Uncovering the mechanism of gastrodin in treating Alzheimer's disease through bioinformatics, machine learning, and virtual gene knockout technology.

Abstract

Alzheimer's disease (AD) is a neurodegenerative disorder characterized by progressive cognitive decline. Gastrodin is a natural compound with neuroprotective activity; however, the precise molecular mechanisms underlying its therapeutic effects on human AD remain unclear. This study employed bioinformatics approaches to systematically elucidate the therapeutic mechanisms of gastrodin in AD. Pharmacokinetic analysis showed that gastrodin exhibited favorable drug-like properties (bioavailability score: 0.55; LogP: -0.85) and an acceptable overall toxicity profile, although predicted nephrotoxicity (0.71) and ototoxicity (0.908) scores suggest potential safety concerns warranting further investigation. Next, differentially expressed genes (DEGs) associated with AD were screened from GEO datasets, yielding 13 core targets that overlapped with gastrodin-responsive genes. Functional enrichment analysis indicated significant enrichment of these targets in signaling pathways closely related to neuroinflammation and cell survival, including the Ras-proximate 1 (Rap1), Ras, mitogen-activated protein kinase (MAPK), and phosphatidylinositol 3-kinase-protein kinase B (PI3K-Akt) pathways. The Ridge model achieved AUCs of 0.898 in the training set and 0.859 in the GSE37263 cohort, but showed modest generalizability in the GSE138260 cohort (AUC = 0.684), highlighting the heterogeneity of external datasets. Further single-cell transcriptomic analysis revealed specific upregulation of insulin-like growth factor 1 receptor (IGF1R) in brain endothelial cells of AD patients. Virtual gene knockout simulations demonstrated that IGF1R knockout substantially disrupted core networks associated with endothelial functions and neurovascular homeostasis, including focal adhesion and calcium signaling pathways. Finally, molecular docking and molecular dynamics (MD) simulations revealed stable binding conformations of gastrodin to EGFR and MET, while the gastrodin-IGF1R complex exhibited a more dynamic binding behavior with conformational flexibility, suggesting an induced-fit binding mechanism. MM/PBSA calculations revealed favorable enthalpic contributions (ΔH = -17.43, -13.63, and -10.41 kcal/mol for EGFR, IGF1R, and MET, respectively) for gastrodin-target binding. In conclusion, this study suggests that gastrodin exerts multi-target effects via coordinated regulation of multiple signaling pathways. Notably, gastrodin targets IGF1R in brain endothelial cells, modulating focal adhesion/calcium signaling networks, thus providing a novel mechanistic hypothesis for AD treatment by restoring neurovascular unit homeostasis.