Aerobic exercise attenuates LPS-induced cognitive dysfunction by reducing oxidative stress, glial activation, and neuroinflammation

Abstract

Physical activity has been considered an important non-medication intervention in preserving mnemonic processes during aging. However, how aerobic exercise promotes such benefits for human health remains unclear. In this study, we aimed to explore the neuroprotective and anti-inflammatory effects of aerobic exercise against lipopolysaccharide (LPS)-induced amnesic C57BL/6J mice and BV-2 microglial cell models. In the in vivo experiment, the aerobic exercise training groups were allowed to run on a motorized treadmill 5 days/week for 4 weeks at a speed of 10 rpm/min, with LPS (0.1 mg/kg) intraperitoneally injected once a week for 4 weeks. We found that aerobic exercise ameliorated memory impairment and cognitive deficits among the amnesic mice. Correspondingly, aerobic exercise significantly increased the protein expressions of FNDC5, which activates target neuroprotective markers BDNF and CREB, and antioxidant markers Nrf2/HO-1, leading to inhibiting microglial-mediated neuroinflammation and reduced the expression of BACE-1 in the hippocampus and cerebral cortex of amnesic mice. We estimated that aerobic exercise inhibited neuroinflammation in part through the action of FNDC5/irisin on microglial cells. Therefore, we explored the anti-inflammatory effects of irisin on LPS-stimulated BV-2 microglial cells. In the in vitro experiment, irisin treatment blocked NF-κB/MAPK/IRF3 signaling activation concomitantly with the significantly lowered levels of the LPS-induced iNOS and COX-2 elevations and promotes the Nrf2/HO-1 expression in the LPS-stimulated BV-2 microglial cells. Together, our findings suggest that aerobic exercise can improve the spatial learning ability and cognitive functions of LPS-treated mice by inhibiting microglia-mediated neuroinflammation through its effect on the expression of BDNF/FNDC5/irisin.

Graphical abstract

Highlights

• •Aerobic exercise training improves spatial learning ability and memory function.
• •Aerobic exercise training induces irisin/FNDC5-mediated upregulation of BDNF/CREB signaling and Nrf2/HO-1 expression in the hippocampus and cerebral cortex.
• •Aerobic exercise training attenuates cognitive dysfunction by reducing BACE-1 in the hippocampus and cerebral cortex.
• •Aerobic exercise training inhibits brain glial activation, neuroinflammation, and neuronal apoptosis in the hippocampus and cerebral cortex.
• •Irisin treatment attenuates LPS-stimulated neuroinflammatory responses in BV-2 microglial cells.

1. Introduction

Neuroinflammation has been crucially implicated in the pathogenesis of neurodegenerative disorders, such as mild cognitive impairment (MCI), Alzheimer's disease (AD), and Parkinson's disease. AD is characterized by the excessive accumulation of extracellular amyloid beta (Aβ) plaques and intracellular neurofibrillary tangles (NFTs) in the brain. Clinically, it manifests as insufficient movement control, language difficulties, memory impairment, personality changes, cognitive dysfunction, and other physical disability that affect daily life and eventually lead to death [1,2]. Approximately 6.7 million people are living with AD. By 2050, this number is projected to double to nearly 13.8 million [3]. However, the pathological mechanisms underlying AD are not completely understood. Among the various neuropathological mechanisms of AD, neuroinflammation is closely involved in the pathogenesis of the disease and is strongly associated with cognitive deficits, oxidative stress, tangle formation, neuronal damage, amyloid deposition, and neuronal death [4,5]. It is well known that neuroinflammation occurs at a very early stage of AD, even before Aβ accumulation and the formation of NFTs. Under the inflammatory condition, microglial activation produces various neurotoxic factors, including proinflammatory cytokines such as interferon-γ (IFN-γ), tumor necrosis factor-α (TNF-α), interleukin-1β (IL-1β), and interleukin-6 (IL-6), as well as inflammatory mediators such as inducible nitric oxide synthase (iNOS) and cyclooxygenase-2 (COX-2). Neuroinflammation and microglial activation can trigger oxidative stress; furthermore, it can lead to increased Aβ plaque and NFTs formation and eventually lead to neuronal damage in the brain of patients with AD [6]. Therefore, the identification of an effective therapeutic strategy that can regulate microglial activation and neuroinflammation-associated molecular targets would offer a promising approach to preventing and treating MCI associated with AD.

Lipopolysaccharide (LPS) triggers neuroinflammation and microglial activation in the brain, impairing cognitive and memory functions. It is extensively used as an amnesic model that induces behavioral and pathological changes in rodents [7]. Studies have demonstrated that intraperitoneal injection of LPS can reduce the expression of nuclear factor erythroid 2-related factor 2 (Nrf2)/heme oxygenase 1 (HO-1); trigger an oxidative stress response by activating NF-κB and MAPK signaling, including c-Jun N-terminal kinase (JNK), p38, and extracellular signal-regulated kinase (ERK); and induce the expression of proinflammatory factors, such as TNF-α, IL-6, IL-1β, iNOS, and COX-2 [8,9]. LPS-induced cognitive impairment is associated with NF-κB- and MAPK-mediated brain-derived neurotrophic factor (BDNF)/cAMP response element-binding protein (CREB) expression in the hippocampus and several cortical regions [10,11]. Further, LPS-induced behavioral and memory impairments are related to neuronal apoptosis indicated by the downregulated expression of antiapoptotic Bcl-2 and elevated expression of pro-apoptotic Bax as well as caspase-3 [12,13]. However, the role of LPS induces memory and cognitive impairments, neuroinflammation, and neuronal apoptosis remains unclear [13].

Physical exercise plays an important role in the prevention of AD by reducing cognitive impairment, Aβ accumulation, and tau phosphorylation, and irisin and BDNF are potential factors contributing to this role [14,15]. Irisin is a novel myokine, has produced by adipose tissue, skeletal muscles, and the cerebral hippocampus [16]. BDNF is an important neurotrophic factor with beneficial effects on brain function in response to exercise [17]. Emerging studies have found that memory function may be improved by the irisin–BDNF axis since irisin could increase BDNF synthesis and further enhance neuroplasticity [18,19]. It has been suggested that acute exercise training immediately and significantly increases the levels of circulating irisin and BDNF [[20], [21]]. Additionally, exercise training significantly enhances the number of BDNF-positive cells in the hippocampus and cerebral cortex, effectively preventing neuronal damage and suppressing excessive inflammatory responses in aging mice [22,23]. Although numerous studies have reported altered BDNF levels in the brain of patients suffering from various brain pathologies, conclusive evidence is still unavailable to determine if these alterations of BDNF levels cause or result from illness onset. Therefore, establishing a causative role of BDNF is crucial in neurodegenerative disorders. However, to date, limited studies have evaluated the effects of aerobic exercise on hippocampal BDNF/FNDC5 expression and its relationship with cognitive function associated with neuroinflammation in LPS-induced amnesic mice. Although presently there are few limitations that are still unexplored (e.g., study type, exercise intensity, and exercise duration etc.). Therefore, the present study aimed to investigate whether, low-intensity aerobic exercise (5 days/week for 4 weeks at a speed of 10 rpm/min) on fibronectin type III domain-containing protein 5 (FNDC5) and BDNF expression in the hippocampus and cerebral cortex and cognitive and memory functions associated with oxidative stress, microglia-mediated neuroinflammation, and neuronal apoptosis in an LPS-treated C57BL/6J mouse model. To the best of our knowledge, this study is the first to reveal a beneficial anti-inflammatory effect of irisin-inhibited LPS-stimulated inflammatory responses in microglial cells.

2. Materials and methods

2.1. Animal handling and treatment schedule

Eight-week-old weight-matched C57BL/6J male mice (N = 24) were supplied by the Daehan Bio-Link, Korea. Animal experiments were performed according to the National Institutes of Health Guide for the Care and Use of Laboratory Animals (Approval number: KU221243). All the animal experiments have been conducted in an isolated and noiseless state and provided a standard environmental condition including food and water ad libitum. The mice were randomly assigned to the following four groups: control group, control intervention; LPS group, LPS injection (0.1 mg/kg); Ex group, aerobic exercise training (10 rpm/min); and LPS + Ex group, LPS injection (0.1 mg/kg) + aerobic exercise training (10 rpm/min). The treatment paradigms and dosage of LPS injection were selected based on previous research to induce memory deficit and neuroinflammation in the AD brain [24]. The LPS group received four intraperitoneal injections of 0.1 mg/kg/b.w. LPS for 4 weeks. The control group was administered with the same volume of saline (sterile 0.9% NaCl). After the behavioral test, the mice were sacrificed, and hippocampal and cortical samples were collected and immediately stored at −80 °C until reverse-transcription (RT)-PCR and Western blot analyses. Fig. 1A shows a summary of the experimental design.

Fig. 1.

Effects of aerobic exercise on LPS-induced spatial learning and cognitive function associated with muscle strength were assessed using the MWM, Y-maze, and muscle strength tests. (A) Shows a summary of the experimental design. (B–D) Total traveled distance and mean escape latency during the probe test. (E) Percentage of spontaneous alterations, (F) total number of arm entries, and (G) muscle strength. Data are expressed as means ± SDs (n = 6 per group). ^###p < 0.001, ^##p < 0.01, and ns—not significant: LPS-treated group vs. control group; *p < 0.05, **p < 0.01, ***p < 0.001, and ns—not significant: LPS + Ex and Ex alone groups vs LPS alone group.

2.2. Aerobic exercise protocol

For the aerobic exercise, the Ex alone and LPS + Ex groups were trained to run on a motorized treadmill apparatus daily from 2:00 p.m. to 4:00 p.m. 5 days/week for 4 week as reported previously (25). All mice were initially trained with the idle treadmill environment. During the experiments, the Ex alone and LPS + Ex groups ran on the treadmill for 1 h at a speed of 10 rpm/min.

2.3. Y-maze test

The Y-maze test is used to assess short-term spatial working memory in rodents. The Y-maze is an experimental apparatus shaped like a Y and consists of a central zone with three identical arms. Each arm has walls that are 3 in high, 3.5 in wide, and 15 in long. Herein, the Y-maze test was conducted according to a previously reported protocol [26]. The percentage of spontaneous alterations and total arm entries was recorded for 5 min. The percentage of alteration was determined as previously described [26].

2.4. Morris water maze (MWM) test

The MWM test is used to evaluate long-term spatial memory in rodents. The MWM apparatus consists of a white circular pool (122 cm in diameter, 20 cm in depth, and 35 cm in height) filled with clear water with a temperature of 25 °C. The MWM tank is painted with a harmless white color to prevent mice from seeing the platform. Mice are carefully released into the water from the edge of the pool, allowing them to swim directly and swiftly to the platform from their initial position. In this study, the test concluded with recording of memory and spatial learning. The data were analyzed using the SMART 3.0 software (Harvard Apparatus, Holliston, MA, USA) [27].

2.5. Grip strength test

The maximal grip strength was measured using a grip strength meter (Bioseb, France). The animals were put in a metal grid, and their tail was caught and pulled back. The force exerted on the grid by the animals’ legs was measured in grams. Three tests were performed, and the mean of the three measurements was calculated.

2.6. Cell culture and treatment

BV-2 microglial cells were cultured in DMEM containing 5% FBS and 1% penicillin/streptomycin solution. BV-2 microglial cells were maintained in a controlled atmosphere with 5% CO₂ at 37 °C. The cells were then treated with various concentrations of irisin (12.5, 25, and 50 nM) for 12 h before exposure to 200 ng/mL LPS.

2.7. Cell viability assay

To assess cell viability, we cultured the BV-2 microglial cells in 24-well plates (6 × 10⁴ cells per well) and treated them with various concentrations of irisin (12.5, 25, and 50 nM) with or without 200 ng/mL LPS. After 24 h, DMEM was removed and 0.5 mg/mL MTT was added, followed by 2 h of incubation at 37 °C. Thereafter, 400 μL DMSO was supplied to the cells. The absorbance was measured at 540 nm using a microplate reader.

2.8. Immunofluorescence study

The BV-2 microglial cells were seeded into 24-well plates and incubated overnight. After washing with pre-warmed PBS, cells were pretreated with 50 nM irisin for 12 h, stimulated with 200 ng/mL LPS for 30 min, fixed in 4% paraformaldehyde, permeabilized with acetone, and blocked with 1% of BSA in PBS. Anti-p–NF–κB primary antibody were then used prior to incubating cells samples at 4 °C overnight. They were washed twice with PBS and incubated with species-specific fluorescent conjugated secondary antibody (A32731 Invitrogen) for 1 h at room temperature. After washed twice with PBS, the cells were then counterstained with DAPI (2 μg/mL). Images were captured by fluorescence microscope and perform further image analysis by NIS-Elements software (BR-2.01.00, NY 11747–3064, NY, USA) [26].

2.9. RT-PCR analysis

RT-PCR was conducted as previously reported [24]. The specific primer sequences used in this study are listed in Supplementary Table 1. In brief, the total RNA samples were isolated from hippocampus and cerebral cortex tissues by using TRIzol reagent (Invitrogen) following the manufacturer's protocols. An equal amount of 2.5 μg of RNA samples was reverse transcribed into cDNA using a First-Strand cDNA Synthesis kit (Invitrogen). The synthesized cDNA was further used as a template to perform PCR for respective targets; the PCR products were separated on 1% agarose gels and images were captured by the Davinch-Gel™ gel imaging system. The relative mRNA expression levels of GCN5, IL-1β, IL-10, and IFN-γ were normalized to the expression of GADPH.

2.10. Western blot analysis

Hippocampal and cerebral cortical tissues from each mouse were isolated and added to 200 μL PRO-PREP lysis buffer (#17081, iNtRON, NJ, USA) for homogenization and centrifugation. The extract of the BV-2 microglial cells was obtained using PRO-PREP lysis buffer. The concentration of the protein sample was measured using the DC protein assay kit according to the manufacturer's instructions (Bio-Rad). Approximately 10 μL of the protein supernatant containing 20 μg total protein was separated on an 8–10% SDS-polyacrylamide gel, the transferred to polyvinylidene difluoride (PVDF) membranes. For immunodetection, the blots were blocked with 3% skim milk in TBS-T for 1 h, then the membrane was incubated with primary antibodies at 4 °C followed by incubation with the secondary antibodies (anti-mouse, anti-rabbit, and anti-goat), for an appropriate amount of time. The primary antibodies used in this study are listed in Supplementary Table 2. Antibodies were diluted as per the manufacturer's instructions. After extensive washing, the protein band was detected using a detection system, and the relative intensity was quantified using the ImageJ software [26,27].

2.11. Statistical analysis

All in vivo and in vitro experiments were independently conducted at least thrice, and all data were expressed as means ± SD. GraphPad Prism 9.0.0 was used for the analyses. Statistical comparison between each group for the in vitro (control, LPS, LPS + irisin 12.5 + irisin 25, and irisin 50 nM) and in vivo experiments (control, LPS, Ex, and LPS + Ex) was conducted using one-way ANOVA analysis. Differences between experimental groups were analyzed by the Tukey multiple comparison test. A value of p < 0.05 was considered statistically significant. ^# (p < 0.05), ^## (p < 0.01), and ^### (p < 0.001) and * (p < 0.05), ** (p < 0.01), and *** (p < 0.001) indicated significant differences between the indicated groups.

3. Results

3.1. Effects of aerobic exercise on LPS-induced behavioral abnormalities

Memory impairment and cognitive deficits are considered a prime candidate mechanism for linking MCI associated with AD progression. To investigate the effects of aerobic exercise, we evaluated spatial learning and cognitive function using the MWM and Y-maze tests and muscle strength using the grip strength test. The treatment summary of the present experiment is illustrated in Fig. 1A. The movement of all mice was analyzed statistically, demonstrating that the total traveled distance and escape latency in the LPS group was increased before locating the platform compared with the control group (p < 0.001) (Fig. 1B), suggesting that LPS treatment diminished spatial learning ability and memory function. In contrast, the LPS + Ex group demonstrated decreased total traveled distance (p < 0.001) and escape latency (p < 0.001) compared with the LPS group (Fig. 1C and D). No changes were observed between the Ex group and the control group, demonstrating that Ex alone can improve cognition to some extent, and aerobic exercise had better performance. In the Y-maze test, the LPS group showed a significantly decreased percentage of spontaneous alteration compared with the control group (p < 0.01). Conversely, the LPS + Ex group exhibited a significantly elevated percentage of spontaneous alteration compared with the LPS group (Fig. 1E). However, significant differences were not observed in the total number of arm entries between the groups (Fig. 1F). Furthermore, the LPS group demonstrated a significantly decreased grip strength compared with the control group (p < 0.01), and this effect was attenuated in the LPS + Ex group compared with that in the LPS group (p < 0.01) (Fig. 1G). These findings confirm that aerobic exercise (5 days/week for 4 weeks at a speed of 10 rpm/min) can improve the spatial learning and cognitive abilities of mice with LPS-induced MCI associated with AD.

3.2. Expression of BDNF/FNDC5/CREB and nrf2/HO-1 in the hippocampus and cerebral cortex

BDNF overexpression could contribute to facilitating neurotrophic processes and is involved in cognitive and memory functions. However, BDNF upregulation could promote axonal and dendritic growth, synaptic plasticity, and neuronal survival and has been discovered as a potential primary candidate for mediating the efficacy of anti-Alzheimer's therapy [14]. Herein, we investigated the effects of aerobic exercise on the protein expression of FNDC5/BDNF/CREB in the hippocampus and cerebral cortex of the LPS-induced amnesic mice. The LPS group showed a significantly decreased expression of BDNF in the hippocampus (p < 0.001) and cerebral cortex (p < 0.01) compared with the control group (Fig. 2A and B) but not FNDC5. The LPS + Ex group demonstrated a significantly increased expression of FNDC5 and BDNF in the hippocampus (p < 0.001, p < 0.05) and cerebral cortex (p < 0.05) compared with the LPS group (Fig. 2C–F). The expression of p-CREB in the hippocampus (p < 0.01) significantly decreased in the LPS group compared with that in the control group (Fig. 2A) but not and cerebral cortex (Fig. 2B). Conversely, the expression of p-CREB in the hippocampus (p < 0.01) and cerebral cortex (p < 0.05) significantly increased in the LPS + Ex group compared with that in the LPS group (Fig. 2G and H). The expressions of FNDC5, BDNF, and p-CREB in the hippocampus (p < 0.001, p < 0.01, and p < 0.01, respectively) and cerebral cortex (p < 0.01, p < 0.01, and p < 0.01, respectively) were higher in the Ex group than in the LPS and LPS + Ex groups. Noticeably, in the hippocampus (p < 0.001) and cerebral cortex (p < 0.01) of the Ex group compared to the control group, FNDC5 expression was significantly higher (Fig. 2C and D). It is plausible that the aerobic exercise selectively influenced FNDC5 expression in specific brain regions, such as hippocampus and cerebral cortex. This regional specificity demonstrates the relevance and effectiveness of the aerobic exercise. Moreover, consistent with an earlier report [28], the expressions of FNDC5/irisin and BDNF in the brain were higher in the aerobic exercise group than in the LPS group. These findings suggest that aerobic exercise (5 days/week for 4 weeks at a speed of 10 rpm/min) plays a neuroprotective role against LPS-induced neurotoxicity and effectively attenuates LPS-induced spatial learning and memory impairments by promoting the BDNF/FNDC5/CREB pathway.

Fig. 2.

(A, B) Effects of aerobic exercise on the protein expression of BDNF/FNDC5/p-CREB and Nrf2/HO-1 in the hippocampus and cerebral cortex of the LPS group. Representative Western blot bands and quantitative analysis of (C, D) FNDC5, (E, F) BDNF, (G, H) p-CREB, (I, J) Nrf2, and (K, L) HO-1 (n = 3 per group). Data are presented as means ± SDs. ^###p < 0.001, ^##p < 0.01, ^#p < 0.01 and ns—not significant: LPS-treated group vs. control group; *p < 0.05, **p < 0.01, and ***p < 0.001: LPS + Ex and Ex alone groups vs. LPS alone group. *p < 0.05, **p < 0.01, ***p < 0.001, and ns—not significant: Ex alone group vs. non-treated control group. (A, B) FNDC5; hippocampus and cerebral cortex [F(3, 8) = 151.0 and 13.91], BDNF; hippocampus and cerebral cortex F(3, 8) = 18.66 and 2.412], p-CREB; hippocampus and cerebral cortex F(3, 8) = 10.19 and 4.054], Nrf2; hippocampus and cerebral cortex [F(3, 8) = 5.579 and 8.733], HO-1; hippocampus and cerebral cortex [F(3, 8) = 12.63 and 27.92].

The activation of Nrf2 protected the hippocampal neurons from oxidative injury. BDNF may protect neurons from ferroptosis-like cell death by induction of Nrf2 signaling, in addition to its involvement in antioxidant defense [29]. In this study, we evaluated the antioxidant effects of aerobic exercise on the protein expression of Nrf2/HO-1 in LPS-induced amnesic mice. According to the Western blotting results, the expressions of Nrf2 and HO-1 in the hippocampus (p < 0.01 and p < 0.05) and cerebral cortex (p < 0.01) were significantly downregulated in the LPS group compared to control the group (Fig. 2A and B). Compared to the LPS group, Nrf2 and HO-1 protein expression of the LPS + Ex group were significantly upregulated in the hippocampus (p < 0.05 and p < 0.01) and cerebral cortex (p < 0.01 and p < 0.01) (Fig. 2I–L). Further, the expressions of Nrf2 and HO-1 in the hippocampus (p < 0.001 and p < 0.01) and cerebral cortex (p < 0.001 and p < 0.01) in the Ex group were markedly higher than those in the LPS and LPS + Ex groups. These findings demonstrate that the regulatory effects of aerobic exercise on the antioxidant defense system are mediated by BDNF-induced upregulation of NRF2/HO-1 signaling.

3.3. Expression of proinflammatory mediators in the hippocampus and cerebral cortex

Earlier reports have demonstrated that microglia-mediated inflammatory responses are another critical pathological feature of AD. Therefore, the alteration of microglia-mediated neuroinflammation has become an important therapeutic target for MCI associated with AD [5,6]. In this study, we analyzed the mRNA expression of inflammatory biomarkers such as iNOS and COX-2 in the hippocampus and cerebral cortex of LPS-induced amnesic mice. Representative RT-PCR findings showed that the mRNA expressions of iNOS and COX-2 in the hippocampus (p < 0.001) and cerebral cortex (p < 0.001) were significantly higher in the LPS group than in the control group (Fig. 3A and B). Conversely, the LPS + Ex group showed significantly reduced mRNA expressions of iNOS and COX-2 in the hippocampus (p < 0.05 and p < 0.05) and cerebral cortex (p < 0.05 and p < 0.05) compared with the LPS group (Fig. 3C–F). We also investigated the mRNA expressions of proinflammatory cytokines such as IL-1β, IL-10, and IFN-γ in the hippocampus and cerebral cortex of LPS-induced amnesic mice. We found that the mRNA expressions of IL-1β, IL-10, and IFN-γ in the hippocampus (p < 0.001) and cerebral cortex (p < 0.001) markedly increased in the LPS group compared with those in the control group (Fig. 3A and B). In contrast, the LPS + Ex group exhibited significantly decreased mRNA expressions of IL-1β, IL-10, and IFN-γ in the hippocampus (p < 0.01, p < 0.05, and p < 0.01, respectively) and cerebral cortex (p < 0.01, p < 0.01, and p < 0.01, respectively) compared with the control group (Fig. 3G–L). The expressions of the inflammatory biomarkers in the hippocampus and cerebral cortex were lower in the Ex group than in the LPS and LPS + Ex groups. These findings suggest that aerobic exercise suppresses microglia-mediated inflammatory responses in LPS-induced amnesic models.

Fig. 3.

(A, B) Effects of aerobic exercise on the mRNA expression of proinflammatory mediators in the hippocampus and cerebral cortex of the LPS group. Representative RT-PCR bands and quantitative analysis of (C, D) iNOS, (E, F) COX-2, (G, H) IL-1β, (I, J) IL-10, and (K, L) IFN-γ (n = 3 per group). Data are presented as means ± SDs. ^###p < 0.001: LPS-treated group vs. control group; *p < 0.05, **p < 0.01, ***p < 0.001: LPS + Ex and Ex alone groups vs. LPS alone group. (A, B) iNOS; hippocampus and cerebral cortex [F(3, 8) = 52.11 and 16.33], COX-2; hippocampus and cerebral cortex [F(3, 8) = 36.26 and 275.1], IL-1β; hippocampus and cerebral cortex [F(3, 8) = 61.93 and 117.0], IL-10; hippocampus and cerebral cortex [F(3, 8) = 24.90 and 72.33], IFN-γ; hippocampus and cerebral cortex [F(3, 8) = 4.352 and 81.45].

3.4. Expression of BACE-1, iba-1 and GCN5 in the hippocampus and cerebral cortex

BDNF has been reported to reduce the expression of BACE-1, which has been associated with Aβ accumulation and plaque formation, leading to cognitive deficits [25]. In the study, Western blot analysis showed that the protein expression of BACE-1 in the hippocampus (p < 0.001) and cerebral cortex (p < 0.01) was significantly elevated in the LPS group (Fig. 4A and B). In contrast, the protein expression of BACE-1 in the hippocampus (p < 0.01) and cerebral cortex (p < 0.001) did not increase in the LPS + Ex group but was lower in the Ex group than in the LPS and LPS + Ex groups (Fig. 4C and D). These findings strongly demonstrate that aerobic exercise might inhibit the protein expression of BACE-1 by increasing the expression of BDNF and therefore prevent amyloidogenic pathway activity and cognitive impairment by enhancing the expression of irisin/FNDC5.

Fig. 4.

(A–J) Effects of aerobic exercise on BACE-1 and microglial activation in the hippocampus and cerebral cortex of the LPS group. Representative Western blot and RT-PCR bands and quantitative analysis of (C, D) BACE-1, (G, H) iba-1, and (K, L) GCN5 (n = 3 per group). Data are presented as means ± SDs. ^###p < 0.001 and ^##p < 0.01: LPS-treated group vs. control group; **p < 0.01 and ***p < 0.001: LPS + Ex and Ex alone groups vs. LPS alone group. (A, B) BACE-1; hippocampus and cerebral cortex [F(3, 8) = 18.44 and 61.72], (E, F) iba-1; hippocampus and cerebral cortex [F(3, 8) = 37.42 and 347.6], (I, J) GCN5; hippocampus and cerebral cortex [F(3, 8) = 19.13 and 31.16].

Microglia plays a prominent role in neuroinflammatory processes in the brain. Therefore, regulating microglial activation would be an effective therapeutic target for neuroinflammation-associated cognitive disorders [24]. To further examine the anti-inflammatory effects of aerobic exercise among the amnesic mice, we investigated the protein and mRNA expressions of iba-1 and GCN5 in the hippocampus and cerebral cortex. Western blot and RT-PCR analyses showed that the LPS group demonstrated increased expressions of iba-1 and GCN5 in the hippocampus (p < 0.001 and p < 0.001) and cerebral cortex (p < 0.001 and p < 0.01) compared with the control group (Fig. 4E and F and 4I and J). In contrast, the LPS + Ex group exhibited significantly suppressed expressions of iba-1 and GCN5 in the hippocampus (p < 0.001 and p < 0.001) and cerebral cortex (p < 0.001 and p < 0.001) compared with the LPS group (Fig. 4G and H and 4K and L). However, no significant differences were observed in the expression of iba-1 and GCN5 in the hippocampus and cerebral cortex between the control and Ex groups. These findings suggest that aerobic exercise might alleviate LPS-induced microglial activation and its related gene in the hippocampus and cerebral cortex.

3.5. Expression of apoptotic markers in the hippocampus and cerebral cortex

We investigated the effects of aerobic exercise on LPS-induced proapoptotic and antiapoptotic protein expressions. The protein expression study showed that proapoptotic and antiapoptotic markers of Bax, caspase-3, and PARP-1 in the hippocampus (p < 0.05 and p < 0.001) and cerebral cortex (p < 0.01, p < 0.05, and p < 0.001, respectively) significantly increased, while that of Bcl-2 (p < 0.05) decreased in the LPS group (Fig. 5A and B). Conversely, the LPS + Ex group showed markedly reduced the protein expressions of Bax, caspase-3, and PARP-1 in the hippocampus (p < 0.05, p < 0.01, and p < 0.01, respectively) and cerebral cortex (p < 0.05, p < 0.05, and p < 0.001, respectively) and upregulated the protein expression of Bcl-2 in the hippocampus (p < 0.05) and cerebral cortex (p < 0.01) compared with the LPS group (Fig. 5C–J). The Ex group exhibited effectively modulated proapoptotic and antiapoptotic protein expressions in the hippocampus and cerebral cortex compared with the LPS and LPS + Ex groups. Compared to the control group, the Ex alone group exhibited much higher the levels of Bcl-2 protein expressions in the hippocampus and cerebral cortex, suggesting that aerobic exercise is a potent antiapoptotic factor that improves neurotrophic factors and cognitive functions. These findings indicate that aerobic exercise has a protective effect against LPS-induced neuronal apoptosis.

Fig. 5.

(A, B) Effects of aerobic exercise on the expressions of proapoptotic and antiapoptotic proteins in the hippocampus and cerebral cortex of the LPS group. Representative Western blot bands and quantitative analysis of (C, D) Bcl-2, (E, F) Bax, (G, H) caspase-3, and (I, J) PARP-1 (n = 3 per group). Data are presented as means ± SDs. ^###p < 0.001, ^##p < 0.01, ^#p < 0.05, and ns—not significant: LPS-treated group vs. control group; *p < 0.05, **p < 0.01, and ***p < 0.001: LPS + Ex and Ex alone groups vs. LPS alone group. (A, B) Bcl2; hippocampus and cerebral cortex [F(3, 8) = 16.80 and 11.52], Bax; hippocampus and cerebral cortex [F(3, 8) = 5.289 and 11.61], caspase-3; hippocampus and cerebral cortex [F(3, 8) = 10.9 and 18.49], PARP-1; hippocampus and cerebral cortex [F(3, 8) = 16.66 and 24.94].

3.6. Effects of irisin on the cell viability, antioxidant, inflammatory responses, and NF-κB/MAPK/IRF3 signaling in the LPS-stimulated BV-2 microglial cells

Since irisin is an exercise-induced myokine, our data confirm that aerobic exercise improves cognitive function and reduces neuroinflammatory responses by activating the protein expression of FNDC5 and BDNF in the hippocampus and cerebral cortex of amnesic mice. Next, we investigated the anti-inflammatory effects of irisin on the LPS-stimulated BV-2 microglial cells. The irisin pretreatment (12.5, 25, and 50 nM) and LPS (200 ng/mL) for 36 h did not significantly affect the viability of the BV-2 microglial cells in comparison with the control intervention (Fig. 6A). Considering the cell viability observed, we chose the anti-inflammatory concentration of irisin to evaluate the effects on the protein expression of antioxidant biomarkers and inflammatory mediators. Western blot analysis showed that LPS treatment synergistically decreased the protein expressions of Nrf2 and HO-1 in the BV-2 microglial cells compared with the control intervention (Fig. 6B). Conversely, irisin pretreatment (12.5, 25, and 50 nM) significantly upregulated the protein expressions of Nrf2 and HO-1 in the BV-2 microglial cells compared with LPS treatment (Fig. 6C). We further analyzed changes in inflammatory mediators expression in LPS-stimulated BV-2 microglial cells. The LPS treatment significantly increased the protein expressions of iNOS and COX-2 in the BV-2 microglial cells in comparison with the control intervention (Fig. 6D). The irisin pretreatment (12.5, 25, and 50 nM) significantly reduced the protein expressions of iNOS and COX-2 in a concentration-dependent manner (Fig. 6E and F). These findings indicate that irisin can enhance the antioxidant effect in LPS-stimulated BV-2 microglial cells through the upregulation of the Nrf2/HO-1 pathway.

Fig. 6.

Effects of irisin on the cell viability, antioxidant biomarkers, and inflammatory responses in BV-2 microglial cells induced by LPS. (A) The cells were pretreated with irisin (12.5, 25, and 50 nM) for 12 h, followed by LPS (200 ng/L) stimulation for 24 h. (B) The antioxidant protein expression of Nrf2 and HO-1 was estimated using Western blot analysis, and the bands were measured using the ImageJ software to quantify the expression of (C) Nrf2 and HO-1. (D) The inflammatory protein expression of iNOS and COX-2 was estimated using Western blot analysis, and the bands were measured using the ImageJ software to quantify the expression of (E) iNOS and (F) COX-2. (G) The protein expression of phosphorylated NF-κB, IκB-α, and IRF3 was estimated using Western blot analysis, and the bands were measured using the ImageJ software to quantify the expression of p–NF–κB, p-IκB-α, and p-IRF3 (H). (I) The immunofluorescence staining analysis of p65 nuclear translocation in BV-2 microglial cells. Nuclei were stained with DAPI. (J) The protein expression of phosphorylated p38, ERK, and JNK was estimated using Western blot analysis, and the bands were measured using the ImageJ software to quantify the expression of (K) p-p38, (L) p-ERK, and (M) p-JNK. All experiments were independently conducted at least three times. Data are presented as means ± SDs. ^###p < 0.001 and ns—not significant: LPS-treated group vs. control group; *p < 0.05, **p < 0.01, ***p < 0.001 and ns—not significant: LPS + Ex and Ex alone groups vs. LPS alone group. (B, D) Nrf2; [F(4,10) = 10.94], HO-1; [F(4,10) = 11.62], iNOS; [F(4,10) = 50.56], COX-2; [F(4,10) = 8.415], (G, J) p–NF–κB; [F(4,10) = 14.96], p-IκB-α; [F(4,10) = 46.59], p-IRF3; [F(4,10) = 18.97], p-p38; [F(4,10) = 10.36], p-Erk; [F(4,10) = 8.536], p-JNK; [F(4,10) = 48.26].

We next explored the effects of irisin on the expressions of NF-κB, IκB-α, and IRF3 in the LPS-stimulated BV-2 microglial cells. Western blot analysis showed that the LPS treatment significantly increased the expressions of NF-κB, IκB-α, and IRF3 in the BV-2 microglial cells in comparison with the control intervention (Fig. 6G). In contrast, the irisin pretreatment significantly reduced the expressions of NF-κB, IκB-α, and IRF3 in a concentration-dependent manner (Fig. 6H). We also confirmed the intracellular localization of p65 in the LPS-stimulated BV-2 microglial cells. Immunofluorescence study showed that the LPS-stimulated cells translocated to the nucleus compared with the control cells. The irisin pretreatment (50 nM) remarkably blocked the translocation of p65 in the BV-2 microglial cells compared with the LPS treatment (Fig. 6I). Since irisin inhibited the inflammatory responses and NF-κB and IκB-α signaling in the LPS-stimulated BV-2 microglial cells, we analyzed the phosphorylation of its downstream signaling pathways. Western blot analysis showed that LPS treatment significantly increased the phosphorylation of p38, ERK, and JNK in the BV-2 microglial cells compared with the control intervention (Fig. 6J). Conversely, irisin pretreatment significantly decreased the phosphorylation of p38, ERK, and JNK in a concentration-dependent manner (Fig. 6K–M). These findings suggest that irisin-induced inhibition of LPS-stimulated inflammatory responses regulates microglial activation by inhibiting the NF-κB, MAPK and significantly suppressing IRF3 signaling pathway.

4. Discussion

Mounting evidence suggests that pharmacological drug supplements can improve memory and cognitive functions; however, their therapeutic efficacy are temporary and limited, and there may be unforeseen adverse effects. It is believed that physical exercise has a notable beneficial effect on brain health and memory function with fewer side effects. In this study, we evaluated the effects of aerobic exercise on the spatial learning ability, memory function, and brain neuroinflammation among amnesic AD mice and the underlying mechanisms of irisin-mediated anti-inflammatory effects in BV-2 microglial cells stimulated by LPS. The main findings were as follows: (1) 4-week aerobic exercise improves the spatial learning ability and memory function of amnesic mice; (2) aerobic exercise induces irisin/FNDC5-mediated upregulation of BDNF/CREB signaling and Nrf2/HO-1 expression in the hippocampus and cerebral cortex of amnesic mice; (3) aerobic exercise has a marked anti-inflammatory effect on the brain of amnesic mice: reducing the expression of proinflammatory cytokines (i.e., IL-1β and IFN-γ) and inflammatory mediators (i.e., iNOS and COX-2), suppressing neuronal apoptosis, and increasing the expression of antiapoptotic markers (i.e., Bcl-2); (4) aerobic exercise inhibits brain glial activation in amnesic mice probably by increasing the protein expressions of BDNF and FNDC5 in the brain of amnesic mice; and (5), irisin treatment attenuates neuroinflammatory responses and activation of NF-κB/MAPK/IRF3 signaling in BV-2 microglial cells induced by LPS.

In an animal study, researchers discovered that LPS-induced neuroinflammation and memory impairment is an important tool for understanding the mechanisms involved in AD pathology and identifying potential therapeutic options. Intraperitoneal LPS injection is one of the most popular methods used to establish animal experiments [30]. Recent research has reported that intraperitoneal injection of LPS (0.1–0.5 mg/kg/b.w.) induces spatial learning and memory dysfunctions and changes in memory-related neuroprotective biomarkers [14,31]. In the present study, LPS (0.1 mg/kg) treatment lowered the expression level of BDNF in the hippocampus and cerebral cortex and the related behavioral experiments have shown that typical characteristics of memory decline in a variety of memory, indicating spatial learning and cognitive impairment. Patients with AD who participate in long-term physical exercise could enhance spatial learning ability and memory function [32,33]. Our research provides evidence of the beneficial effects of physical exercise on spatial learning ability and memory function and the underlying mechanisms in amnesic mice. In the MWM test, the LPS group showed a longer latency and traveled distance, indicating loss of spatial memory, while the LPS + Ex group showed a significantly shorter latency and total traveled distance than the LPS group. These findings indicate that the 4-week aerobic exercise training significantly improves the spatial learning and cognitive abilities of AD mice. LPS injection also affected the hippocampal working memory and learning ability by suppressing the percentage of spontaneous alteration and yielded a similar number of total arm entries. In contrast, the LPS + Ex group showed an increased percentage of spontaneous alteration, indicating improved spatial working memory and cognitive function. The LPS group showed less grip strength, and this effect was attenuated in the LPS + Ex group. Similarly, Yuan et al. showed that long-term aerobic exercise produces beneficial effects on the spatial learning memory and cognitive function of LPS-induced APP/PS1 mice [34]. An increase in hippocampal BDNF expression and improvement in cognitive function during physical exercise has also been reported in the Y-maze and MWM tests in transgenic AD rats [35]. However, the cross-sectional and longitudinal study confirmed that handgrip strength is associated with special learning and cognitive function [5]. LPS-treated amnesic mice have learning and memory deficits. We found that treadmill and voluntary exercise training for 5 or 9 weeks significantly improved brain and cognitive functions by enhancing neurogenesis, synaptic plasticity, gene expression, and hippocampal neuronal density and attenuating neuroinflammation, consistent with numerous current reports [[36], [37], [38]]. Therefore, aerobic exercise can be beneficial in preventing the decline and improving the memory performance of individuals with MCI associated with AD.

BDNF is closely implicated in the exercise-induced enhancing effect on brain function, but muscle activity may also contribute to inducing the expression of BDNF in the brain by increasing the expression of circulating factors such as irisin/FNDC5 [39]. Enhanced BDNF expression in the hippocampus and cortex is known to improve spatial learning and cognitive function; increase synaptic plasticity, synaptic terminal density, and axonal and dendritic growth; and contribute to neuronal survival and differentiation [40,41]. Previous research has demonstrated that the hippocampal and cortical protein expressions of BDNF, FNDC5, and p-CREB decrease in aged and LPS-injected mice [26,37,42]. In this study, LPS injection decreased the expressions of BDNF, FNDC5, and p-CREB in the hippocampus and cerebral cortex of the amnesic mice, while LPS injection with aerobic exercise training increased the expressions of these markers. These findings indicate that 4-week aerobic exercise training significantly increases hippocampal and cortical expressions of BDNF, FNDC5, and p-CREB in AD mice. This is similar to previous findings that aerobic and voluntary exercise training increased the expressions of BDNF, FNDC5, and p-CREB in the hippocampus of AD animals, accompanied by improvement of spatial memory learning and synaptic function [[36], [37], [38]]. Another study reported that FNDC5 has been shown to regulate neuronal survival, neural development, and differentiation in embryonic stem cells [43]. Belviranli et al. showed that endurance training improves cognitive performance and increases the expressions of BDNF and irisin/FNDC5 in trained athletes and indicated a possible link between physical exercise-induced cognitive performance and the expressions of BDNF and irisin [44]. In the present study, we also observed that the hippocampal and cortical protein expressions of Nrf2 and HO-1 were downregulated in the LPS group and that 4-week aerobic exercise increased the protein expressions in the LPS + Ex group. Recent research has also shown that LPS-induced oxidative stress is associated with a decreased expression of Nrf2 and HO-1 and that exercise training upregulates the expressions of Nrf2 and HO-1 and prevents LPS-induced oxidative damage in sepsis mice [45]. Ryu et al. reported that aerobic exercise elevated the expressions of Nrf2 and HO-1 in mouse models of AD. In addition, cognitive and memory functions improved based on the Y-maze and MWM test findings [46].

Neuroinflammation plays a crucial role in the pathogenesis of AD. Microglia-mediated neuroinflammation can lead to increased Aβ accumulation and tau phosphorylation, consequently activating microglia and astrocytes. Activated microglia then produce various proinflammatory mediators that induce changes in the hippocampus involved in spatial learning and memory [47]. In animal experiments, sustained expressions of IFN-γ and IL-1β in the hippocampus can activate microglia and astrocytes to trigger a vigorous inflammatory response, ultimately leading to neuronal cell death and hippocampus-mediated memory impairment [48]. Some studies have also demonstrated that prolonged expressions of iNOS and COX-2, along with increased production of NO and ROS, may contribute to spatial working memory impairments and cognitive deficits in AD mice [35,49,50]. Moreover, increased expressions of iNOS, COX-2, IL-1β, and IFN-γ in the brain tissues have been reported in animal models of LPS-induced neuroinflammation [35,51]. Previous study has reported that LPS treatment not only to induction of IL-1β, IFN-γ, iNOS, and COX-2 expression but also response to stimulates an early induction of anti-inflammatory cytokines such as IL-10 expression [24], portentous this may be a mechanism to counterbalance the inflammatory response. In the present study, LPS treatment significantly increased the mRNA expression of proinflammatory mediators iNOS, COX-2, IL-1β, IL-10, and IFN-γ in the hippocampus and cerebral cortex of the amnesic mice, but 4-week aerobic exercise training reduced the mRNA expression of IFN-γ, IL-10, IL-1β, COX-2, and iNOS, and helped recover memory deficits. Consistent with the earlier literature, we found that LPS injection increased significantly in the hippocampal expression level of iNOS, COX-2, IL-1β, and IFN-γ, which was attenuated by treadmill exercise training [24,51]. Subsequently, to further evaluate neuroinflammation, we examined microglial activation by quantifying the protein and mRNA expression of iba-1 and GCN5. LPS treatment increased the protein expression of iba-1, indicating microglial activation induced by LPS as reported previously [52]. Consistent with the report by Mota and Kelly [24], the present findings show that the neuroprotective and anti-inflammatory effects of aerobic exercise modulate neuroinflammation by reducing the protein and mRNA expression of iba-1 and GCN5 microglial activation, thus protecting against LPS-induced deficits in spatial learning and memory. It has been reported that irisin/FNDC5 attenuates brain neuroinflammation and microglial activation and improves hippocampal learning and memory by increasing the expression of BDNF. BDNF plays a vital role in multiple aspects of anti-inflammatory effects by reducing the release of IFN-γ and IL-1β and attenuating the iNOS and COX-2 expression in the brain [53].

Increases in the expressions of AD-related neurotoxic proteins (Aβ protein peptide and protein tau as well as extracellular Aβ plaques) may cause neuroinflammation. Such inflammation could trigger the production of APP, BACE-1, and Aβ in the hippocampus of AD rats [25]. Zameer et al. showed that injection of streptozotocin in rats impaired their cognitive abilities and increased the deposition of Aβ by elevating the enzymatic expression of BACE-1 and significantly augmenting neuroinflammation with oxidative stress [54]. In the present study, we also observed that LPS treatment increased the hippocampal and cortical protein expressions of BACE-1 and that 4-week aerobic exercise training decreased the expressions of BACE-1 in the AD mice. Previous research has shown that aerobic exercise alleviates brain neuroinflammation-induced memory impairment by lowering the expression of BACE-1 and elevation in the expression of BDNF [25]. Therefore, the expression of BACE-1 is strongly influenced by that of BDNF. In Aβ-treated animals, the expression of BDNF is significantly reduced, while that of BACE-1 is increased. In contrast, exercise training has been shown to markedly elevate the expressions of BDNF in both healthy and Aβ-induced animals [55,56], which supports the present findings. Apoptosis plays an important role in the progression of AD. According to the findings of previous studies, neuronal apoptosis is induced by oxidative stress, neuroinflammatory response, and apoptosis [57]. Caspase-3 is one of the most widely studied caspases, and it is a key executor of apoptosis. Activation of caspase-3 is involved in neuronal apoptosis and regenerative failure in AD brain [58]. Bcl-2 is a survival factor that can block both necrotic and apoptotic cell death. Bcl-2 acts upstream to prevent caspase-3 activation, inhibits free radical formation, and blocks the proapoptotic actions of other members of the Bcl-2 family such as Bax [59]. Decreased protein expression of Bcl-2 and increased protein expression of Bax were observed in ischemic rats [60]. The present study showed that the expression of Bax, caspase-3, and PARP-1 protein was increased, while the expression of Bcl-2 protein was decreased following LPS injection. These results indicate that LPS-induced apoptotic neuronal cell death in the hippocampus and cerebral cortex. Tong et al. [61] reported that wheel running exercise for 3 weeks reduced the expression of mRNAs of apoptosis-associated genes, such as the Bcl-2 gene family, and neuronal death. Likewise, both 4 weeks and 6 weeks of exercise training increased the expression of Bcl-2, decreased the expression of Bax, and inhibited apoptosis in both AD mice and rats [62,63]. In this study, we confirmed that 4-week aerobic exercise can enhance learning memory and attenuate hippocampal and cortical neuronal apoptosis in AD mice by increasing the expression of Bcl-2 and reducing that of Bax, caspase 3, and PARP-1. These results suggest that aerobic exercise training not only has the neuroprotective effects, but also may have other beneficial anti-apoptotic effects by inhibiting neuronal apoptosis and improving cognitive function.

To further examine the mechanism underlying the antioxidant and anti-inflammatory effects of irisin on the LPS-exposed BV-2 microglial cells. Accordingly, we investigated the therapeutic potential of irisin on the Nrf2/HO-1 and NF-κB/MAPK/IRF3 signaling pathways. Microglia-mediated neuroinflammation is initiated by a variety of pathogenic stimuli such as viruses, bacteria, and LPSs. LPS can activate the NF-κB and MAPK signaling in microglial cells increases the production of proinflammatory cytokines and the expression of inflammatory mediators, such as TNF-α, IL-6, iNOS, and COX-2 [27,64]. IRF3 is a transcription factor that regulates the innate immune response and activation of inflammation and other pathological conditions [65]. Many studies have demonstrated that Nrf2/HO-1 signaling regulates anti-oxidation and anti-inflammation by directly driving the clearance of excessive ROS and indirectly inhibiting the production of cytokines [66]. Therefore, Nrf2/HO-1 and NF-κB/MAPK/IRF3 signaling is considered a crucial element in neuroinflammatory processes and may act as an effective target for anti-neuroinflammatory therapy. Our in vitro experiments revealed that irisin significantly activated Nrf2/HO-1 signaling and promoted the protein expressions of Nrf2 and HO-1 in the LPS-stimulated BV-2 microglial cells. Furthermore, irisin treatment reduced the expressions of iNOS and COX-2 in the BV-2 microglial cells induced by LPS and no significant difference on cell viability was found. Our study also confirmed that irisin treatment blocked the LPS-stimulated elevation in the expression of phosphorylated p38, ERK, and JNK as well as the phosphorylation of IRF3 and inhibited the degradation and phosphorylation of IκB and NF-κB microglial cells. These findings suggest that irisin treatment regulates microglia-mediated neuroinflammation through the NF-κB/MAPK/IRF3 signaling pathways.

5. Conclusion

The results of the present study found that aerobic exercise can improve the cognitive function and spatial learning ability affected by LPS by inhibiting microglia-mediated neuroinflammatory responses, increasing the level of antioxidant biomarkers Nrf2/HO-1 expression, and regulating the expression of proapoptotic and antiapoptotic biomarkers. The mechanism of action may be related to the increased expression of FNDC5 and its possible association with the upregulated expression of BDNF in the hippocampus and cerebral cortex of mice with MCI associated with AD. Additionally, the in vitro experiment demonstrated that the anti-inflammatory effects of exercise-induced irisin ameliorate inflammatory responses and microglial activation through inhibition of the NF-κB/MAPK/IRF3 signaling pathways in BV-2 microglial cells induced by LPS. The present findings suggest that aerobic exercise of appropriate intensity and duration should be recommended for patients with MCI associated with AD because physical inactivity increases the detrimental effects on brain function, while physical exercise improves cognitive function and spatial learning ability and protects the brain.

Future prospects

However, the current study has some limitations, our in vivo experiment results showed that aerobic exercise might have a beneficial effect on the improvement of memory function and modulate the microglia-mediated neuroinflammatory responses through increasing BDNF synthesis. Therefore, further studies are required to evaluate the detailed mechanisms underlying the modulatory action of aerobic exercise training on the neuronal synaptic changes, neurotransmitter levels, and amyloid deposition and their relationship with cognitive improvement. The roles of astrocytes in neuroinflammation still need to be investigated. In addition to evaluating the age, gender, exercise intensity, and what kind of physical exercise promotes brain health and function, is worth further investigation.

CRediT authorship contribution statement

Jae-Won Choi: Methodology, Investigation, Data curation. Sang-Woo Jo: Resources. Dae-Eun Kim: Resources. Il-Young Paik: Validation. Rengasamy Balakrishnan: Writing – review & editing, Writing – original draft, Supervision, Project administration, Investigation, Funding acquisition, Conceptualization.

Declaration of competing interest

None

Appendix A. Supplementary data

The following are the Supplementary data to this article.

Data availability

Data will be made available on request.