OBJECTIVE To explore the mechanism of Rhodiolae Crenulatae Radix et Rhizoma improving diabetic cognitive dysfunction(DCD) by network pharmacology and molecular docking, and establish a diabetic mouse model for validation.

METHODS The active ingredients of Rhodiolae Crenulatae Radix et Rhizoma were obtained from the TCMSP and HERB databases, and the relevant targets were obtained by SwissTargetPrediction and SEA database. GeneCards, TTD, and OMIM databases were used to search for targets related to DCD. Using the “Wei Sheng Xin” website, the common targets of active ingredients of Rhodiolae Crenulatae Radix et Rhizoma and the related targets of DCD were obtained by the Venn diagram. The protein interaction network was constructed and analyzed using the STRING platform. The core targets were screened after visualization with Cytoscape software, and the PPI network model was constructed. Then the intersecting targets were imported into the Metacape website for GO functional enrichment and KEGG pathway enrichment analysis. Molecular docking was performed using AutoDock software to analyze the interaction between salidroside(SAL) and the core targets. Fifty male C57BL/6J mice were randomly divided into 5 groups: normal group, model group, treatment group, solvent control group, and activator group. After 4 weeks of a high-fat diet, streptozotocin(150 mg·kg⁻¹) was intraperitoneally injected to establish a type 2 diabetes model in mice. The expression levels of core target proteins were detected by Western blotting, the expression levels of inflammation-related factors in the hippocampus were detected by ELISA, and the spatial memory and learning functions of mice were assessed using the water maze test.

RESULTS Network pharmacology screening identified 16 monomeric compounds, three of which showed potential as drugs. After screening, the main active ingredient of Rhodiolae Crenulatae Radix et Rhizoma for treating DCD was determined to be SAL. A total of 80 potential targets of SAL were obtained, with 8 968 DCD-related targets and 42 intersecting targets, including 10 core targets. GO and KEGG enrichment analysis indicated that Rhodiolae Crenulatae Radix et Rhizoma treatment of DCD mainly regulates the inflammatory response process through the EGFR/PI3K/AKT signaling pathway. Molecular docking results showed that SAL had a higher binding affinity with the core targets IL-6, IL-2, and FGF2. Compared with the model group, treatment with SAL significantly downregulated the expression levels of EGFR, p-PI3K/PI3K, and p-AKT/AKT proteins(P<0.05) and obviously inhibited the expression of IL-6, IL-2, and FGF2 factors at both mRNA and protein levels(P<0.05). The results of the water maze test showed that, compared with the model group, treatment with SAL remarkably improved the learning and memory abilities of diabetic mice(P<0.05). Compared with the treatment group, the activator group significantly reversed the related indicators, reducing the protective effect of SAL on DCD.

CONCLUSION The possible mechanism by which Rhodiolae Crenulatae Radix et Rhizoma improves DCD is that SAL regulates the expression of IL-6, IL-2, and FGF2 by modulating the EGFR/PI3K/AKT signaling pathway, thereby reducing neuroinflammatory response.