Rotating magnetic field improved cognitive and memory impairments in a sporadic ad model of mice by regulating microglial polarization

Abstract

Neuroinflammation, triggered by aberrantly activated microglia, is widely recognized as a key contributor to the initiation and progression of Alzheimer’s disease (AD). Microglial activation in the central nervous system (CNS) can be classified into two distinct phenotypes: the pro-inflammatory M1 phenotype and the anti-inflammatory M2 phenotype. In this study, we investigated the effects of a non-invasive rotating magnetic field (RMF) (0.2T, 4Hz) on cognitive and memory impairments in a sporadic AD model of female Kunming mice induced by AlCl₃ and D-gal. Our findings revealed significant improvements in cognitive and memory impairments following RMF treatment. Furthermore, RMF treatment led to reduced amyloid-beta (Aβ) deposition, mitigated damage to hippocampal morphology, prevented synaptic and neuronal loss, and alleviated cell apoptosis in the hippocampus and cortex of AD mice. Notably, RMF treatment ameliorated neuroinflammation, facilitated the transition of microglial polarization from M1 to M2, and inhibited the NF-кB/MAPK pathway. Additionally, RMF treatment resulted in reduced aluminum deposition in the brains of AD mice. In cellular experiments, RMF promoted the M1-M2 polarization transition and enhanced amyloid phagocytosis in cultured BV2 cells while inhibiting the TLR4/NF-кB/MAPK pathway. Collectively, these results demonstrate that RMF improves memory and cognitive impairments in a sporadic AD model, potentially by promoting the M1 to M2 transition of microglial polarization through inhibition of the NF-кB/MAPK signaling pathway. These findings suggest the promising therapeutic applications of RMF in the clinical treatment of AD.

Introduction

Alzheimer’s disease (AD) is a prevalent neurodegenerative disorder, affecting a substantial number of individuals aged 65 and above, with an estimated increase to 13.8 million by 2060, thus imposing a significant burden on families and societies worldwide. This burden is particularly pronounced in countries with aging populations, such as China [1]. AD patients manifest profound cognitive impairment and memory loss, primarily attributed to extensive neuronal loss within the central nervous system (CNS). These neuronal losses arise from the formation of senile plaques, characterized by extracellular amyloid deposition and neurofibrillary tangles resulting from the abnormal phosphorylation of intracellular tau protein [2]. Importantly, these pathological changes observed in AD are also common contributors to other forms of dementia [1].

Microglia, which reside within the CNS and keep it in a homeostatic state, are extremely active immune effector cells [3]. Under physiological circumstances, resting microglia actively scout the CNS environment through their motility [4], clean pathogens and misfolded proteins by phagocytosis, and regulate the remodeling of synapses, thus maintaining the normal structure and function of the CNS [5, 6]. In turn, the microenvironment in the brain also affects the plasticity of the microglia [7]. Microglia can differentiate into two distinct polarized phenotypes based on the changing microenvironment, much like the polarization of peripheral macrophages [8]. Neurodegeneration can be triggered by M1-type or pro-inflammatory microglia, which release neurotoxic substances, pro-inflammatory cytokines, such as interleukin-6 (IL-6), interleukin-1β (IL-1β), tumor necrosis factor (TNF-α), nitric oxide (NO), and reactive oxygen species (ROS) [9]. M2 microglia, an alternative activated form of microglia, are assumed to be the anti-inflammatory phenotype, which releases anti-inflammatory cytokines such as interleukin-10 (IL-10), transforming growth factor β1 (TGF-β1), interleukin-4 (IL-4), and have enhanced phagocytosis. Thus, the activation of M2 microglia contributes to the resolution of inflammation, tissue repair, and neuroprotection and maintains the homeostasis of the brain microenvironment [10, 11].

Neuroinflammation is one of the main elements that can accelerate the development of many neurodegenerative diseases including AD, in which microglia play a crucial role [12, 13]. The transient acute inflammatory response, a typical physiological reaction to a variety of possible stimuli, is advantageous to the CNS in the early stages of AD. When the stimuli are removed, the inflammation disappears, and the CNS returns to a homeostatic state [14]. However, microglia can be trapped in a state of sustained activation if the stimuli persist, such as Aβ, other extracellular or intracellular products of APP metabolism, and abnormal alterations in nerve fibers [15–17]. Then, continuously activated M1 microglia will keep producing high levels of pro-inflammatory factors, leading to the formation of long-term chronic neuroinflammation, which can be neurotoxic and create a vicious cycle that may exacerbate pathological changes and cause neurodegeneration and cognitive decline [18, 19]. Since neuroinflammation has a non-negligible impact on the onset and development of AD, several medications, such as rosiglitazone [20], benzimidazole [21], and several pyrimidine compounds [22], are available to suppress microglia activation or promote the conversion of microglia from M1 to M2 type. Nevertheless, due to the blood–brain barrier, the existing anti-inflammatory medicines of the CNS that target microglia activation or polarization transition do not have high efficacy [23].

Magnetobiology is an emerging interdiscipline that focuses on the biological effects and mechanisms of various magnetic fields. Magnetic fields are usually divided into static and variable magnetic fields. As a kind of time-varying magnetic field, a rotating magnetic field (RMF) is a kind of magnetic field that changes direction with a constant angular velocity and a constant period. RMF is generated in two main ways: rotating permanent magnets driven by motors and alternating currents passing through circular array coils. The RMF devices we use have been described in detail before [24]. As a non-invasive physical treatment tool, the biological effects of RMF biological effects have been discovered, including inhibiting the growth of cancer cells [25, 26] and harmful microorganisms [27, 28], promoting wound healing [29], protecting hematopoietic function [30, 31], increasing bone density [32], accelerating the recovery of osteonecrosis of the femoral head [33], improving enzyme activity [34], and regulating immune function [35]. Besides, in the study conducted by Hernández-Hernández et al., 6.4 mT, 7 and 10 Hz RMF promoted neurite growth in chromaffin cells [36]. In 2014, Suszyński et al. verified that 20–150 μT, 40 Hz RMF promoted regeneration of crushed sciatic nerve in rats [37]. In 2016, Chen et al. discovered that 0.02T, 6Hz RMF can promote the regeneration of the CNS in decapitated planarian Girardia Sinensis [38]. In addition, RMF can maintain the health status and extend the lifespan of Caenorhabditis elegans by improving the activities of antioxidase, reducing age-related pigment accumulation, delaying Aβ-induced paralysis, and increasing resistance to heat and oxidative stress [39]. However, so far, there is no report on whether RMF can improve cognitive disorders in AD by regulating inflammatory response.

In this study, we investigated whether RMF could ameliorate chronic inflammation in the brain by modulating microglial polarization, thereby exerting a neuroprotective effect and ameliorating the symptoms of AD. We hope our work can provide new ideas and new exploration for clinical treatment of AD.

Materials and methods

RMF device

Previously published studies have described the structure of the RMF exposure device in detail [24, 38]. Briefly, the core structure of the RMF device is composed of two parallel cylindrical NdFeB permanent magnets (each of them has a diameter of 98 mm, a height of 72 mm, and the distance between their centers is 270 mm) with opposite directions of magnetization and an iron plate at the bottom of the magnets; its rotation was driven by a motor (Fig. 1E, F). Besides, the device has a custom-made plastic cover for safety isolation and the placement of magnetic exposure subjects (Fig. 1A, B). The intensity near the upper surface of the magnets is about 0.4T, but at the level of animals and cells, the intensity is about 0.2T, which is simulated by ANSYS Maxwell 2020 software and measured by a Hall magnetometer (ETM-13-Achsen, Geneva, Switzerland), and its rotating frequency is 4 Hz (Fig. 1G). The RMF device can be used for the magnetic exposure of cells (Fig. 1C) and animals (Fig. 1D).

Fig. 1.

The RMF exposure apparatus. A Side view of RMF apparatus. B Top view of RMF apparatus. C The RMF exposure apparatus in the cellular room. D The RMF exposure apparatus in the animal room. E The core structure of RMF apparatus. F Travel direction of the magnetic induction line. G Distribution of magnetic induction intensity

Animal and experimental design

Previous studies have indicated a higher susceptibility of females to AD [1, 40–42], along with an elevated expression of pro-inflammatory cytokines [43]. Moreover, Kunming (KM) mice, which possess a close genetic homology to humans, have been extensively employed across various research fields, including immunology, genetics, pharmacology, and neurology [44]. These mice exhibit a diverse range of neurological signs and symptoms, rendering them suitable for investigating dementia, anxiety neurosis, stroke, and age-related memory deterioration [45–47]. Notably, KM mice demonstrate sensitivity to D-galactose (D-gal), which facilitates the establishment of AD models [48]. Consequently, female KM mice were selected as the experimental subjects in this study.

Female KM (20-week-old) were bought from the SPF (Beijing) Biotechnology Co., Ltd [SCXK (J) 2019–0010]. All experimental mice were housed in plastic cages in a temperature-controlled room (room temperature 23 ± 1 °C, humidity 55 ± 5%) with a 12-h light/12-h dark cycle and allowed access to food and water freely. After one week of adaptation, the animals were randomly divided into four groups (n = 7): (1) CTRL group (wild type), (2) RMF group (wild type + RMF), (3) AD group, and (4) AD + RMF group.

All mice were housed in a dedicated pathogen-free animal house at Shenzhen University. The Shenzhen University Animal Ethics and Welfare Committee has reviewed and confirmed that the use of animals and experimental protocols were by the guidelines established by the Animal Ethics and Welfare Committee. Each procedure was designed to minimize animal suffering and to limit the number of animals used.

AD model building and treatment

The induction of the sporadic AD model using aluminum chloride (AlCl₃) and D-gal is a well-established approach, and the selection of AlCl₃ and D-gal doses was guided by previous experimental reports [49–54]. In the AD model group and AD + RMF group, animals received orally AlCl₃ (100 mg/kg b.wt./day) dissolved in drinking water. Meanwhile, they also received subcutaneous injection with D-gal (120 mg/kg b.wt./day). According to the experience of our research group, the RMF and AD + RMF groups were exposed to RMF (4 Hz, 0.2 T) for 2 h daily. The CTRL and AD group mice were exposed to sham RMF (4 Hz, same device as the former RMF device, but not magnetized) for the same schedule. AD model building and RMF treatment are performed simultaneously. All the treatments above were performed for 90 days.

BV2 microglial cell culture

BV2 cells are immortalized murine-derived microglial cell lines obtained by stable transfection of c-myc oncogene using a J7-retrovirus_ENREF_14 and can exhibit phenotypic and functional properties of primary microglia [55]. The BV2 cells were cultured in DMEM (high glucose) medium supplemented with 10% heat-inactivated FBS, 100 U/mL of penicillin, and 100 μg/mL of streptomycin and maintained in a humidified incubator with 5% CO₂ at 37 °C.

Preparation of oligomeric Aβ_1–42

Oligomeric Aβ_1–42 was prepared as previously described [56]. Aβ_1-42 lyophilized powder was removed from − 20 °C and placed in equilibrium at room temperature for 30 min. Then, the peptide was dissolved in 222 μL HFIP (1,1,1,3,3,3-hexafluoro-2-propanol) and incubated for 30 min at room temperature. The HFIP was then removed by evaporation. Then, the peptide was dissolved in 44.4 μL anhydrous DMSO. After sonication in a water bath for 10 min and centrifugation at 1000 rpm for 30 s, phenol red-free F12 medium precooled at 4 degrees for 24 h was added into the solution.

Behavioral analysis

Morris water maze (MWM) test

The MWM test was conducted as previously described to evaluate mice’s capacity for spatial learning and memory [57]. A big circular plastic pool of 120 cm in diameter and 60 cm in height was used as the main component of the water maze. The pool was filled to a depth of 35 cm with opaque black water of 25 ± 2 °C. The pool was divided into four equal quadrants (I, II, III, and IV), and in each trial, the center of quadrant IV (the target quadrant) had a hidden circular platform with a diameter of 9 cm that was positioned 1 cm below the water surface. For five days straight, each mouse completed four pieces of training each day. The longest training lasted 60 s. If the mouse did not get to the platform in 60 s, it would be manually directed there and given 10 s to stay. After the navigation test, the mice were given 60 s to swim and identify the platform in a detection test, which involved removing the hidden platform. Escape latency, swimming speeds, swimming tracks, times crossing the target quadrant, and the distance traveled within the target quadrant were recorded and analyzed using the Morris maze analysis system (ZS Dichuang, Beijing, China).

Novel objection recognition (NOR) test

The NOR test is based on mice’s propensity to interact more naturally with novel objects than those they are familiar with. In the habituation phase, mice were given 5 min to freely explore the behavioral field (a 40 cm × 40 cm × 40 cm white plastic box, empty) before the testing day. During the training phase, the mice were placed in a box with two identical objects on two corners and allowed to explore for 5 min. After a 24-h retention period, a novel object was placed in the proper spot during the test session; the mice were reintroduced to the box and given 5 min to investigate it. The time spent exploring and sniffing each object was recorded. The results were expressed as a recognition index (RI = Tn/(Tn + Tf)). Between two adjacent trials, the box was cleaned with 70% alcohol to eliminate olfactory cues. An automated video tracking system (EthoVision XT V14) was used to record and analyze the mice’s exploratory activities.

Immunohistochemistry

Mice were deeply anesthetized with isoflurane and perfused with ice-cold PBS containing heparin (10 U/mL) for cardiopulmonary perfusion before sacrifice. Their brains were immediately removed and segmented along the sagittal plane. The left cerebral hemisphere was fixed overnight in 10% paraformaldehyde and processed for paraffin-embedded sections. For immunostaining, 5-μm coronal serial sections were deparaffinized, and antigen was retrieved with citrate buffer (0.01 M, pH 6.0) at 95 °C for 20 min. Sections were then incubated with 3% H₂O₂ and washed three times with 1 × PBS. The sections were subsequently permeabilized and blocked with 10% normal goat serum in 0.3% Triton X-100 PBST for 1 h at room temperature. After that, the sections were immune-stained with anti-NeuN (Abcam, USA) and anti-6E10 (Biolegend, USA), respectively, followed by the incubation with appropriate HRP-labeled secondary antibodies, and visualized with diaminobenzidine (DAB). After dehydration with alcohol and xylene, the resin was used to seal tissue section samples. Finally, images were obtained using a microscope (Olympus, Tokyo, Japan). The quantitative images of NeuN-positive neuron cells were selected in the similar region of the CA1 and cortex, respectively. The number of labeled NeuN-positive neuron cells was quantified with ImageJ software (National Institutes of Health, USA). In addition, to determine the burden of amyloid plaque, similar locations in the CA1, CA3, and cortex were selected, and the percent area of positive labeling was analyzed at × 20 magnification.

Immunofluorescence staining

The right cerebral hemisphere was fixed overnight in 10% paraformaldehyde and processed for frozen sections. The sections were then removed from − 20 °C and placed at room temperature for 20 min and washed 3 times with 1 × PBS. Then, the sections or BV2 cells were kept in a blocking solution (1% BSA and 0.1% Triton X-100) for 1 h at room temperature and then incubated with primary antibodies against CD86, anti-CD206, anti-Iba 1, and anti-6E10 overnight at 4 °C, respectively, followed by the staining of Alexa Fluor 488, 594, 647 secondary antibodies for 2 h at 37 °C. Finally, the sections were counterstained with DAPI and examined using a fluorescence microscope (Olympus, Tokyo, Japan). Mean fluorescence intensity was used to analyze the positive labeling for CD86, CD206, and Iba 1. To further determine the number of microglia around plaques, quantitative images of amyloid plaques in the cortex were randomly selected, and the CD206-positive microglia within a radius of 30 μm from the edge of plaques were quantified using ImageJ software (National Institutes of Health, USA).

Nissl staining

After deparaffinized and hydration, brain tissue sections were stained with Cresyl Violet (Beyotime, Shanghai, China) for 1 h in a 56 °C oven. The sections were then dehydrated in ethanol, transparentized in xylene, and sealed with central resin. Images were obtained using a microscope (Olympus, Tokyo, Japan).

Lumogallion staining

Lumogallion (Macklin, California, USA) was dissolved in 0.1M acetic acid buffer (pH = 5.2) to prepare a 10-μM solution. The samples were then incubated with 10 μM lumogallion for 1 h away from light, followed by thoroughly washing twice with acetic acid buffer solution (15 min each time). Images were obtained using a fluorescence microscope (Olympus, Tokyo, Japan). ImageJ software (National Institutes of Health, USA) was applied to analyze the mean fluorescence intensity.

Assay of NO production

NO production in cells and serum was measured by the Griess method via NO assay kit (Beyotime Biotech Inc., Jiangsu, China) according to the manufacturer’s instructions. Nitrite levels were normalized by protein concentrations, and the data were shown as a percentage of control levels.

Measurement of reactive oxygen species

Intracellular generation of ROS in BV2 cells and serum were determined using the 5-(and-6)-chloromethyl-2’,7’-dichlorodihydrofluorescein diacetate (H2DCFDA) detection assay kit (Abcam, USA) according to the manufacturer’s guidelines.

Enzyme-linked immunosorbent assay

The blood was centrifuged at 1000 g for 15 min at 4 °C, and the serum was extracted. Aliquots of serum were obtained and stored at − 20 °C for further examination. The cytokine concentrations in serum and cell supernatant, including IL-6, TNF-α, TGF-β1, and IL-4, were assessed via an enzyme-linked immunosorbent assay (ELISA) kit (Invitrogen, Waltham, MA, USA) according to the manufacturer’s guidelines.

Cell viability assay

1 × 10⁴ BV2 cells were seeded in each well of 96-well plates. After 24 h of cell growth, the cells received different concentrations (0 μM, 5 μM, 10 μM, 20 μM, and 40 μM) of Aβ_1-42 treatment for 24 h and were exposed to RMF for 0 h, 1 h, 2 h, 3 h, and 4 h. Then, the cell viability was assessed using a Cell Counting Kit-8 assay (MedChemExpress, MCE, USA) according to the manufacturer’s guidelines. The absorbance was measured at 450 nm using a microplate reader (Bio Tek, Shenzhen, China).

Flow cytometry analysis

BV2 cells were collected after receiving 24-h Aβ_1–42 oligomer stimulation and 2-h RMF treatment and labeled by cell surface markers, including fluorochrome-conjugated CD86 and CD206 (BD Biosciences, USA) antibodies. The cells were then softly mixed with the markers and incubated at 4 °C in the dark for 30 min. Finally, a total of 50,000 fluorescence events were aggregated. Negative controls were unlabeled cells.

Quantitative PCR

Total RNA from the cerebral cortex, hippocampus, and BV2 cells were extracted by TRIZOL (Sigma-Aldrich, USA), respectively, and reverse transcription of the RNA (1000 ng) to cDNA was conducted by Evo M-MLVRT Kit with gDNA Clean for qPCR assay (AG, Guangzhou, China). Quantitative PCR was accomplished utilizing the SYBR ®Green Premix Pro TaqHS qPCR Kit (Rox Plus) (AG, Guangzhou, China). The conditions for amplification were 95 °C for 30 s, 40 cycles at 95 °C for 5 s, and 60 °C for 30 s. GAPDH served as a control. All primers were manufactured by BGI (Wuhan, China), and the sequences are shown in Table 1.

Table 1.

Primer sequence

Gene		Primer sequence
GAPDH	Forward	5’-CTTGTGCAGTGCCAGCC-3’
	Reverse	5’-GCCCAATACGGCCAAATCC-3’
IL-6	Forward	5’-CTCCCAACAGACCTGTCTATAC-3’
	Reverse	5’-CCATTGCACAACTCTTTTCTCA-3’
IL-1β	Forward	5’-CACTACAGGCTCCGAGATGAACAAC-3’
	Reverse	5’-TGTCGTTGCTTGGTTCTCCTTGTAC-3’
TGF-β1	Forward	5’-CCAGATCCTGTCCAAACTAAGG-3’
	Reverse	5’-CTCTTTAGCATAGTAGTCCGCT-3’
TNF-α	Forward	5’-ATGTCTCAGCCTCTTCTCATTC-3’
	Reverse	5’-GCTTGTCACTCGAATTTTGAGA-3’
Arg-1	Forward	5’-CATATCTGCCAAAGACATCGTG-3’
	Reverse	5’-GACATCAAAGCTCAGGTGAATC-3’
iNOS	Forward	5’-ATCTTGGAGCGAGTTGTGGATTGTC -3’
	Reverse	5’-TAGGTGAGGGCTTGGCTGAGTG-3’
IL-4	Forward	5’-TACCAGGAGCCATATCCACGGATG-3’
	Reverse	5’-TGTGGTGTTCTTCGTTGCTGTGAG-3’
IL-10	Forward	5’-TTCTTTCAAACAAAGGACCAGC-3’
	Reverse	5’-GCAACCCAAGTAACCCTTAAAG-3’

Western blot analysis

The brain tissues and BV2 cells were homogenized in RIPA containing protease and phosphatase inhibitors. The lysate was further homogenized by sonication and then centrifuged at 5000 rpm for 10 min at 4 °C. After quantifying the protein concentration using the BCA protein assay kit (Thermo Fisher Scientific, Waltham, MA, USA), the samples were denatured by the metal bath at 95 °C for 10 min. Then, the proteins were separated by 7.5 ~ 10% gels and transferred into PVDF membranes. The PVDF membranes were blocked with 5% non-fat milk at room temperature for 1 h and incubated overnight with specific antibodies (anti-IL-6, anti-iNOS, anti-TGF-β1, anti-Arg-1, anti-NeuN, anti-TLR4, anti-Myd88, anti-p-IKKα/β, anti-p-IкBα, anti-p-NF-кB p65, anti-p-JNK, anti-p-p38, anti-p-ERK, anti- IKKα/β, anti-IкBα, anti-NF-кB p65, anti-JNK, anti-p38, and anti-ERK) (Cell Signaling Technology, CST, USA) at 4 °C. After washing with TBST three times, the blot was incubated with the corresponding secondary antibody for 1 h at room temperature. Finally, the blots were visualized using the ECL Protein Blot Kit (MeilunBio®, China) and a Gel Imaging System (Tanon-5220, Shanghai, China) and analyzed using Image J software.

TUNEL staining

The TUNEL staining was conducted using the TUNEL kit (Vazyme, Nanjing, China) according to the manufacturer’s guidelines. The sections were removed from − 20 °C, left at room temperature for 20 min, and washed three times with 1 × PBS. Then, the slices were digested with 100 μL protease K (20μg/mL) and incubated at room temperature for 10 min. The sections were placed in a jar containing 1 × equilibration buffer and left at room temperature for 20 min. After that, 50 μL of TdT buffer (34 μL ddH₂O, 10 μL 5 × equilibration buffer, 5 μL BrightGreen labeling mix, and 1 μL recombinant TdT enzyme) was added, and the sections were incubated for 1 h at 37 °C away from light and washed twice in PBS. The slides were then immersed in a dyeing tank containing PI solution (1 μg/mL), left at room temperature for 5 min, and washed three times with 1 × PBS. TUNEL-positive cells were visualized under a fluorescence microscope (Olympus, Tokyo, Japan) and counted using ImageJ software (National Institutes of Health, USA).

Total RNA extraction and RNA sequencing

After the mice were euthanized, the half-brain hippocampal tissues were isolated and transferred into the grinding tubes. Then, tissue samples were ground into powder with liquid nitrogen and transferred into the lysis buffer, which were further grounded for 30 s and left for 5 min. After centrifuging at 12,000 g for 5 min at 4 °C, the supernatants were transferred to the centrifuge tubes with 300 µL chloroform/isoamyl alcohol (24:1), mixed, and thoroughly shaken. Then, the samples were centrifuged at 12,000 g for 8 min at 4 °C. After that, the supernatants were transferred to 1.5-mL centrifuge tubes, and 2/3 volume of isopropyl alcohol was added to each tube. The tubes were gently inverted and mixed and were placed at − 20 °C for more than 2 h. Then, the samples were centrifuged at 17,500 g for 25 min at 4 °C. After discarding the supernatants, the precipitations were washed twice with 0.9-mL 75% ethanol and dissolved with 20-µL RNase-free water.

The Illumina HiSeq (BGI Genomics, BGI-Shenzhen, Shenzhen, China) was used to index, assemble, and sequence the data. Base-calling and demultiplexing were done using a custom Python demultiplexing tool and Illumina’s BCL2FASTQ software. Each gene’s residual standard deviation against its average log count was plotted to assess each gene’s performance, and the result was a trend line that closely resembled the residuals. Benjamini–Hochberg adjusted p-values less than or equal to 0.05 were chosen to identify treatment differences following the differential expression analysis. Global variations in well-known Gene Ontology (GO) keywords and the Kyoto Encyclopedia of Genes and Genomes (KEGG) were found for each comparison retrieved by Limma using the R/Bioconductor tool GAGE8. The GO item data with a Benjamini–Hochberg corrected P-value of less than or equal to 0.05 was displayed using Heatmap 3. Additionally, annotated graphs for the disturbed KEGG pathway were created using p-values less than or equal to 0.05. In these graphs, the estimated log twofold variation of the term’s genes was significantly perturbed either in one direction relative to the background or in the other direction relative to other genes in the phrase.

Abbreviations

All abbreviations used in the text are listed in Table 2.

Table 2.

Abbreviations

Abbreviations	Definition
AD	Alzheimer’s disease
Aβ	amyloid β
AlCl3	aluminum chloride
CNS	central neural system
IL-6	interleukin-6
IL-1β	interleukin-1β
TNF- α	tumor necrosis factor a
NO	nitric oxide
ROS	reactive oxygen species
IL-10	interleukin-10
TGF-β1	transforming growth factor β1
IL-4	interleukin-4
RMF	rotating magnetic field
Arg-1	arginase-1
HFIP	1,1,1,3,3,3-hexafluoro-2-propanol
Iba 1	ionized calcium-binding adapter molecule 1
NOR	novel objection recognition
ELISA	enzyme-linked immunosorbent assay
CCK-8	cell counting kit-8
MWM	Morris water maze
CTRL group	wild-type group
RMF group	wild-type mice received RMF treatment
AD group	wild-type mice received orally AlCl3
AD + RMF group	wild-type mice received orally AlCl3 and RMF
PBS	phosphate buffer solution
DAB	diaminobenzidine
PVDF	polyvinylidene fluoride
HE	hematoxylin–eosin
DG	dentate gyrus
CA1	Cornu ammonis 1
CA3	Cornu ammonis 3
MAPKs	mitogen-activated protein kinases
ERK	extracellular signal-regulated kinases
JNK	c-jun N-terminal kinases
MyD88	myeloid differentiation factor 88
NF-κB	nuclear factor kappa B
TLR4	toll-like receptor 4

Statistical analysis

The statistical analysis was performed using GraphPad Prism 9.0, and all data were presented as mean ± SEM. The variance between the two groups was analyzed using an unpaired, two-tailed Student’s t-test. To analyze the data from the four groups, a one-way analysis of variance (one-way ANOVA) followed by Tukey’s post hoc tests was applied. For the escape latency line graphs in the Morris water maze (MWM) results, a one-way ANOVA was conducted to assess the escape latency for each day across the four experimental groups. Statistical significance was considered at a P-value less than 0.05 (P < 0.05).

Result

RMF treatment improved learning and memory impairments in AD model mice

After a 3-month treatment, the spatial learning and memory abilities of the AD model mice were examined with Morris’s water maze. As can be seen in the behavioral track images in Fig. 2A, the mice in the AD group were disoriented, whereas the other three groups explored around the platform. In the probe trial, the escape latency of the AD group was significantly longer than that in the CTRL group since the 4th day. However, the escape latency of AD model mice remarkably decreased after RMF treatment (Fig. 2B). Besides, compared to the mice in the CTRL group, AD model mice also exhibited fewer times crossing the target quadrant and shorter distances traveled in the target quadrant. However, these behavioral performances in the AD model of mice were significantly improved after RMF treatment (Fig. 2C, D). Specifically, there was no difference among the four groups in swimming speed; therefore, changes in athletic ability were excluded as the cause of the observed improvement (Fig. 2E). Besides, the effect of RMF on cognitive impairment was also validated through new object recognition experiments. As shown in Fig. 2E, the mice in the AD group were unable to discriminate between novel objects and familiar objects significantly, while RMF treatment increased the tendency of AD mice to interact with a new object rather than with a familiar object.

Fig. 2.

RMF treatment ameliorated learning and memory deficits in AD model mice. A Representative track images of mice in the probe test of the Morris water maze. B Escape latency to get to the platform during the training (*P < 0.05 and ***P < 0.001: AD vs. CTRL; #P < 0.05 and ##P < 0.01: AD vs. AD + RMF). C Target (platform) entries in the probe test. D Distance spent in the target quadrant in the probe test. E Swimming speed in the probe test. F Novel object recognition index in the novel object recognition test. Data are expressed in the form of means ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001; ns, not significant. n = 6

RMF treatment alleviated the damage to hippocampal morphology and prevented synaptic and neuronal loss in AD model mice

The loss of neurons and synapses at the hippocampus and cortex is closely related to spatial cognitive ability [58]. Nissl staining showed that the hippocampus in the CTRL group had normal morphology and neatly arranged pyramidal cell layers. Compared to the CTRL group, the hippocampus in the AD group had slightly smaller volume and fewer nissl bodies. Besides, pyramidal cells showed disorganized arrangement, nuclear division, hyperchromatic nuclei, irregular shape, and eosinophilic changes of the CA1 area in the AD group. After RMF treatment, the morphology of the hippocampus was almost the same as that in the CTRL group, and the degree of morphological changes in the hippocampus was also reduced (Fig. 3A). To further verify the neuronal damage, immunohistochemistry was used to examine the number of NeuN-positive cells in the hippocampus and cortex. The number of NeuN-positive cells in the hippocampus and cortex was significantly decreased in the AD group compared with the CTRL and RMF groups, while the neuronal loss was significantly alleviated after RMF treatment (Fig. 3B, D, E). Western blot of NeuN also further confirmed the immunohistochemical results (Fig. 3C, F, G).

Fig. 3.

RMF treatment attenuated hippocampal morphological damage and neuronal death. A Representative images of Nissl staining in the hippocampus and the CA1 region. B Representative images of NeuN staining in the hippocampal CA1 and cortex were examined by Immunohistochemistry. Scale bars = 50 μm. Quantification of NeuN-positive cells in CA1 (C) and cortex (D). E–G levels of NeuN in the hippocampus and cortex were examined by Western blot and quantified with Image J software. Data were performed as means ± SEM. *P < 0.05, **P < 0.01 and ***P < 0.001; ns, not significant. n = 6

RMF treatment relieved the cell apoptosis in the hippocampus and cortex in AD model mice

In AD, the CNS always experiences varying degrees of cell death especially neurons [59]. Tunnel staining results revealed that the rate of apoptosis was not statistically different between the CTRL and RMF groups, but that it was much higher in the DG and cortical areas in the AD group. After RMF treatment, the apoptosis rate in the AD + RMF group was dramatically reduced compared to the AD group (Fig. 4).

Fig. 4.

RMF treatment relieved hippocampal and cortical cell apoptosis. A, B Representative images of TUNEL staining in the hippocampal DG (A) and cortex (B) region. Scale bars = 100 μm. C, D Statistical analysis of apoptosis rate in the hippocampal DG (C) and cortex (D) region. Data were performed as means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant. n = 5

RMF treatment induced M2 microglial polarization and suppressed M1 microglial activation in AD model mice

The etiology of AD is closely related to neuroinflammation, which is characterized by increased microgliosis and persistent release of pro-inflammatory cytokines [60]. In this study, immunofluorescence was applied to identify microglial activation (ionized calcium-binding adaptor molecule 1, Iba 1) and the expression of microglial type markers to ascertain in each group. Furthermore, M1 microglia are identified by the markers CD86 and IL-6, while M2 microglia are identified by CD206 and TGF-1. Immunofluorescence and immunohistochemistry assay of Iba1 revealed that after treatment with AlCl₃ and D-gal, microglia were significantly activated, and their morphology changed significantly. The CA1 region and cortical microglial somas in the AD group were substantially larger than those in the CTRL group, whereas the protrusions shrank and the number of the branches increased, while RMF treatment dramatically reduced microglial activation without significant alterations (Fig. 5A). Western blot analysis consistently revealed a decrease in hippocampus and cortical Iba-1 expression levels after RMF therapy (Fig. 5C–E). Besides, according to the immunofluorescence results, RMF treatment significantly decreased the fluorescence intensity of CD86 (Fig. 5F, H) while significantly increasing the fluorescence intensity of CD206 (Fig. 5G, I) in the AD + RMF group.

Fig. 5.

RMF treatment inhibited M1-type microglia polarization and promoted M2-type microglia polarization. A, B Co-immunofluorescence staining of the CA1 (A) and cortex region (B) for Iba 1 (activated microglia marker) (green) and DAPI (blue). C–E Expression levels of Iba 1 in the hippocampus and cortex were examined by Western blot and quantified with ImageJ software. F–G Co-immunofluorescence staining of the CA1 region for M1 marker CD86 (red) (F), M2 marker CD206 (red) (G), and DAPI (blue). Scale bars = 50 μm. H, I Statistical graph of (F) and (G), respectively. Data were performed as means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant. n = 5

RMF treatment improved the pro-inflammatory environments and promoted the secretion of anti-inflammatory cytokines in the AD mice

The role of RMF in inflammatory regulation was evaluated through ELISA, qPCR, and Western blot. The levels of pro-inflammatory cytokines, such as IL-6 and TNF-α, were significantly increased in the AD group compared to the CTRL group, while the levels of anti-inflammatory cytokines, such as IL-4 and TGF-1, were significantly reduced. RMF treatment, however, significantly suppressed the release of pro-inflammatory cytokines and elevated the secretion of anti-inflammatory cytokines. When compared to the CTRL group, the levels of two kinds of cytokines were not significantly different between the RMF group and the AD + RMF group (Fig. 6A). The results of qPCR showed that RMF treatment promoted the expression of anti-inflammatory cytokines (TGF-β1, Arg-1, IL-4, IL-10) (Fig. 6C) and reduced the expression of inflammatory factors in the brain (IL-6, IL-1β, TNF-α, iNOS) (Fig. 6B). The ELISA and qPCR results were further verified by Western blot of IL-6, iNOS, TGF-β1, and Arg-1 (Fig. 6D-E).

Fig. 6.

RMF treatment regulated the secretion of pro-inflammatory/anti-inflammatory cytokines in the AD mice. A The levels of IL-6, TNF-α, IL-4, and TGF-β1 in the serum of AD model mice were determined using corresponding ELISA kits. B Relative mRNA levels of IL-6, IL-1β, iNOS, and TNF-α in the hippocampus of AD model mice. C Relative mRNA levels of TGF-β1, Arg-1, IL-4, IL-10 in the hippocampus of AD model mice. D–F Expression levels of IL-6, iNOS, TGF-β1, and Arg 1 in the hippocampus and cortex were examined by Western blot and quantified with ImageJ software. Data were performed as means ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001; ns, not significant. n = 6

RMF treatment reduces Al deposition in the brain of AD mice

Aluminum (Al) was recognized as a neurotoxic metal element for the first time in the early 1970s [61]. Since AlCl₃ was used to induce the AD model, it was also investigated whether RMF treatment can enhance Al clearance and inhibit Al accumulation in the AD brain. By examining the results of Lumogallion staining, it was discovered that Al deposition was evident in the AD group, particularly in the CA3 region, while there was essentially no Al deposition in the mouse brains of the CTRL and RMF groups. More importantly, Al accumulation in the brains of mice in the AD + RMF group was dramatically reduced after RMF treatment (Fig. 7A, B).

Fig. 7.

RMF treatment reduced Al deposition in the brain of AD mice. A Representative images of Lumogallion staining in the hippocampal CA3 and cortex region. Scale bars = 50 μm. B statistical graph of (A). Data were performed as means ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001; n = 6

RMF treatment reduced ROS levels in the brain and NO production in the serum of AD mice

ROS plays a key essential role in the pathogenesis of AD due to their deleterious effects including oxidative damage to cellular DNA, lipids, and proteins and the capability of inducing inflammatory responses [62]. Thus, the degree of central and peripheral inflammatory response was assessed by measuring ROS levels in the brain and NO levels in the serum, respectively. The results showed that ROS and NO levels were significantly increased in the AD model group compared with the CTRL group, while both decreased significantly after RMF treatment (Fig. 8).

Fig. 8.

RMF treatment reduced ROS levels in the brain and NO production in the serum of AD mice. A ROS level in the hippocampal CA3 and cortex region. Scale bars = 50 μm. B Statistical graph of (A). C The level of NO in serum was determined using NO assay. Data are means ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001; ns, not significant; n = 6

RMF treatment reduced Aβ deposition in the AD model mice

To assess the effect of RMF on pathological changes in amyloid, the brain sections were immune-stained with antibodies of Aβ. A large number of amyloid plaques were mainly distributed in the cortex and hippocampus of AD model mice, while amyloid plaques cannot be detected in the same brain regions of mice in the CTRL and RMF groups (Fig. 9A, B). After RMF treatment, the average area occupied by Aβ-positive plaques significantly decreased in the AD + RMF group. In addition, immunofluorescent double staining for Aβ and CD206 was used to locate amyloid plaques and M2-type microglia. The results showed that in the AD group, the expression of CD206 in microglia was significantly lower than that in the control group. After RMF treatment, the expression of CD206 was significantly up-regulated, M2 microglia were recruited near amyloid plaques, and more antennae protruded into the plaques (Fig. 9C).

Fig. 9.

RMF treatment reduced Aβ deposition in the AD model mice. A Representative images of Aβ (6E10) staining in the hippocampal CA1 and CA3 region. Scale bars = 100 μm. B Quantification of Aβ plaques in the cortex and hippocampal CA1 and CA3 region. C Co-immunofluorescence staining of Aβ (6E10) (red) and M2 marker CD206 (green). Scale bars = 20 μm. D The number of CD206-positive microglia around plaques. Data were performed as means ± SEM. *P < 0.05; ns, not significant. n = 6

RMF promoted the polarization transition of microglia from M1 to M2 by inhibiting the NF-кB/MAPK pathway

To clarify the mechanism by which RMF promoted the polarization transition from M1 to M2 in microglia, RNA from mouse hippocampal tissue was sequenced using whole-transcriptome sequencing. Volcano maps for the differential gene in AD and AD + RMF groups showed that the number of differential genes is large in AD and AD + RMF groups (Fig. 10A). By classifying the differential genes KEGG pathway, it was found that the altered genes were mainly focused on Alzheimer’s disease and TLR4-Myd88-MAPK-related signaling pathways (Fig. 10B, C). Subsequently, the expression of MAPK-related genes was clustered, and the heat map was plotted; the results showed that the expression of MAPK pathway-related genes was significantly up-regulated in the AD group (Fig. 10D). Finally, the changes of NF-кB and TLR4-Myd88-MAPK signaling molecules were further verified by Western blot. As shown in Fig. 9, RMF significantly down-regulated the expression of TLR4 and Myd88 and phosphorylated IKKα/β, IкBα, NF-кB p65, JNK, p38, and ERK (Fig. 11A–I).

Fig. 10.

The whole-transcriptome sequencing of hippocampal tissues. A Volcano maps of differential genes between AD and AD + RMF group. B, C Classification map of differential gene KEGG pathway. D Heatmap for the differential gene in MAPK signal pathway. n = 3

Fig. 11.

RMF treatment inhibited NF-κB and MAPK pathways in vivo. A–I Expression levels of TLR4, Myd88, p-IKKα/β, IKKα/β, p-IкBα, IкBα, p-NF-кB p65, NF-кB p65, p-JNK, JNK, p-p38, p38, p-ERK, and ERK in the hippocampus were examined by Western blot and quantified with ImageJ software. Data were performed as means ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001, ns, not significant. n = 3

RMF treatment promoted amyloid phagocytosis in BV2 cells

AD pathology is characterized by the deposition of Aβ peptides including Aβ_1-42 in extracellular amyloid plaques. Amyloids can form a variety type of protein polymers, including monomers, dimers, trimers, oligomers, protofibrils, and fibrils. However, there is growing evidence that small oligomers formed by early aggregates of Aβ directly involved in neuronal functional damage and death [56]. Thus, Aβ_1-42 oligomers were used to pre-treat BV2 cells to establish a cell model of AD in this study. Aβ_1-42 oligomers were prepared according to the protocol of Violetta et al. [56] and verified their oligomeric morphology by Western blot (Fig. 12A). Then, the CCK-8 assay was used to determine the optimal concentration of Aβ_1-42 oligomers for cell pre-treatment. It was found that cell viability was significantly inhibited by Aβ_1-42 concentrations above 40 μM. Therefore, 20 μM Aβ_1-42 was chosen (Fig. 12B). Besides, to exclude the effect of RMF itself on cell viability, CCK-8 assay was used to detect cell viability after RMF treatment for 0 ~ 4 h; the results showed that RMF had no considerable influence on cell viability (Fig. 12C). In addition, to verify whether RMF can promote the phagocytic ability of microglia, the level of Aβ in BV-2 cells was detected by immunofluorescence (Fig. 12D, E) and Western blot (Fig. 12F, G). The results showed that compared with the Aβ group, the content of Aβ in BV2 cells was significantly increased after RMF treatment in the Aβ + RMF group.

Fig. 12.

RMF treatment promotes amyloid phagocytosis in BV2 cells. A Identification of Aβ_1-42 oligomers was determined by Western blot using 6E10 antibody. B Viability of BV2 cells pre-treated by Aβ_1-42 in different concentrations after 24 h. C Viability of BV2 cells treated with RMF at different times. D Co-immunofluorescence staining of BV2 cells of Iba 1 (red), 6E10 (green), and DAPI (blue). E Statistical graph of (D). F, G Expression levels of 6E10 in BV2 cells after Aβ_1-42 pre-treatment were examined by Western blot and quantified with ImageJ software. Scale bars = 50 μm. Data were performed as means ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001; ns, not significant, dimensionless unit. n = 3

RMF treatment inhibited the release of pro-inflammatory cytokines and promoted the secretion of anti-inflammatory cytokines in vitro

ELISA, qPCR, and Western blot were utilized to verify the effects of RMF on the release of inflammatory factors. Cell supernatants from each group at 3 h, 6 h, 12 h, and 24 h after Aβ stimulation were collected to conduct an ELISA assay, and the results showed that the concentration of IL-6 was significantly increased at each time point in the AD group, which decreased significantly after RMF treatment (Fig. 13A). In addition, in the cell supernatants collected at 3 h, 12 h, and 24 h, the level of TGF-β1 in the Aβ + RMF group showed an increase compared to the Aβ group, but it was not statistically significant (Fig. 13B). In the cell supernatant collected 6 h after Aβ stimulation, the TGF-β1 content in the Aβ group was significantly decreased compared to the other three groups. Then, cellular RNA after 6 h and 24 h of Aβ pre-treatment was extracted, and qPCR assays were performed; the results showed that the mRNA expression levels of IL-6 (Fig. 13C), IL-1β (Fig. 13D), and TNF-α (Fig. 13E) were significantly increased after 6 h and 24 h of Aβ stimulation, and the mRNA level of iNOS did not change after 6 h of Aβ stimulation and 2 h of RMF treatment, but the level of iNOS decreased significantly after 24 h of Aβ stimulation (Fig. 13F). The mRNA levels of anti-inflammatory cytokines TGF-β1 and IL-10 showed a significant increase (Fig. 13G, J) after RMF treatment, while the mRNA level of Arg-1 had no change after 6-h Aβ stimulation and 2-h RMF treatment. The mRNA level of Arg-1 in the Aβ + RMF group was significantly increased compared to the Aβ group after 6 h of Aβ stimulation and 2 h of RMF treatment (Fig. 13H). As for IL-4, its mRNA levels showed an increased trend but no statistical difference after RMF treatment (Fig. 13I). Western blot of IL-6, iNOS, TGF-β1, and Arg-1 also further validated the results of ELISA and qPCR (Fig. 13K). In addition, the level of ROS in cells and the content of NO in cell supernatant after 24 h of Aβ treatment were measured, and the results showed that RMF significantly inhibited ROS and NO production (Fig. 15A, B). Besides, the ratio of M1 and M2 microglia was verified by immunofluorescence and flow cytometry. It was discovered that the percentage of M1 BV2 cells was dramatically decreased following RMF treatment (Fig. 14A, C), while the number of M2 microglia was significantly elevated (Fig. 14B, D). Apart from the production of pro-inflammatory cytokines, the migration of microglia to the site of inflammatory stimulation also plays an important role in the development of neuroinflammation [63, 64]. Thus, it was examined whether RMF affected the migration of BV2 cells using a trans-well assay. The results showed that RMF significantly inhibited the migration of BV2 cells induced by Aβ stimulation (Fig. 15C, D).

Fig. 13.

RMF treatment inhibited the release of pro-inflammatory cytokines and promoted the secretion of anti-inflammatory cytokines in vitro. A, B The levels of IL-6 (A) and TGF-β1 (B) in the supernatant collected in 3 h, 6 h, 12 h, and 24 h after Aβ stimulation were determined using corresponding ELISA kits. C–J Relative mRNA expression levels of IL-6 (C), IL-1 β (D), iNOS (E), TNF- α (F), TGF-β1 (G), Arg-1 (H), IL-4 (I), and IL-10 (J) in BV2 cells. K-O Protein expression levels of IL-6, iNOS, TGF-β1, and Arg 1 in BV2 cells were examined by Western blot and quantified with ImageJ software. Data were performed as means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant. n = 5

Fig. 15.

RMF treatment attenuated ROS, reduced NO production, and inhibited BV2 cell migration. A Intracellular ROS levels in Aβ_1-42 pre-treated BV2 cells were determined using H2DCFDA assay. B The levels of NO in the supernatant were determined using NO assay. C The changes in cell migration were determined by trans-well assay. D The ratio of cell migration. Scale bars = 50 μm. Data were performed as means ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant. n = 3

Fig. 14.

RMF treatment increased the proportion of CD206-positive cells and reduced that of the CD86-positive cells. A, B Immunofluorescence staining of BV2 cells with M1 marker CD86 (A) and M2 marker CD206 (B) (green); C, D The percentage of CD86 and CD206 positive BV2 cells were assessed by flow cytometry. Scale bars = 50 μm. Data were performed as means ± SEM. *P < 0.05, **P < 0.01; ns, not significant. n = 3

RMF treatment inhibited TLR4/NF-кB/MAPK pathway in vitro

Similarly, the expressions of TLR4/NF-кB/MAPK signaling pathway-related proteins in vitro were also investigated by Western blot. The results were consistent with that in vivo. After 24 h of Aβ stimulation, the expression levels of TLR4, Myd88, p-IKKα/β, p-IкBα, p-NF-кB p65, p-JNK, p-p38, and p-ERK significantly increased, and their phosphorylation levels were significantly inhibited after RMF treatment for 2 h (Fig. 16).

Fig. 16.

RMF treatment inhibited the activation of NF-κB and MAPK pathways in vitro. A–I Expression levels of TLR4, Myd88, p-IKKα/β, IKKα/β, p-IкBα, IкBα, p-NF-кB p65, NF-кB p65, p-JNK, JNK, p-p38, p38, p-ERK, and ERK in BV2 cells were examined by Western blot and quantified with ImageJ software. Data were performed as means ± SEM. *P < 0.05, **P < 0.01, and ***P < 0.001; ns, not significant. n = 5

Discussion

Non-invasive treatments like magnetic fields for neurodegenerative diseases including AD have attracted the attention of researchers for decades because of their excellent safety and unique effects [65]. Nevertheless, the effects of RMF on AD pathology and their underlying mechanisms remain to be investigated. This study evaluated the effects of RMF on cognitive function in sporadic AD model mice and explored their underlying mechanisms. Sporadic AD, which accounts for the vast majority of AD (about 95%), usually develops after the age of 65 and is called late-onset AD. Environmental factors, including metal elements, play a crucial role in the pathogenesis of sporadic AD, such as Al. Al has been recognized for its ability to accelerate the aggregation of Aβ protein, induce neuronal degeneration, and impair spatial learning and memory functions in the brain [66]. The dosage of Al utilized in previous studies has varied, and cognitive impairment and hippocampal damage have been observed at different levels of dosage [67]. Cao et al. demonstrated that a dosage of 300 mg/kg AlCl₃ resulted in significant deficits in learning and memory, loss of dendritic spines, increased inflammatory factors, and decreased levels of immunomodulators and brain-derived neurotrophic factors. However, a dosage of 50 mg/kg did not exhibit a significant effect on cognitive function [68]. Considering the balance between successful modeling and animal survival time, a dosage of 100 mg/kg AlCl₃ was selected in this study based on preliminary experiments and existing literature [69]. On the other hand, D-gal, being poorly metabolized in the body, accumulates and leads to oxidative stress and inflammation in various tissues, including the brain [70]. Wu and Ryu et al. utilized a dosage of 120 mg/kg D-gal to establish an AD model in mice, which has been demonstrated to induce neural senescence and impair cognitive functions [53, 54]. In this study, a combination of Al and D-gal was employed to establish a mouse model that partially replicates the primary pathological characteristics of sporadic AD, including Aβ aggregation, neuronal degeneration, and cognitive impairment. The selection of 100 mg/kg AlCl₃ and 120 mg/kg D-gal dosages was based on previous research findings and the necessity to strike a balance between modeling success and animal survival time. The behavioral tests conducted in this study, such as the MWM and the NOR tests, unequivocally demonstrated that 3 months of RMF treatment significantly improved the spatial learning and memory abilities of AD mice. These findings not only validate the effectiveness of the chosen dosages of AlCl₃ and D-gal in inducing cognitive impairment but also support the evaluation of the therapeutic effects of RMF, consistent with prior research reporting improved spatial cognition in AD mice following exposure to magnetic fields [71, 72]. Additionally, the findings indicate that RMF treatment enhanced cognitive and memory functions, as well as hippocampus-cerebral cortex neuronal connections in AD mice [73].

Neuroinflammation, primarily driven by continuously activated microglia, is considered a key pathological feature of AD and an early event in its initiation [74]. It has been shown that the neuroinflammatory response of microglia is closely associated with the progression of several AD-related neuropathological changes, including Aβ deposition, tau pathology, neuronal and synaptic loss, and memory impairments [75, 76]. Western blot and immunofluorescence results in this study showed that the expression levels of M1-specific markers (Iba 1, CD86, and IL-6) were significantly decreased, and the expression levels of M2-specific markers (CD206 and TGF-β1) were significantly increased after RMF treatment. At the same time, the number of activated M1-type microglia was also significantly reduced after RMF treatment, while the percentage of M2-type microglia significantly increased. These results suggest that RMF can exert anti-neuroinflammatory effects by reducing the number of activated microglia and promoting their polarization shift.

Activated M1-type microglia play a key role in the neuroinflammatory response by activating the MAPK-NF-κB signaling pathway. TLR4 is expressed on the surface of microglia, which can be activated in both MyD88-dependent and independent pathways [77]. After identifying the pathogen, toll-like receptors (TLRs) can promote the rapid activation of the host's innate immunity by increasing the production of pro-inflammatory cytokines such as TNF-α, IL-6, and IL-23 and chemokines such as IL-1β, CCL2, and CCL8 [78]. The family of mitogen-activated protein kinases (MAPKs), which includes extracellular signal-regulated kinases (ERK), p38 MAPK, and c-Jun N-terminal kinases, is a collection of signaling molecules that is crucial for the expression of inflammatory cytokines. Activation of the members from the MAPK pathway is a key mediator of neuroinflammation, which plays an important role in neurodegenerative diseases [79]. Overactivated MAPK induces inflammatory signaling cascades in microglia, leading to NF-κB activation and expression of various pro-inflammatory cytokines such as IL-6, TNF-α, IL-1β, and iNOS [80]. TNF-α is associated with synaptic and neuronal damage, which ultimately leads to dementia [81]. IL-1β activates p38 and NF-κB pathways, further exacerbating neuronal cell dysfunction [82]. IL-6 is a cytokine that can activate the ERK/MAPK signaling pathway, resulting in the further development and maintenance of inflammation [83]. In contrast, by producing anti-inflammatory factors, including Arg-1, IL-10, and Ym1, M2-type microglia can reduce neuroinflammation [84, 85]. Nitric oxide (NO) generation is inhibited by Arg-1, which helps to reduce neuroinflammation [86]. IL-10 exhibits neuroprotective effects by preventing the production of pro-inflammatory cytokines, including IL-6, IL-8, and TNF-α [87]. Ym1 may antagonize inflammation and promote tissue repair by slowing leukocyte adhesion [88]. Thus, it is suggested that inhibition of the inflammatory M1 phenotype or promotion of the beneficial M2 phenotype may be helpful to AD treatment. Our results showed that after 3 months of RMF treatment, pro-inflammatory cytokines including IL-6, IL-1β, TNF-α, NO, and ROS were reduced in the serum and hippocampus of AD model mice, while anti-inflammatory cytokines Arg-1, TGF-β1, IL-10, and IL-4 were increased. In addition, in the Aβ-stimulated BV2 cell model, RMF also inhibited the production of pro-inflammatory cytokines and increased the levels of anti-inflammatory cytokines, like the findings in vivo. These results provide corroborating evidence that RMF can promote the M1 to M2 transition of microglia, thus alleviating the neuroinflammatory process in the AD model of mice.

After that, whole-transcriptome sequencing of RNA from mouse hippocampal tissue was performed. After enrichment by KEGG, it was found that differentially expressed genes in AD and AD + RMF group mice were mainly enriched in the MAPK signaling pathway, which was further validated by Western blot. Both in vivo and in vitro results showed that RMF treatment significantly inhibited the phosphorylation of p38, JNK, and ERK and suppressed the phosphorylation, degradation, and nuclear translocation of IKKα/β and IкBα, which in turn inhibited the activation of NF-κB and the expression of various pro-inflammatory cytokines including TNF-α and IL-1β.

Chronic neuroinflammation has been considered to be associated with decreased microglia phagocytic clearance and sustained accumulation of Aβ in the AD brain [89–91]. In response to inflammatory factors and Aβ, microglia transform into M1-type microglia and release pro-inflammatory factors, including TNF-α, IL-1β, and IL-6. TNF-α can upregulate the expression of microglia Aβ_1-42 chemokine receptors, enhance the chemotactic effect on Aβ, and promote plaque formation [92]. IL-1β plays a decisive role in the early formation of plaques, which enhances the synthesis and secretion of β-APP through the protein kinase C pathway and promotes the production and deposition of Aβ [93]. In addition, IL-1β can also induce microglia activation and proliferation, which further promotes the overgrowth of dystrophic neural axons and leads to neuroinflammatory plaque generation [94]. IL-6 is able to inhibit post-translational modification of amyloid precursors by activating α2 macroglobulin and triggering the deposition of pathogenic β4 amyloid in senile plaques [83]. However, in this study, the levels of these pro-inflammatory cytokines were significantly decreased after RMF treatment, suggesting that RMF may reduce amyloid deposition by inhibiting the production of pro-inflammatory cytokines. Besides, M2-type microglia are anti-inflammatory microglial cells, which are known to alleviate neuroinflammation through the production of anti-inflammatory factors, such as IL-4, IL-10, and Ym1 [84, 85]. It was reported that IL-4 activated TREM2, enhanced microglial autophagy, down-regulated TLR4, and blocked activation of the MAPK pathway, which can potentially enhance microglial phagocytosis, promote the clearance of Aβ, and ameliorate cognitive impairment [95]. In vivo results of this study showed that the number of M2 microglia increased, and they accumulated near Aβ deposits, while the Aβ plaques burden in the hippocampus of mice decreased after RMF treatment. In addition, the in vitro results showed the same tendency, most surprisingly of all, the phagocytic ability of BV2 cells was strongly enhanced after RMF treatment. Therefore, we hypothesized that the reduction of Aβ plaques in the hippocampus after RMF treatment might be associated with enhanced phagocytosis ability of microglia and the reduction of pro-inflammatory cytokines, which may be caused by the activation of M2 microglia.

However, the present study has several limitations. Firstly, the effect of RMF was only tested on female KM mice. Although women have a higher risk of AD compared to men due to female-specific factors [96–98], a comparative study utilizing male and female mice is necessary to assess gender-specific impacts. Secondly, an AD animal model was induced in this study using D-gal and AlCl₃. While this approach is straightforward and cost-effective, alternative sporadic AD mouse models may provide better options as the model employed in this study fails to accurately replicate the pathological alterations observed in sporadic AD patients. Additionally, the selected doses of AlCl₃ and D-gal were based solely on previous studies due to the lengthy modeling process, and the sample size was limited. Moreover, the expression of inflammatory cytokines in cultured BV2 cells did not fully resemble that in primary microglia. Furthermore, the current study only examined the effects of RMF on microglia-mediated inflammation, but future research should investigate whether RMF influences other types of brain cells, such as neurons, astrocytes, and vascular endothelial cells [60, 99].

Conclusions

In conclusion, this study provides novel insights into the beneficial effects of RMF on a sporadic AD mouse model, shedding light on the underlying mechanisms. Specifically, our findings demonstrate that RMF treatment significantly ameliorated spatial learning and memory deficits in the sporadic AD model while effectively mitigating synaptic and neuronal loss and reducing Aβ plaque accumulation. These effects are likely mediated through the suppression of neuroinflammatory responses via inhibition of the TLR4-Myd88-MAPK-NF-κB signaling pathway and the regulation of microglial polarization transition from M1 to M2 phenotype (Fig. 17). Our results highlight the potential of RMF as a promising therapeutic strategy for modulating inflammatory responses, preventing neurodegeneration, and attenuating chronic inflammation in sporadic AD and other neurodegenerative diseases.

Fig. 17.

A proposed mechanism that underlies the ameliorative effects of RMF on the cognitive and memory impairments of AD mice in this study. A AlCl₃ and D-gal stimulated inflammation by activating the TLR4-Myd88-MAPK-NF-κB signaling pathway, resulting in overactivated M1-type microglia and subsequent neuronal loss. B RMF exerted anti-inflammatory effects by inhibiting the TLR4-Myd88-MAPK-NF-κB signaling pathway and promoting the M2-type polarization shifting of microglia to play a neuroprotective role

Funding

This project was funded by the National Natural Science Foundation of China (NSFC) (No.81772002) and the Shenzhen Science Technology and Innovation Commission (No. JCYJ20170818143334365, JCYJ20230808104859039).

Declarations

Conflict of interest

The authors declare no competing interests.