brains showed lower levels of GluN3A in the cortex compared to normal aging subjects.ConclusionOur results identified a novel amyloid‐independent mechanism that the deficiency of GluN3A alone causes a slight lifelong Ca2+ dyshomeostasis, leading to AD pathophysiology and amyloid pathology. The long‐term GluN3A modulation of NMDARs signifies new therapeutic targets and possible preventive interventions for a sporadic type of AD.