Male 5 × FAD mice (stock #34848, Qianbi Biotechnology Co., Ltd., China) and female C57BL/6 wild-type mice were bred, and the offspring genotyping was performed by polymerase chain reaction analysis of tail DNA. The primers used included APP (oIMR 3610: AGGACTGACCACTCGACCAG, oIMR 3611: CGGGGGTCTAGTTCTGCAT), PS1(oIMR 1644: AATAGAGAACGGGCAGGAGCA, oIMR 1645: GCCATGAGGGCACTAATCAT). Both male and female mice were used in this study, and they were housed in same-sex of 3–5/cage under a 12 h light–dark cycle (lights on at 8:00 am) with access to food and water at a consistent ambient temperature (21 ± 1 °C). All experimental protocols and animal handling procedures were conducted according to the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health. This study was approved by the Ethical Committee on Animal Experimentation, Fujian University of Traditional Chinese Medicine (FJTCMIACUC2019031).

This study was divided into two parts. 3-months-old 5 × FAD female and male mice were randomly divided into five groups (n = 10/group):the 5 × FAD group, 5 × FAD + EA group(5 × FAD + EA), 5 × FAD + Wnt5a interference virus group(5 × FAD + shWnt5a), 5 × FAD + EA + Wnt5a interference virus group(5 × FAD + EA + shWnt5a) and 5 × FAD + EA + empty virus group(5 × FAD + EA + shC) in each part,. Moreover, age and sex-matched littermate WT mice were classified as the WT group. Mice were evaluated for cognitive behavioral tests, including open field test (OFT), Object Location Task (OLT), Touch Screen Unique Nonmatching to Location (TUNL) task.

All viruses were obtained from BrainVTA Co., Ltd. (Wuhan, China), and refer to the literature to determine the program of interfering with virus sequence and injection [23]. The rAAV-U6-DIO-shRNA (Wnt5a)-CMV-EGFP-SV40 polyA virus were injected into the DG area of the bilateral hippocampus of 5 × FAD + shWnt5a group and 5 × FAD + EA + shWnt5a group; the rAAV-U6-DIO-shRNA (scramble)-CMV-EGFP-SV40 polyA virus were injected into the same location of 5 × FAD + EA + shC group. Coordinates of hippocampal DG area: from Bregma AP: − 2.00 mm, ML: ± 1.3 mm, DV: − 2.3 mm. The virus was delivered using a syringe pump at a rate of 20 nl/min for 10 min, for a total of 200 nl/infusion. The syringe was then raised.2 mm, and remained in place for 10 min after each injection to allow for virus diffusion, and was then slowly retracted. Virus expression was detected using a fluorescence microscope after 21 days of virus injection.

According to a previous study [27], the needle was inserted into the acupoints of Baihui (DU20) and Shenting (DU24) of the 5 × FAD + EA group, 5 × FAD + EA + shWnt5a group, and 5 × FAD + EA + shC group for 2 mm. and the electrical stimulator (G6805; SMIF, Shanghai, China) was connected, 2/20 Hz, 1 mA, 30 min/day, 5 times/week, 4 week. The WT group, 5 × FAD group, and 5 × FAD + shWnt5a group mice were fixed under the same conditions but with no EA treatment.

All mice received a single dose of 50 mg/kg BrdU (BOSTER Bio, ED1100) intraperitoneally on the 1st, 3rd, 5th …28th day of electroacupuncture (twice a day). 100 mg of BRDU powder is dissolved into a 10 mg/ml solution by physiological saline.

The experimental scheme was set according to the reference [28], OFT was carried out in a white plastic chamber (40 × 40 × 50 cm) with a camera mounted on the top to record the behavior and activity tracking of mice (Supermaze, Softmaze, China). The chamber bottom area is divided into 25 small cells by 3 equidistant horizontal parallel lines and 3 equidistant vertical parallel lines, with the middle 9 small cells defined as the central area. The mouse was singly placed in the center and allowed to move freely to habituate the arena for 5 min and the variables recorded were total Distance, central area distance, and time spent in the central part.

According to the reference [29], OLT consists of training sessions (10 min) and test session (5 min), separated by 1 h intervals. During the training session, we placed mice in an arena that contained two identical objects placed on the same side (A1 and A2). During the test session, object A2 was moved to another top so that A1 and A2 were placed diagonally. The mouse was placed in the chamber again and the time spent exploring two objects in 5 min was recorded as TA1 and TA2. Exploration is defined as the mouse sniffing or touching with its forepaws within 2 cm of the object. The test results are expressed by the identification index: identification index = TA2/(TA1 + TA2) × 100%.

The touchscreen experimental chamber consists of a touchscreen display, an infrared detector, an indoor lighting lamp, a trough tray, and a liquid dispenser. Strawberry milk was used as the liquid food reward for mice in this study (Fig. 4A). The TUNL test can only be conducted on mice after at least 3 weeks of pre-training. According to the performance of mice, 5–6 days of training should be conducted every week. With reference to the training method of Kim [30], the whole training is divided into two parts:

Pre-training stage: (1) Weight control: All mice needed food restriction and began maintaining 85–90% of free-feeding body weight throughout the experiments. (2) Environmental habituation: each mouse will explore freely for 40 min to habituate the environment. (3) Initial touch training: mice needed to complete 30 trials that touching a bright white square on the screen would result in more food rewards in 60 min. (4) Must touch training: mice need to complete 30 trials that touching will be rewarded only when white squares light up on the screen in 60 min. (5) Must initial training: the mice need to learn that the prompt light must be on at the end of the delay, and close to the food trough can trigger the random appearance of white squares. The mouse must touch 30 white squares within 45 min to receive a reward. (6) Correct touch training: The mouse must touch the white square to get a reward. If not, it will be punished by strong light. The mice completed 48 tests within 30 min for two consecutive days, and the correct rate exceeded 80% (Fig. 4B).

TUNL task: TUNL task consists of a sampling stage and a selection stage, as shown in (Fig. 4C, D). After startup, a white square appears randomly in one of the five positions on the screen. The activation was completed by blocking the infrared beam near the reward trough. After the mouse touched the white square with its nose, the stimulus was removed and began to delay. Then block the infrared beam to start the selection phase. In the selection phase, two white squares were presented, one in the wrong position in the sampling phase (wrong) and the other in the new position (correct). Touching the right position will be rewarded. The start delay (15 s) was triggered when the reward was eaten, and then the next test started. Touching a square (wrong) stimulus causes a timeout and a light penalty, after which there is a delay period before the next test. Record the total test times and the times of touching the correct position of each mouse in 45 min or the times of touching the correct position in 36 tests.

Immunohistochemistry was performed with paraffin-embedded slices and immunohistochemistry kit (MXB biotechnologies, KIT-9720) to measure Aβ. Primary antibody Aβ1-42 (Proteintech, 25524-1-AP, 1:100) was used, and DAB chromogenic solution was added, counterstained with hematoxylin, and the slices were sealed with neutral gum. The number of Aβ plaques in the DG area of the hippocampus was counted by light microscopy. Nine brain slices (three mice, three consecutive slices per mouse) from each group were taken for counting the Aβ plaque number of DG area in a microscopic field of × 100.

Immunofluorescence used paraffin-embedded sections with a thickness of 4 μm. And 0.3% Triton was added to the brain slices to permeabilize the cell membrane for 25 min. The DNA was denatured with hydrochloric acid at 37 °C for 1 h, after which the tissue was covered with 5% fetal bovine serum plus 5% goat serum blocking solution and incubated at 37 °C for 1 h. Then, primary antibodies, BrdU (Abcam, ab8955, 1:100) and CaR (Cell Signaling Technology, CST92635, 1:100), were used over at night 4 °C. Goat anti-Mouse 555 (Abcam, ab150114, 1:500), Goat anti-Rabbit 633 (Thermo Fisher, A21070, 1:500) were used as second antibodies, and nuclei were visualized with DAPI. Cells expressing both BrdU and CaR were defined as new immature GCs. The number of BrdU+/CaR+ cells in the hippocampal DG was counted by laser scanning confocal microscopy.

The Golgi staining was performed with the Golgi staining kit (Fdneurotech, PK401A). The soaked brain tissue was embedded in a frozen section embedding agent (100 um). The slides were dried at room temperature in the dark for at least 12 h. Subsequently, the slides were washed twice in double steamed water for 4 min each time. After that, they were successively dehydrated in 50%, 75%, and 90% ethanol for 4 min each time. Each gradient of ethanol was dehydrated and then placed into absolute ethanol for further dehydration, 4 times for 4 min each time. After completion of dehydration, the cells were placed in xylene 3 times for 4 min each for transparency. The stained slides were sealed and stored in the dark. Dendrites in the DG were photographed under a microscope at a magnification of × 630.

Total RNA was extracted from brain tissues using TRIzol reagent (Vazyme Biotech Co., Ltd) and converted to cDNA using the HiScript® II RT SuperMix for qPCR (+ gDNA wiper) kit (Vazyme Biotech Co., Ltd. R223). qPCR was performed with ChamQ Universal SYBR qPCR Master Mix (Vazyme Biotech Co., Ltd. Q711). All primers were purchased from SunYa biological company, Wnt5a-F(GGAACGAATCCACGATAAGG), Wnt5a-R(CAGACACTCCATGACACTTACAG). GAPDH-F(TGGAAAGCTGTGGCGTGATG), GAPDH-R(TACTTGGCAGGTTTCTCCAGG). Relative mRNA levels were calculated by normalization to the level of GAPDH. Relative gene expression was analyzed based on the fold change (the 2−ΔΔCt method).

Taken the mouse hippocampus and adjusted the concentration to 5 μg/μl after BCA quantification. Prepared 10% SDS-PAGE gel, electrophoresis, membrane transfer, and sealed 5–8% milk for 2 h. According to the instructions, incubate PVDF membrane in the primary antibody solution (GAPDH, Proteintech, 60004–1-lg, 1:5000; CaR, Cell Signaling Technology, CST92635, 1:1000; Prox1, Abcam, ab199359, 1:1000; Wnt5a, Abcam, ab229200, 1:1000; DVL2, 12037-1-AP, 1:500; p-DVL2, ybio, YB73041,1:500; FZD2, Proteintech, 24272-1-AP, 1:1500; CaMKII,Abcam, ab52476, 1:1000; p-CaMKII,Abcam, ab32678, 1:1000) over at night 4 ℃. secondary antibodies (goat anti-rabbit, Proteintech, SA00001-2, 1:5000; Proteintech, goat anti-mouse, SA00001-1, 1:5000) at room temperature for 1 h. Use developer to cover PVDF film, develop in Bio-Rad imager, use Image lab software to calculate protein integrated density, and then put it into SPSS for statistical analysis.

All test data in this study were analyzed in SPSS25.0 statistical software. If the measurement data conform to the normal distribution, the experimental data shall be expressed as mean ± standard deviation. One-way ANOVA was used for comparison between data groups. The LSD method was used to compare the two groups when the variance was homogeneous. When the conflict is homogeneous, the Games Howells method compares two groups. P < 0.05 was considered statistically significant.