摘要
Vascular cognitive impairment and vascular dementia (VCI/VaD) arise from chronic cerebral hypoperfusion, yet the molecular events that convert a vascular insult into neuronal loss remain undefined. Using the asymmetric common carotid artery stenosis (ACAS) mouse model, we identify vascular hydrogen peroxide (H₂O₂) as the initiating driver of this transition. Longitudinal laser speckle imaging reveals progressive hypoperfusion over 35 days, accompanied by an early rise in H₂O₂ (POD 0–14) potentially sourced from vascular NADPH oxidase 4 (NOX4). This H₂O₂ impairs cerebrovascular function by upregulating the vasoconstrictor endothelin-1 (ET-1) and suppressing NO-dependent vasodilatory pathways, leaving vessels unable to dilate on demand. Once vascular H₂O₂ crosses a threshold, astrocytes undergo a mild-to-severe reactive transition and induce iNOS. We propose that astrocytes release NO to restore flow, but because the vasculature is unresponsive, the excess NO instead drives oxidative and nitrosative injury (8-OHdG, 3-nitrotyrosine) and neuronal apoptosis. Severe reactive astrocytes impair CA1 synaptic transmission, excitability, and spike probability, culminating in neurodegeneration and memory deficits. Scavenging H₂O₂ with KDS12025 preserves cerebrovascular function, prevents the astrocyte switch, protects neurons and white matter, and rescues cognition. Pharmacological (L-NIL) and genetic (shRNA) inhibition of iNOS further shows that astrocytic nitrosative stress is necessary for neuronal death and cognitive impairment despite persistent vascular dysfunction, dissociating the executioner step from the upstream insult. Together, these findings define a gliovascular H₂O₂–iNOS axis in which vascular oxidative stress gates a neurotoxic astrocyte state, and nominate H₂O₂, NOX4, ET-1, and iNOS as therapeutic targets for vascular dementia.
