Abstract

Introduction: Cognitive dysfunction is a common complication of the central nervous system in diabetic patients. To explore the molecular mechanism of diabetic cognitive dysfunction (DCD). Material and methods: The db/db mice and high-glucose (HG)-treated human brain astrocytes (SVG p12) and mouse hippocampal neuron cells (mHNCs) were used for mechanistic exploration. Learning and memory abilities in mice were tested using the Morris water maze assay. The expression of lncRNA NEAT1 and miR-30e-5p and the mRNA levels of FLT1, NGF, and GPX4 were detected by RT-qPCR. Western blotting was used to analyze FLT1 and p-tau protein levels. Cell viability and apoptosis were evaluated using the CCK-8 method and flow cytometry. Ferroptosis was assessed by measuring Fe2+ and 4-HNE levels and GPX4 expression. Target sites among NEAT1, miR-30e-5p, and FLT1 were detected using a dual-luciferase reporter assay, and their predictive value for DCD was assessed using ROC curves and binary logistic regression analysis. Results: The db/db mice showed significant cognitive deficits. NEAT1 and FLT1 were highly expressed in the hippocampus of db/db mice and HG-treated nerve cells, whereas miR-30e-5p was expressed at low levels. Silencing NEAT1 rescued HG-induced nerve cell viability impairment and ferroptosis. NEAT1 directly targeted miR-30e-5p and negatively regulated it, and miR-30e-5p directly targeted FLT1. The miR-30e-5p inhibitor partially reversed the effects of NEAT1 silencing, whereas silencing FLT1 rescued the effects of the miR-30e-5p inhibitor. NEAT1, miR-30e-5p, and FLT1 were classifiers for differentiating patients with diabetes mellitus from those with DCD, with areas under the curves (AUCs) of 0.718, 0.816, and 0.716, respectively. Their combination showed more reliable diagnostic performance (AUC = 0.896). Conclusions: The NEAT1/miR-30e-5p/FLT1 axis may mediate DCD pathogenesis by regulating ferroptosis in brain nerve cells under HG conditions. These molecules may serve as