Exploring the mechanism of action of Kaixin San in treating dementia based on network pharmacology and molecular docking

Background: Dementia is a prevalent neurodegenerative disorder characterized by cognitive decline, imposing heavy socioeconomic burdens globally. Kaixin San, a classic traditional Chinese medicine formula, has been clinically used for cognitive impairment, but its underlying therapeutic mechanism against dementia remains poorly understood. Therefore, this study aimed to elucidate the therapeutic mechanism of Kaixin San against dementia using integrated network pharmacology and molecular docking approaches. Methods: Active constituents of Kaixin San were retrieved from TCMSP and HERB databases, with target prediction conducted via PubChem, SwissTargetPrediction, and UniProt. Dementia-associated targets were acquired from GeneCards, OMIM, TTD, and PharmGKB. The intersection of the active ingredient targets of action and dementia target of Kaixin San were taken by Veen diagram, and they were imported into the STRING database and analyzed to construct the PPI network. Core targets were subsequently screened via 3 different Cytoscape plugins CytoNCA, CytoHubba, and MCODE. Functional enrichment analyses (GO and KEGG) were executed on the DAVID platform. Cytoscape facilitated the construction of both “drug-component-target-disease” and “formula-target-pathway-disease” networks. Molecular docking simulations employed AutoDock Vina, with visualizations rendered in ChimeraX. Results: We identified 87 bioactive components and 743 putative targets of Kaixin San, with 403 overlapping targets selected as potential therapeutic targets (common targets) for dementia. Four pivotal constituents emerged: Kaempferol, 1,2,3,6,7-pentamethoxyxanthone, 1-hydroxy-3,7-dimethoxyxanthone, and 1-Peroxyferolide.Using different plugin methods, the PPI network revealed 7, 10, and 15 core targets respectively, with AKT1, STAT3, and ESR1 exhibiting the highest centrality. GO functional enrichment yielded 71 cellular component terms, 18 molecular function terms, and 30 biological process terms. KEGG pathway enrichment analysis indicated 76 signaling pathways, notably featuring the ErbB cascade. Molecular docking demonstrated binding affinities mostly ≤ −5.0 kcal/mol, confirming stable ligand-target interactions, particularly between 1-Peroxyferolide and ESR1. Conclusion: Kaixin San likely mitigates dementia through flavonoid/xanthone-mediated modulation of core targets (AKT1, STAT3, ESR1) and critical pathways (ErbB, PI3K-AKT), thereby alleviating behavioral deficits, conferring neuroprotection, and attenuating amyloid-β (Aβ) deposition.