The study was conducted in accordance with the Declaration of Helsinki, and approved by the by the Animal Experimental Ethics Inspection of Guangzhou Sport University (protocol code 2024DWLL-38 and 10 July 2024 of approval). Healthy 8-week-old male C57BL/6J (n = 11) and ApoE^−/− mice (n = 22) were purchased from Cyagen Biotechnology Co., Ltd (Suzhou, China). All mice were housed and trained in a specific pathogen-free (SPF) laboratory animal facility at the Guangzhou sport university science experiment center, with ad libitum access to water and maintained on a 12-h light/dark cycle. Following the experiments, the mice were euthanized via cervical dislocation under 2.5% isoflurane anesthesia.

In this study, ApoE^−/− mice subjected to HFHCD were designated as AS group, whereas C57BL/6J mice maintained on a standard diet served as the wild-type (WT) control group. After a 6-week feeding period, initial cognitive function assessments were conducted for both groups, utilizing the Y-maze, open-field, and three-chamber social tests. Subsequently, three mice from each group were randomly selected for dissection. Aortic atherosclerotic plaques were examined using optical microscopy. The remaining mice (n = 16) in the AS group were randomly allocated into AS-SED and AS-EX subgroups (n = 8 per group), while continuing their original diet, and the WT group persisted on the standard feed (n = 8). The AS-EX subgroup participated in a 10-week aerobic exercise intervention, with the AS-SED and WT groups serving as sedentary controls (Figure 1A). Upon completion of the 10-week treadmill aerobic exercise intervention, behavioral assessments were repeated, followed by cerebral blood flow evaluations. Subsequently, the aortas were excised for Oil Red O staining to evaluate plaque area, and the complete brains along with the hippocampi were collected from mice across all three experimental groups.

Building upon previous research and made modifications (Pynn et al., 2004), this experiment implemented a moderate intensity aerobic exercise protocol using a zero-degree incline treadmill. The mice were subjected to exercise at a speed of 14 m/min. Considering that the subjects involved in the experiment were atherosclerotic model mice, we arranged for them to rest for 2 min after every 15 min of exercise. This cycle was repeated 4 times, resulting in a total exercise duration of 68 min per session. The exercise was scheduled for 5 days per week over a period of 10 weeks.

The experiment was performed using a Y-shaped maze. Each mouse was positioned at the starting point of one arm of the maze and permitted to explore freely for a duration of 8 min. The parameters evaluated included: (1) the total number of arm entries, and (2) the count of spontaneous alternations. Locomotor activity was quantified by the total number of arm entries, whereas the number of spontaneous alternations served as an indicator of the mice’s short-term memory capacity (Farkas et al., 2022).

The experiment is performed in an open square arena, where the subject animal is positioned at the center, and its spontaneous activity is monitored using a video tracking system over a duration of 15 min. The primary parameters measured during the test include the total distance traveled, the duration of time spent in the central area, and the duration of time spent in the peripheral corners. These metrics are instrumental in assessing the animal’s exploratory behavior, anxiety levels, and motor capabilities (Wang et al., 2023).

The three-chamber social test is primarily used to assess the social capabilities, social memory, and preference for novelty in mice (Kaidanovich-Beilin et al., 2011). The experimental setup consists of three transparent rectangular chambers. The experiment is divided into two phases:

Phase 1: A novel mouse (A) is placed in a circular wire cage on one side, while the wire cage on the opposite side remains unoccupied. At the commencement of this phase, the test mice are introduced from the central chamber and permitted to freely explore all three chambers for a duration of 10 min. Upon the conclusion of this phase, the three chambers are disinfected with alcohol to remove any residual odors left by the mice. Phase 2: A second novel mouse (B) is introduced into the previously unoccupied wire cage, and the test mice are once again introduced from the central chamber to explore the three chambers for an additional 10 min. During the experiment, an interaction with the wire cage is deemed a valid investigative behavior when the test mice make direct contact with the cage or remain within a 3–5 cm proximity to it. The following metrics are calculated: the frequency of sniffing behaviors directed towards mice A and B, expressed as a percentage of total interactions.

Effective anesthesia is initially achieved through isoflurane inhalation, after which the mice are securely positioned on a specially designed experimental board. The hair on the mice’s heads is subsequently removed using depilatory cream, followed by a careful incision of the scalp to fully expose the skull. Any residual hair and debris on the skull surface are meticulously cleaned using cotton swabs moistened with distilled water. Following this preparatory procedure, LSCI is employed (Simopto, China), wherein a laser is directly applied to the surface of the mice’s skull, and a continuous series of 30 speckle images is captured for each subject.

Following the measurement of cerebral blood flow, the thoracic cavity of the mouse was accessed to expose the heart. A perfusion needle was inserted into the apex of the heart, and a swift incision was made in the right auricular region. Saline perfusion was initiated, and upon observing a lightening of the liver color, 4% paraformaldehyde (PFA) was introduced until the limbs exhibited rigidity. Subsequently, the aorta and intact brain tissue were excised. The intact brain tissue was then immersed in 4% PFA fixative for a duration of 24 h to ensure adequate tissue fixation. In parallel, hippocampal tissue from the remaining mice was isolated and promptly placed into sterile centrifuge tubes containing RNA preservation buffer to preserve RNA integrity for subsequent transcriptome sequencing.

The excised aortas were fixed in 4% PFA for a duration of 24 h. Following fixation, the specimens were rinsed three times with phosphate-buffered saline (PBS), with each wash lasting 3 min. Subsequently, the aortic specimens underwent incubation in an Oil Red O staining solution at 40°C for a period of 1 h. Post-staining, the samples were washed twice with 75% ethanol. During the differentiation process, the samples were rinsed twice with distilled water to eliminate any residual ethanol. Finally, the stained aortic samples were promptly observed and photographed to document and preserve the experimental results.

The hippocampal tissues of mice in the AS-SED and AS-EX groups were used to extract total RNA, and four samples from each group were analyzed using RNA sequencing. The RNA sequencing and subsequent data analysis were conducted by Applied Protein Technology Co., Ltd.

In this experiment, total RNA was extracted from hippocampal tissue using the Trizol reagent (Magen, China). Reverse transcription was performed on the extracted RNA using the RT SuperMIX for RT-qPCR kit (G3337, Servicebio, China) to synthesize the corresponding cDNA. PCR was conducted using the SGExcel FastSYBR RT-qPCR Premix Kit (B532955, Sangon biotech, China). The primer sequences are listed in Table 1.

The ABI StepOne Plus cycler, along with StepOne software, was used for amplification and detection. The qPCR procedure began with a 3-min step at 95°C, then proceeded with 40 cycles of 5 s at 95°C and 20 s at 60°C. GAPDH was used as the internal standard control to normalize gene expression using the 2^−ΔΔCt method (Yang et al., 2022).

Following a 24-h fixation period in 4% PFA, the brain tissues underwent dehydration through a gradient series of sucrose solutions with concentrations of 15%, 20%, and 30%, each for 24 h. Subsequently, the tissues were cryopreserved in OCT embedding medium (Servicebio, China) and sectioned into 16 μm continuous coronal slices using a cryostat microtome (Leica Microsystems, Germany). The primary antibodies utilized included anti-Iba1 antibody (1:500, GB12105, Servicebio, China), anti-CD68 antibody (1:200, D194559, Sangon Biotech), and anti-IL-33 antibody (1:100, 12372-1-AP, Proteintech, China). The secondary antibodies employed were Alexa Fluor 488-conjugated goat anti-mouse IgG fluorescent secondary antibody (green, 1:400, GB25301, Servicebio, China), CY3-conjugated goat anti-mouse IgG fluorescent secondary antibody (red, 1:300, GB21301, Servicebio, China), along with DAPI staining solution (G1012, Servicebio, China). The examination and imaging were performed using an Olympus confocal microscope (OLYMPUS FV31S-SW, Japan).

Hippocampal tissues from mice were lysed utilizing a lysis buffer (P0013B, Beyotime Biotechnology). Protein concentrations were subsequently determined for Western blot (WB) analysis. The primary antibodies employed included anti-IL-33 antibody (1:1000, 12372-1-AP, Proteintech, China), anti-IκB alpha (18220-1-AP, Proteintech, China), anti-phospho-IκB alpha (68999-1-Ig, Proteintech, China), anti-NF-κB (GB11997, Servicebio, China), anti-phospho-NF-κB (GB113882, Servicebio, China), and anti-Gapdh (GB12002, Servicebio, China). The secondary antibody used was anti-Rabbit IgG (GB111738, Servicebio, China). Image-Pro Plus version 6.0 was used to quantify the chemiluminescent signal.

Data analysis was performed with GraphPad Prism 9.5.1 software. Each experiment or group’s sample number (n) is noted in the figure legends. Among three groups, statistical analysis was performed using one-way ANOVA with Tukey’s multiple comparisons test. Comparisons between two groups were analyzed using unpaired and two-tailed Student’s t-test. All data are presented as the mean ± SEM. Video tracking system (shxinruan, China) for recording and analyzing behavioral tests. Values of P < 0.05 were considered statistically significant.

The experimental design and grouping are shown in (Figure 1A). Macroscopic examination of the aorta demonstrated that 6 weeks on HFHCD (pre-intervention modeling phase) resulted in the formation of significant atherosclerotic plaques in the AS group, whereas the WT group, which was fed a standard diet, did not develop such plaques. This finding initially confirmed the successful establishment of the atherosclerotic model in the AS mice (Figure 1B). In the Y-maze test (Figures 1C,D), both the total number of entries and the number of alternations made by the AS group were significantly lower than those of the WT group, indicating impairments in spatial memory and decision-making abilities in the AS mice. Furthermore, results from the three-chamber social test revealed deficits in social behavior in the AS mice (Figures 1E,F). During both the first and second phases of the test, the WT group mice exhibited a significantly higher percentage of sniffing interactions with mice A and B compared to the AS group mice, suggesting that AS may adversely affect the social cognitive function of AS mice. In conclusion, the open-field test showed that AS group mice traveled significantly less distance than WT group mice, suggesting reduced spontaneous activity in the AS mice. Furthermore, the AS group mice exhibited a marked decrease in time spent in the central area and an increase in time spent in the peripheral corners. This behavior may suggest heightened anxiety levels, as exploration of the central area is generally associated with lower anxiety, whereas exploration of the peripheral corners is linked to higher anxiety (Figures 1G,H). These findings indicate that 6 weeks of HFHCD resulted in an increased atherosclerotic plaque area in ApoE^−/− mice and the onset of cognitive impairments. Thus, mice with atherosclerotic pathology and differences in cognitive function meet criteria for subsequent experiments before proceeding with exercise intervention.

The AS-EX group showed a notable decrease in atherosclerotic plaque area as indicated by Aortic Oil Red O staining, compared to the AS-SED group (Figures 2A,B). Following the exercise intervention, the Y-maze test (Figures 2C,D) demonstrated that the AS-EX group exhibited a significantly higher number of total entries and alternations relative to the AS-SED group, suggesting that exercise improved spatial memory and decision-making abilities in AS mice. The results from the three-chamber social test further corroborated the positive effects of exercise intervention on the social cognition of AS mice (Figures 2E,F). During the initial phase of the test, the AS-EX group spent a significantly greater proportion of times sniffing mouse A compared to the AS-SED group. In the subsequent phase, the AS-EX group also spent a significantly greater proportion of times sniffing mouse B compared to the AS-SED group. Lastly, the open field test indicated a significant increase in the total distance traveled by the AS-EX group compared to the AS-SED group, suggesting that exercise intervention substantially enhanced the spontaneous activity of AS mice. Furthermore, the AS-EX group exhibited a marked increase in the duration spent in the central area and a reduction in time spent in the peripheral corners relative to the AS-SED group, indicating an amelioration of anxiety levels in AS mice following the exercise intervention (Figures 2G,H). Cerebral blood flow was assessed in mice via the sagittal suture to provide an indication of global cerebral perfusion (Figure 2I). The blood flow index in the AS-SED group was significantly lower than that observed in the WT group, whereas the AS-EX group demonstrated a significant enhancement in cerebral blood flow compared to the AS-SED group (Figure 2J). These findings suggest that exercise facilitates improvements in cerebral blood flow in AS mice.

Based on the results of RNA-Seq experiments, 103 differentially expressed genes were identified in the hippocampal tissue of AS-SED and AS-EX mice, with 33 genes being upregulated and 70 genes downregulated (Figure 3A). To visually represent the statistical significance and magnitude of expression changes of these differential genes, we constructed a volcano plot and a heatmap (Figures 3B,C). To further elucidate the potential biological functions of these differentially expressed genes, we conducted Gene Ontology (GO) enrichment analysis and presented the top 10 functional terms, ranked by p-value, in a bubble chart (Figure 3D). This analysis revealed associations of these genes with specific biological processes, molecular functions, or cellular components. Additionally, we identified the top 20 significantly enriched pathway terms and depicted them in a KEGG enrichment bubble chart (Figure 3E). To validate the key genes identified by RNA-Seq analysis, we chose genes from the inflammation-related signaling pathways such as the IL-17 signaling pathway (differential genes: CCL17 and MUC5B), the cAMP signaling pathway (differential genes: EDN1 and HCAR1), the cytosolic DNA-sensing pathway (differential gene: IL-33), and the cytokine-cytokine receptor interaction (differential genes: IL-33 and CCL17). The 5 genes expression data were showed by the heatmap of RNA-Seq (Figure 3F). The qRT-PCR analysis revealed that the IL-33 gene exhibited the most pronounced differential expression among the three experimental groups of mice. Notably, IL-33 expression in the hippocampus was significantly elevated in the AS-SED group relative to the WT group. Conversely, the AS-EX group demonstrated a marked reduction in IL-33 expression compared to the AS-SED group (Figure 3G). These results align with the expression patterns of the IL-33 gene identified in the RNA-Seq data.

To further explore the potential mechanism of aerobic exercise interventions in reducing IL-33 expression at the cellular level within the hippocampal tissue of AS mice, we employed IL-33 immunofluorescence staining techniques firstly. Our findings revealed a significantly elevated number of IL-33-positive (IL-33⁺) cells in the hippocampal tissue of AS-SED mice compared to the WT group. Conversely, the AS-EX group exhibited a substantial reduction in IL-33⁺ cells relative to the AS-SED group (Figures 4A,B). Moreover, the IL-33 protein also exhibited the same expression trend (Figures 4C,D).These results, when considered alongside previous studies, suggest that aerobic exercise effectively diminishes IL-33 levels in the hippocampal tissue of AS mice.

IL-33 has the capability to trigger the NF-κB signaling cascade (Faas et al., 2021), upregulate numerous pro-inflammatory genes (Mattson, 2005), and is involved in the upregulation of pro-inflammatory cytokines, chemokines, and neurotoxin mediators, such as IL-1β, IL-6, interleukin-8 (IL-8), interferon-gamma (IFN-γ), TNF-α, C-C Motif chemokine ligand 2 (CCL2), glia maturation factor (GMF), nitric oxide (NO), and reactive oxygen species (ROS), which are implicated in the inflammatory processes and cognitive decline characteristic of neurodegenerative diseases (Xiong et al., 2014; Kopitar-Jerala, 2015; Molofsky et al., 2015; Saggu et al., 2016; Ju Hwang et al., 2019). Therefore, WB analyses were performed to assess the expression and phosphorylation levels of key proteins within the NF-κB signaling pathway in the hippocampal tissue of mice across three experimental groups (Figures 4C,D). The findings revealed no statistically significant differences in the protein expression levels of NF-κB and IκBα among the groups. However, relative to the WT group, the AS-SED group exhibited a marked increase in the phosphorylation levels of p-NF-κB/NF-κB and p-IκBα/IκBα, indicating activation of the NF-κB signaling pathway. Conversely, when compared to the AS-SED group, the AS-EX group demonstrated a significant reduction in these phosphorylation levels, suggesting that aerobic exercise intervention effectively attenuates the activation of NF-κB in the hippocampal tissue of AS mice.

Aerobic exercise was found to attenuate the expression of inflammatory markers IL-6, TNF-α, and IL-1β in hippocampal tissue (Figure 5A). The data indicated that, relative to WT mice, the levels of IL-6, TNF-α, and IL-1β were significantly elevated in the hippocampus of AS-SED mice, demonstrating a pronounced upregulation of these inflammatory factors in AS mice. In contrast, when compared to the AS-SED group, the expression of these inflammatory markers was markedly reduced in the hippocampus of AS mice subjected to AS-EX, suggesting that aerobic exercise intervention effectively mitigates the expression of inflammatory factors in the hippocampal tissue of AS mice.

Furthermore, Iba1 is expressed in both resting and activated microglia and is widely recognized as a marker for these cells (Ito et al., 2001). CD68 serves as an optimal marker for identifying M1-type activated microglia (Murphy et al., 2005). Consequently, immunofluorescence staining for CD68 and Iba1 was conducted on the hippocampal regions of mice from the three experimental groups, with the results presented in Figures 5B–D. In comparison to the WT group, the AS-SED mice exhibited a significant increase in the percentage of Iba1-positive (Iba1⁺) and CD68-positive (CD68⁺) microglia among Iba1⁺ microglia in the hippocampal CA1 (cornu amoni 1), CA2, CA3, and dentate gyrus (DG) regions. Conversely, when compared to the AS-SED group, the AS-EX mice demonstrated a significant decrease in the percentage of double-positive microglia for Iba1⁺ and CD68⁺ in the hippocampal CA1, CA3, and DG regions. However, no statistically significant difference was observed in the CA2 region between the two groups.