Specific-pathogen-free (SPF)-grade 5×FAD transgenic mice were obtained from the Model Animal Research Institute of Nanjing University (Animal Qualification Certificate No. 201400975) and kept under SPF conditions. The animal experiments were approved by the Animal Care and Use Committee of Shanghai University of Traditional Chinese Medicine (ethics number: PZSHUTCM210702001).

5×FAD transgenic mice were cross-bred with C57BL/6J mice. The mice were raised in separate cages according to their gender after 20 days of life. Their tails were docked at 30 days for genetic identification of mouse breeds, and 5×FAD mice were selected as the Acanthopanax senticosus treatment group (AS group), the donepezil hydrochloride treatment group (donepezil group), and the model group (AD group), while the C57BL/6J mice in the same litter were the blank group (WT group), with 12 mice in each group. After 2 months of regular feeding of the mice, the AS group was changed to a diet supplemented with 1.69 mg/kg of Acanthopanax senticosus (Heilongjiang Ussuri River Harbin Branch, batch no. 20210501), and the donepezil group was changed to a diet containing 3.8 g/kg of the drug, donepezil hydrochloride tablets (Sibohai, Phyllanthus Bio-Technology Co. Ltd, batch no. 21030004). All the drug-containing diets were manufactured by Jiangsu Synergy Pharmaceutical and Biological Engineering Co. Ltd., and animal tissues were acquired after 3 months of continuous feeding.

The core targets obtained from the PPI network were mapped to the CTN network to find the key compounds corresponding to the core targets. The 3D protein structure of the core target was retrieved from the PDB database (https://www.rcsb.org/) as a receptor, and the receptor protein was subjected to pre-docking preparation operations such as dehydrogenation and hydrogenation by UCSF Chimera 1.16 software. Using the abovementioned key compounds as ligands, molecular docking was performed using Auto Dock Vina 1.1.2 software, and the pairs with binding energies ≤-7.5 were visualized by Pymol 2.4.0 software.

mRNA was extracted from the cerebral cortex of each group of mice using the TRIzol method. cDNA reverse transcription was performed at 25°C for 5 min, 55°C for 10 min, and 85°C for 5 s. The PCR primers were purchased from Sangong Bioengineering (Shanghai) Co., Ltd., and the sequence of the primers (5′ to 3′) is shown in Table 1 . The reaction program was set at 95°C for 30 s, 95°C for 15 s, and 60°C for 30 s in the quantitative PCR instrument (C1000 Touc, BIO-RAD). The results were expressed as 2^-ΔΔCT values and analyzed for statistical differences.

An appropriate amount of mouse hippocampal tissue stored at -80°C in the refrigerator was taken out, RIPA lysis buffer was added, and the tissue was disrupted with an ultrasonic crusher and centrifuged at 12,000 rpm for 15 min. The supernatant was aspirated and 5× Protein Loading Buffer (Beyotime, P0015) was added. The protein samples were heated in a water bath at 100°C for 10 min. Electrophoresis was performed at 120 V for 80 min, and the membranes were transferred to a new membrane. After blocking with 5% BSA solution for 2 h, the primary antibody was incubated. The antibodies used for each sample were TRAF6 (#8028, CST, 1:1,000), anti-MAP3K7 (K008561P, Solarbio, 1:1,000), anti-p-MAP3K7 (K006194P, Solarbio, 1:1,000), P38 (#8690, CST, 1: 1,000), p-P38 (#4511, CST, 1:1,000), and HSP27 (#95357, CST, 1:1,000), overnight at 4°C in the refrigerator. After three washes in TBST solution, secondary antibodies were used: anti-rabbit IgG, HRP-conjugated antibody (#7074, CST, 1:2,000) was incubated for 1 h. The protein band signals were captured using the FL1000 Intelligent Imaging Scanning System (Thermo Fisher Scientific, USA). The protein band intensities were quantified using Image J (v1.45f) software.

The experimental data obtained from all behavioral experiments were expressed as mean ± SD ( x ¯ ± s ) , and SPSS 26.0 software was used for statistical analysis. One-way ANOVA was used for two-way comparisons between multiple groups and when P<0.05 is a statistically significant difference criterion.

The effects of AS on anxiety, nervousness, and other emotions of mice were analyzed by recording the distance traveled, average speed of movement, and number of times crossing the central zone of different groups of freely moving mice in the open-field experiment, and the results show that there was no significant difference between the distance traveled and the average speed of movement of mice in the WT, AD, donepezil, and AS groups in the behavioral experiments (P< 0.05) ( Figures 1A, B ). The number of times the mice in the AD group crossed the central zone was significantly lower than that of the WT group (5.38 ± 1.19 vs. 10.12 ± 3.31; P< 0.01), while the number of times the mice in the AS group traversed the central zone was significantly higher than that of the AD group (11.38 ± 3.11 vs. 5.38 ± 1.19; P< 0.001), whereas there was no significant difference in the donepezil group (6.25 ± 1.75 vs. 5.38 ± 1.19) ( Figure 1C ).

Mice are naturally curious and exploratory of new things, and the effect of spikenard on learning and memory impairment in demented mice could be determined by conducting new object recognition experiments. The experimental results showed that in the training phase (0 h) ( Figure 1D ), there was no significant difference in the percentage of time spent exploring unfamiliar toys (0.60 ± 0.05 vs. 0.53 ± 0.11 vs. 0.55 ± 0.06 vs. 0.57 ± 0.04; P > 0.05) and the percentage of number of explorations (0.57 ± 0.07 vs. 0.54 ± 0.11 vs. 0.56 ± 0.07 vs. 0.60 ± 0.04; P > 0.05) among the mice in the WT, AD, donepezil, and AS groups, indicating that there was no difference in the inherent curiosity for novelty among the mice in each group.

During the test phase (1 h) ( Figure 1E ), compared with the WT group, the AD group showed a decrease in the percentage of number of times of curious exploration of object Z (0.45 ± 0.07 vs. 0.55 ± 0.07; P< 0.05) and a decrease in the percentage of time spent exploring (0.46 ± 0.09 vs. 0.56 ± 0.05), but there was no significant difference; compared with AD group, the percentage of time spent exploring and the percentage of number of times of curious exploration of object Z are increased in the AS group (0. 56 ± 0.08 vs. 0.45 ± 0.07; 0.60 ± 0.07 vs. 0.46 ± 0.09; P< 0.05); compared with the AD group, the donepezil group had a significantly higher percentage of times exploring object Z (0.58 ± 0.05 vs. 0.45 ± 0.07; P< 0.01) and an increased percentage of time exploring object Z, but there was no statistically significant difference and it was lower than the AS group.

By searching the TCMIP database, the HERB database, and the Chinese Pharmacopoeia (2020 edition), 25 active ingredients ( Table 2 ) and 395 targets for AS were obtained. GeneCards database (score ≥5), OMIM database, and TTD database were used to retrieve 3563 AD-related targets. The intersection of the two databases yielded 245 targets common to AS and AD ( Figure 2A ), which belonged to a total of 25 chemical components in AS, and a compound–target network (CTN) was constructed using Cytoscape-v3.7.2 software ( Figure 2B ).

Protein interaction pairs (PPI pairs) of the abovementioned 245 targets screened in IntAct (https://www.ebi.ac.uk/intact/), BioGrid (https://thebiogrid.org/), and STRING databases were transferred to Cytoscape software to construct the protein–protein interaction (PPI) network for AS of AD. The topological parameters of each node in the PPI network were calculated using the MCODE algorithm, and a total of 19 core protein interaction cluster networks were screened, of which 85 targets related to AS for AD were identified in these 19 core protein interaction clusters. The core nodes were screened by the mean values of degree, degree centrality (DC), closeness centrality (CC), betweenness centrality (BC) and other parameters, and finally nine key targets were obtained, including APP, NTRK1, ESR1, CFTR, CSNK2A1, EGFR, ESR2, GSK3B, and PAK1, which suggest that these targets play an important role in AS for AD.

The results of the GO enrichment analysis of these nine key targets showed that ( Figure 2D ) the biological process (BP) of AS treatment of AD was highlighted in the processes of protein phosphorylation, protein autophosphorylation, positive regulation of protein phosphorylation, cellular response to estradiol stimulation, and learning or memory; cellular composition (CC) was mainly in the processes of macromolecular complexes, axons, plasma membrane, cell surface, early endosomal membranes and receptor complexes, etc.; and molecular function (MF) focuses on the role of enzyme binding, protein serine/threonine/tyrosine kinase activity, ATP binding, binding of the same proteins, estrogen receptor activity, and nitric oxide synthase regulatory activity.

The KEGG pathway analysis identified 22 relevant biological pathways that may be significantly affected by AS in the treatment of AD ( Figure 2C ). AS was found to play a role in some cancer pathways, such as breast and endometrial cancer as well as in Alzheimer’s disease pathways and pathways associated with neurodegenerative diseases. Others play important roles in the ErbB signaling pathway, the Ras signaling pathway, the MAPK signaling pathway, the PI3K-Akt signaling pathway, and so on.

To further explore the potential mechanism of AS for the treatment of AD, these nine key targets and their corresponding 16 compounds in AS were subjected to molecular docking operations. The results ( Table 3 ) showed that the binding free energies of each compound docked to the proteins were all less than -5.0 kJ/mol, and the lower the free binding energy, the higher the affinity between the protein receptor and the small molecule ligand, and the more likely the interaction would occur. This indicates that the compounds in AS have high affinity with all relevant proteins. Docking results below -7.5 kJ/mol were visualized using Pymol software ( Figure 3 ).

According to the statistical results of qRT-PCR ( Figure 4A ), the mRNA content expression of APP, NTRK1, ESR1, CFTR, CSNK2A1, EGFR, ESR2, GSK3B, and PAK1 in the hippocampus of the brains of mice in the AD group was significantly elevated compared with those of mice in the WT group (P< 0.01, P< 0.001, P< 0.0001). The mRNA expression of APP, NTRK1, ESR1, CFTR, CSNK2A1, EGFR, ESR2, GSK3B, and PAK1 in the AS group showed a significant decrease compared with that of the AD group (P< 0.001, P< 0.0001). It was demonstrated that the key targets of AS for the treatment of AD derived from network pharmacology were plausible and effective.

According to the results of the KEGG biological function analysis, the most relevant pathway for AS treatment of AD is involved in the MAPK pathway. The TRAF6, MAP3K7, p-MAP3K7, P38, p-P38, and HSP27 proteins in this pathway were selected for western blot analysis. The analysis showed that the expression level of TRAF6 protein in the hippocampal tissue of the AD group was significantly increased compared to that of the WT group, while the relative protein levels of p-MAP3K7/MAP3K7, the relative protein levels of p-P38/P38, and the expression of HSP27 protein were decreased. In contrast, in the hippocampus of mice after therapeutic intervention with AS, it can be found that mice in the AS group have a significant decrease in the expression level of TRAF6 protein and a notable increase in the expression level of MAPK phosphorylation and the HSP27 protein compared with the AD group; all the results were significant ( Figure 4B ). This suggests that AS may treat AD by regulating the phosphorylation process of the MAPK signaling pathway.