Introduction

Alzheimer’s disease (AD) is the most prevalent neurodegenerative disorder and cause of dementia in the elderly. Its pathological hallmarks include the deposition of intercellular senile plaques and formation of intracellular neurofibrillary tangles in the brain [1]. Despite decades of research demonstrating that amyloid-beta (Aβ) is the main component of extracellular amyloid plaques [2], which are considered a crucial risk factor for AD pathogenesis [3], there is further need for effective treatments for AD. Regardless, the exact cause of AD remains unclear. Therefore, understanding the molecular bases of the transcriptome, epigenome, and epitranscriptome of AD pathogenesis holds promise for uncovering novel insights into the disease.

Emerging evidence suggests that the epigenetic dysregulation of N6-methyladenosine (m6A) modification plays a significant role in the pathogenesis of various neurodegenerative diseases, including AD [4]. As the most common reversible RNA modification, m6A methylation plays a crucial role in regulating gene expression and RNA metabolism [5,6,7,8]. In the brain, proper m6A modulation is essential for maintaining synaptic plasticity, neurotransmission, protein homeostasis, and neuronal function, contributing to neurological diseases [7,8,9,10]. However, the specific roles of m6A modification across various cell types are not fully understood. Recent studies demonstrate that m6A methylation is associated with AD progression. Transcriptome-wide profiling has revealed widespread changes in m6A methylation patterns in AD, particularly in transcripts involved in Aβ metabolism, Tau pathology, and microglial activation as well as compromised neuronal survival [11, 12]. Notably, m6A-modified transcripts are enriched in microglia-associated immune pathways, implicating m6A as a key regulator of microglial activation and neuroinflammatory response [13, 14]. Accordingly, m6A may contribute to AD pathology by modulating the microglia-mediated clearance of Aβ plaques and other cellular debris [15]. Although m6A dysregulation in microglia is associated with Aβ clearance and neuroinflammation, there is a lack of direct evidence demonstrating that m6A-modified transcripts specifically regulate the microglial phagocytosis of Aβ in AD-like conditions. Overall, these findings collectively suggest that m6A dysregulation may exacerbate AD progression [16, 17], although its precise impact requires further investigation.

Therefore, in this study, we elucidated the associations of abnormal m6A modifications with the pathogenesis and progression of early-stage AD. Accordingly, we utilized the APP/PS1 mouse model of AD and examined the mRNA m6A levels in the cortex and hippocampus during various stages of AD progression. We found that m6A mRNA modification was elevated in 6-month-old AD mice. To examine the changes in m6A modification in the hippocampus at the genome-wide level, we analyzed mRNA m6A methylation profiles in 6-month-old AD and wild-type (WT) mice by methylated RNA immunoprecipitation sequencing (MeRIP-seq). In addition, we experimentally validated the association between m6A modification and the gene expression of 2 key microglial genes, CD9 and Cebpβ, which are involved in microglia-mediated phagocytosis. Together, these findings provide direct molecular evidence linking m6A modification to specific changes in gene expression involved in Aβ clearance. These novel insights advance our understanding about the epitranscriptomic regulation of microglial functions in AD and may help identify new therapeutic targets for AD treatment.

Materials and methods

Animal ethics and genotyping

All animal experiments were approved and conducted in accordance with the regulations of the Animal Care Committee of the Shenzhen Institute of Advanced Technology. We used male, APP/PS1, heterozygous, double-transgenic mice (B6C3-Tg [APPswe, PSEN1dE9] 85Dbo/J; stock number 2010-0001) and WT littermates aged 3, 6, and 9 months. We prepared primary microglia from murine newborn ICR/CD1 mice on postnatal days 3–5. We purchased all mice from GemPharmatech and housed them in specific pathogen-free conditions with 50% humidity, a constant temperature of 23°C, and a 12/12-h light/dark cycle.

We conducted all experiments blinded to animal genotype, with 2 tail sections per animal prepared and examined as previously described [18].

Sample collection and RNA extraction

We collected whole cortical and hippocampal tissues from 3-, 6-, and 9-month-old AD and WT mice after dissecting fresh brains on ice in RNase-free conditions. We immediately placed the collected samples into a − 80°C freezer for subsequent analysis.

We extracted the total RNA from the above tissues using TRIzol (Thermo Fisher, MA, USA, #15596026) according to the manufacturer’s instructions followed by isopropanol precipitation. We stored the extracted total RNA at −80°C until use.

Dot blot assay

We purified and enriched mRNA from total RNA using a Dynabeads mRNA Direct Purification Kit (Thermo Fisher, MA, USA, #61011). We performed an m6A dot blot assay as previously described [19, 20]. In brief, we applied 100 ng denatured mRNA to each dot well in a Minifold I Dot-Blot System (Acrylic) (Capitol Scientific, TX, USA, #WHA-10447900) according to the manufacturer’s protocol. We crosslinked the membrane with denatured RNA samples in CL-1000 Ultraviolet Crosslinker (Spectrum, UK, #341-10262). We then blocked and incubated the crosslinked membrane with anti-m6A antibody (Synaptic Systems, Germany, #212B11). We subsequently incubated the membrane with secondary antibody and performed visualization using enhanced chemiluminescence (GE Healthcare, IL, USA, #45-002-401). We quantified dot intensity using ImageJ software (version 1.8.0).

Methylated RNA immunoprecipitation sequencing and identification of m6A peaks

We performed m6A MeRIP-seq as previously described [20]. We pooled total RNA extracted from the hippocampi of eight 6-month-old mice. We fragmented mRNAs into 100–200-bp fragments using RNA Fragmentation Reagent (Thermo Fisher, MA, USA, #AM8740). We conducted immunoprecipitation using rabbit polyclonal anti-m6A antibody (Synaptic Systems, Germany, #212B11) to pull down m6A-modified RNA followed by the addition of protein G Dynabeads (Thermo Fisher, Lithuania, #10009D) for the immunoprecipitation process. We constructed the cDNA library from RNA pull-down and input RNA using a NEBNext Ultra RNA Library Prep Kit for Illumina (BioLabs, UK, #E7530S).

We performed data visualization and sequencing analysis as previously described [19, 21, 22]. In brief, we subjected input data to quality control using FastQC (version 0.11.7) [23]. We trimmed the immunoprecipitation and input reads to remove 5′ and 3′ adaptor bases and aligned them to the reference genome using HISAT2 (v2.1.0) [24]. We then used the aligned reads to call peaks in the MeRIP regions. We identified statistically significant MeRIP-enriched regions (i.e., peaks) for each sample using a p-value threshold of 0.05 [25]. We used ExomePeak (version 2.13.2) to remove PCR artifacts and extended reads [26]. We identified and defined high-confidence m6A peaks on the basis of annotation to the Ensembl mm10 transcriptome (release 96) with the criteria of fold-change >2 and p < 0.05. Finally, we visualized the m6A peaks with the Integrated Genomic Viewer (IGV) Genome Browser [27, 28]. We performed m6A motif discovery using HOMER (version 4.7) with the parameter length set to 8 [29]. We classified m6A transcripts into 3 categories: (1) hypermethylated transcripts with significantly increased m6A peak intensity compared to the control group (i.e., adjusted p < 0.05 and log2[fold-change] > 1); (2) hypomethylated transcripts with significantly decreased m6A peak intensity compared to the control group (i.e., adjusted p < 0.05 and log2[fold-change] < −1); and (3) unchanged transcripts with no statistically significant differences in m6A peak intensity between the experimental and control groups (i.e., adjusted p ≥ 0.05 or log2[fold-change] between −1 and 1).

Functional pathway enrichment analysis

To analyze m6A-modified transcripts, we performed Gene Ontology (GO) enrichment analysis, Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, and Reactome gene set analysis of m6A-modified transcripts using Metascape (https://metascape.org/) with the following parameters: minimum overlap = 3, minimum enrichment = 1.5, and p < 0.05. We identified differential m6A modification enrichment patterns using Metascape with a threshold of p < 0.05 and fold-change >1.5 or <0.667. To elucidate the general organizing principles of functional cellular systems and predict protein interactions, we constructed a protein–protein interaction (PPI) network using the STRING database (http://STRING-db.org/) and GeneMANIA (http://genemania.org/). We performed modular analysis and visualization of the PPI network using GeneMANIA with the default parameters as follows: degree cutoff = 2, node score cutoff = 2, K-score = 2, and maximum depth = 100.

Real-time quantitative PCR

To analyze the expression levels of genes with m6A modifications, we employed the real-time quantitative PCR (RT-qPCR). We performed cDNA synthesis using a RevertAid H Minus First Strand cDNA Synthesis Kit (QIAGEN, Germany, #K1632). We analyzed the RT-qPCR data using a CFX96 Touch Real-Time PCR Detection System (Bio-Rad, CA, USA, #CFX96). We designed gene-specific primers using Primer3 (version 0.4.0) (Table 1) [30]. We normalized relative mRNA expression levels to the averages of the housekeeping gene, GAPDH. We performed statistical analysis of gene expression using the comparative 2−ΔΔCT method.

Table 1 List of primers used in RT-qPCR.
Full size table

Isolation and culture of microglia

After anesthesia with pentobarbital, we isolated primary microglia from ICR mouse pups at postnatal days 3–5. We collected and dissected the cortices and minced them into 2–3-mm pieces using a sterile scalpel or scissors. To obtain single-cell suspensions, we dissociated the minced cortices using a papain dissociation system (Worthington, NJ, #33E23641). We then isolated microglia and purified them by Percoll density gradient centrifugation [31]. We seeded the purified cells onto poly-l-lysine–coated chamber slides (Sigma-Aldrich, Germany, #A-005-C) and cultured them in DMEM supplemented with 10% FBS and 1% penicillin–streptomycin. We incubated the cells at 37 °C in a 5% CO2 atmosphere at 100% humidity for 7–8 days for further use.

Short hairpin RNA construction and cell transfection

The short hairpin RNAs (shRNAs) against CD9 and Cebpβ were synthesized by Life Technologies (CA, USA) (Table 2) and then cloned separately into pAAV vector. We transfected primary microglia (cultured for 7–8 days) using Lipofectamine 3000 according to the manufacturer’s instructions. Briefly, we seeded the cells onto 8-well chamber slides at 25,000 cells per cm2 and transfected them with vectors at 400 ng per well supplemented with P3000 at 0.8 μL per well and Lipofectamine 3000 at 1 μL per well.

Table 2 list of siRNA used in transfection.
Full size table

Knockdown efficiency validation and cell viability assay

We assessed the knockdown efficiency after shRNA transfection by western blot analysis. We transfected primary microglia with shRNAs against CD9 and Cebpβ, and harvested the cells after 48 h. We prepared cell lysates using RIPA buffer, quantified them by BCA assay, and subjected them to SDS-PAGE following PVDF membrane transfer. We incubated membranes with anti-Cebpβ (Abcam, UK, #AB32358) or anti-CD9 (Cell Signaling Technology, MA, USA, #98327S) and β-actin (Cell Signaling Technology, MA, USA, #4970S) followed by HRP-conjugated secondary antibodies. We detected the membranes using an enhanced chemiluminescence system (Thermo Fisher, USA, MA, #32209).

We assessed cell viability after shRNA transfection by MTT assay. Briefly, we incubated cells transfected with shRNAs against CD9 and Cebpβ in MTT solution (5 mg/mL) for 4 h and dissolved formazan crystals in DMSO. We measured absorbance at 570 nm. Viability was expressed as a percentage of the control.

Microglial phagocytosis assay

We assessed the phagocytic activity of microglia by measuring the uptake of PE-labeled Zymosan beads (Novus Biologicals, CO, USA, NBP2-54860). Briefly, we cultured primary microglia on an 8-well chamber slide at 25,000 cells per cm2 for 7–8 days. For cell stimulation, we prepared Aβ oligomers from ultrapure fluorescein Aβ (1-42) peptides (rPeptide, GA, USA, #A-1119) to a final concentration of 100 μM according to the manufacturer’s instructions. We transfected the cells with the candidate shRNAs for 24 h followed by stimulation with either 1 μg/mL LPS or 0.5 μM fluorescein Aβ for 24 h. After stimulation for 24 h, we co-incubated the microglia from the control group (treated with PBS) and the LPS group with overloaded PE-labeled Zymosan beads for 3 h in an 8-well chamber slide.

To visualize the uptake of PE-labeled Zymosan beads, we washed away the floating PE-labeled Zymosan beads using DPBS and then fixed the microglia with Cytofix fixation buffer (BD Pharmingen, CA, USA, #554655) at room temperature according to the manufacturer’s instructions. We obtained fluorescent images using a Zeiss LSM880 scanning confocal microscope with a 20× immersion objective.

Statistical analysis

All data are presented as mean ± SEM. We performed statistical comparisons using Student’s t test and 2-way ANOVA in GraphPad Prism 9. The levels of significance were as follows: *p < 0.05, **p < 0.01, and ***p < 0.001. The statistical results shown in the figures are derived from at least 3 independent experiments. We conducted each experiment with biological replicates to ensure reliability and reproducibility of the data. We included technical replicates within each experiment to account for any variability in the experimental procedure.

Results

Epitranscriptomic m6A modifications in Alzheimer’s disease mice

To investigate the changes in m6A modification during AD progression, we first quantified the m6A levels in mRNA samples extracted from the cortices and hippocampi of both WT and AD mice at various ages by dot blot assay. We observed a significant increase in m6A level in the hippocampus in 6-month-old AD mice (Fig. 1A, B). To elucidate the role of m6A-mediated epitranscriptome modifications in AD, we performed m6A-specific MeRIP-seq on hippocampal tissues from 6-month-old WT and AD mice (Fig. 1C). We analyzed the distribution of m6A peaks in the hippocampus and identified them within the transcriptome. In WT mice, the m6A peaks were most abundant in coding sequences (47.45%) and untranslated regions (5′-UTR: 13.09%, 3′-UTR: 18.71%) followed by exons (20.75%) (Fig. 1D). In AD mice, the m6A peaks were most abundant in coding sequences (42.04%) and untranslated regions (5′-UTR: 11.59%, 3′-UTR: 26.8%) followed by exons (19.57%) (Fig. 1E). Further analysis of m6A methylation peaks in the hippocampi of AD and WT mice revealed enrichment of m6A peaks in the 3′- and 5′-UTRs in both groups (Fig. 1F). Notably, the AD mice exhibited larger m6A peaks in the CDS and 3′-UTRs than the WT mice (Fig. 1F). The distribution of m6A reads in the hippocampus differed between the WT and AD mice, indicating disease-related changes in m6A peaks within coding sequences and UTRs in the AD mice.

Fig. 1: Alterations in m6A modification of the epitranscriptome in Alzheimer’s disease model mice.
figure 1

A Dot blot assay results of global mRNA m6A abundance in total RNA from the hippocampi and cortices of 3-, 6-, and 9-month-old Alzheimer’s disease (AD) mice (i.e., APP/PS1 mice) and wild-type (WT) mice. B Quantification of 3 independent dot blot experiments, normalized to the average m6A modification levels in WT mice (n ≥ 10 from 3 independent experiments). C Experimental design and flowchart for methylated RNA immunoprecipitation sequencing (MeRIP-seq). D, E Percentages of m6A peaks in the AD and WT mice. F Distribution of m6A peaks across mRNA regions in the hippocampus in the AD (red) and WT (blue) mice (n = 8 mice per group from 1 independent experiment). Values are the mean ± SEM of 3 technical measurements from 2 biological replicates for m6A qualitative analysis and quantification. We performed statistical comparisons using Student’s t test (n ≥ 10; *p < 0.05).

Full size image

Changes in m6A modulation alter cellular function in Alzheimer’s disease mice

To investigate disease-associated m6A alterations in the hippocampal epitranscriptome, we analyzed the m6A modification profile of the hippocampus in both the WT and AD mice. Accordingly, we identified 2039 hypermethylated (i.e., p < 0.05, log2[fold-change] > 1) and 1374 hypomethylated (i.e., p < 0.05, log2[fold-change] < −1) m6A-modified transcripts in the AD mice compared to the WT mice (Fig. 2A). GO analysis of m6A peak-modified transcripts revealed significant enrichment in biological functions critical to immune responses, including “regulation of cellular progresses” and “metabolic processes,” “response to stimuli,” “immune system processes,” and “homeostasis” (Fig. 2B). These findings suggest that m6A modifications play an important role in regulating immune cell activation in the AD brain, potentially influencing disease progression through an inflammatory response mechanism. Further “cellular component” analysis indicated that the m6A-modified transcripts (i.e., p < 0.05, fold-change < 0.667 or >1.5) were enriched in genes associated with neuronal and immune-related structures, including ”post- and pre-synapse,” “axon,” “cell body,” “GABA-ergic synapse,” “cell leading edge,” “transport vesicle,” and “perinuclear region of cytoplasm” (Fig. 2C). These findings suggest potential roles of m6A modification in the regulation of synaptic function, cellular communication, and immune cell interactions in the AD brain. In addition, “molecular function” analysis showed that the m6A-modified transcripts were notably involved in pathways critical for immune activity in the brain, including “kinase binding,” “cell adhesion,” “molecule binding,” “molecular adaptor activity,” “protein domain specific binding,” “GTPase regulator activity,” “phospholipid binding,” and “kinase activity” (Fig. 2D). Of note, these functions are essential for immune cell signaling, neuroinflammation, and glial function in neurodegenerative diseases. Next, we conducted motif analysis of m6A peak-modified transcripts, comparing the AD and WT mouse brains (Fig. 2E). The results revealed differences in the m6A motif sequence in the hippocampus in the AD mice (Fig. 2E). These differences suggest a regulatory mechanism that modulates gene expression and contributes to molecular processes implicated in AD pathogenesis [32].

Fig. 2: Impacts of m6A modulation on gene expression and cellular function in Alzheimer’s disease.
figure 2

A Volcano plots showing differentially methylated m6A peaks identified through de novo detection. Horizontal and vertical lines show the scoring thresholds (i.e., −4, −2, 0, 2, and 4). B Gene Ontology (GO) analysis of “biological process” revealed the top 20 enriched pathways in the hippocampus in Alzheimer’s disease (AD) mice (i.e., APP/PS1 mice) compared to that in wild-type (WT) mice (p < 0.05, fold-change < 0.667 or >1.5). C GO analysis of “cellular component” identified the top 20 enriched pathways related to m6A-modified genes in the hippocampus in the AD mice compared to that in the WT mice (p < 0.05, fold-change < 0.667 or >1.5). D GO analysis of “molecular function” identified the top 20 enriched pathways for m6A-modified genes in the hippocampus in the AD mice compared to that in the WT mice (p < 0.05, fold-change < 0.667 or >1.5). E De novo motif discovery identified m6A motifs in the hippocampus in the AD and WT mice. (n = 8 per group, pooled).

Full size image

Identification of m6A-modified transcripts associated with immune features in Alzheimer’s disease mice

To further explore the functional impacts of m6A modification on AD, we performed Reactome pathway enrichment analysis on differentially methylated m6A peaks in the hippocampi of AD and WT mice (Fig. 3). The results revealed that the m6A-modified genes (i.e., p < 0.05, fold-change < 0.667 or >1.5) were enriched in pathways associated with neuronal function and brain disorder regulation, including “neuronal system function,” “protein–protein interactions (PPIs) at synapses,” “nervous system development,” “membrane trafficking,” “axon guidance,” and “MAPK family signaling cascades” (Fig. 3A). In addition to neuronal pathways, several immune-related pathways were enriched, including “Rho GTPases,” “micro GTPases and RHOBTB3,” “adaptive immune system,” and “diseases of signal transduction by growth factor receptor (GFR)” (Fig. 3A). These findings suggest that m6A modification plays potential roles in modulating immune regulation and neuronal function in AD pathology [15]. Further cell-type-specific analysis showed distinct patterns of m6A methylation in microglia, various neuronal subtypes, and astrocytes (Fig. 3B), highlighting the involvement of m6A in immune regulation within the AD brain. To further understand the functional impact of m6A RNA modification on neuroinflammation, we analyzed m6A-modified genes associated with immune function in microglia-enriched transcripts. GO enrichment analysis identified significant immune-related pathways, including “VEGFA–VEGFR2 pathway,” “cytokine signaling in immune system,” “hemostasis,” “MAPK family signaling cascade,” and “MHC antigen presenting and death receptor signaling” (Fig. 3C). In addition, 16 m6A-modified immune genes were strongly associated with inflammatory cytokine activation and phagocytosis (Fig. 3D). Of note, the m6A-modified transcripts of Sirpa, CD47, Cebpβ, and CD9, which are key regulators of phagocytic activity in antigen-presenting cells, were involved in “Fc gamma receptor-dependent phagocytosis,” “positive regulation of cell adhesion,” and “cell–cell communication” (Fig. 3D). These results collectively highlight the roles of m6A modifications in modulating microglial activation, phagocytic functions, and immune responses in AD pathogenesis.

Fig. 3: Alteration of m6A modification is involved in inflammatory response in Alzheimer’s disease pathogenesis.
figure 3

A Reactome enrichment analysis of high-confidence, m6A-modified transcripts in the mouse hippocampus (p < 0.05, fold-change < 0.667 or <1.5). B Reactome enrichment analysis of m6A-modified transcripts by cell type (p < 0.05, fold-change < 0.667 or >1.5). C Network of the top 20 Gene Ontology terms and KEGG pathways focusing on the adaptive immune system (p < 0.05, fold-change < 0.667 or >1.5). D Heatmap of enriched KEGG pathways and Gene Ontology terms in microglial clusters. We used Fisher’s exact test for peak calling (p < 0.05, fold-change < 0.667 or >1.5).

Full size image

Changes in m6A methylation regulate microglial phagocytic activity

To elucidate the roles of m6A modification in immune regulation and Fc-gamma receptor-dependent phagocytosis, we examined m6A-modified immune-related genes in the hippocampi of WT and AD mice. Using STRING 10, we constructed a functional PPI network. The analysis identified 3 main network nodes representing the central interactions among Cebpβ, CD9, CD47, and Sirpa in the Fc-gamma receptor-mediated phagocytic pathway (Fig. 4A). Next, we employed GeneMANIA to analyze the 4 central immune-responsive genes within the interaction network. The results indicate that these m6A-modified genes are highly involved in phagocytosis, including “cell adhesion molecule binding,” “negative regulation of cell–cell adhesion,” “negative regulation of cell activation,” and “regulation of cell–substrate adhesion” (Fig. 4B). These findings indicate that m6A RNA modifications play a critical role in modulating microglial phagocytosis in the AD brain, potentially influencing neuroinflammation and disease progression.

Fig. 4: Roles of m6A modifications in microglial phagocytic activity and immune response regulation.
figure 4

A Functional protein–protein interaction (PPI) network of m6A modification-related adaptive immune system in Alzheimer’s disease (AD) mice (i.e., APP/PS1 mice) compared to wild-type (WT) mice. Different colors represent network nodes. B We used GeneMANIA analysis to identify genes with similar functions to the 4 hub genes in (A) and construct an interactive functional association network. Interactive functional association networks show connections with the regulation of coagulation and circulation.

Full size image

Alteration of gene expression levels m6A-modified transcripts in Alzheimer’s disease mice

Among the m6A-modified transcripts of the phagocytic pattern, we examined 4 representative genes: Cebpβ, CD9, CD47, and Sirpa. We confirmed the existence of m6A binding peaks in the transcripts of these genes. IGV visualization analysis revealed significant enrichment of m6A peaks specifically in the AD mice (Fig. 5A–D). To evaluate the impacts of AD-mediated m6A modification on the expression of Cebpβ, CD9, CD47, and Sirpa transcripts, we performed gene expression analysis by RT-qPCR. Accordingly, hypermethylation of Cebpβ, CD9, and CD47 was associated with increased gene expression in the AD mice (Fig. 5E). These findings suggest that m6A modifications play a regulatory role in phagocytic gene expression during AD pathogenesis.

Fig. 5: Changes in m6A modifications affect gene transcription levels.
figure 5

A–D IGV Genome Browser views of m6A distribution and abundantly enriched peaks in Cebpβ, CD9, CD47, and Sirpa transcripts. E RT-qPCR quantification of Cebpβ, CD9, CD47, and Sirpa expression in the cortex and hippocampus in 6-month-old Alzheimer’s disease (AD) mouse models (i.e., APP/PS1 mice). Values are the mean ± SEM of 3 technical measurements from 2 biological replicates for gene expression. (n ≥ 8; Student’s t test, *p < 0.05, **p < 0.01).

Full size image

Suppression of Cebpβ and CD9 impairs microglial phagocytosis

To gain deeper insights into the regulatory effects of the candidate m6A-modified genes, we examined the functions of 2 representative genes, Cebpβ and CD9, in cultured primary microglia in vitro. We initially confirmed that the shRNA approach effectively silenced gene expression without affecting the viability of either shCebpβ- or shCD9-transfected cells (Supplementary Figs. 1 and 2). Specifically, we investigated the impacts of Cebpβ and CD9 on microglial phagocytosis by using shRNAs to target individual transcripts and tracking the uptake of PE-labeled Zymosan beads or Aβ fluorescence according to immunofluorescence intensity (Fig. 6A, B). We quantified the fluorescence intensity of PE-labeled Zymosan and Aβ fluorescence uptake for each individual transcript following LPS or Aβ stimulation. Accordingly, suppression of either Cebpβ or CD9 significantly reduced microglial phagocytic activity following stimulation with either LPS or Aβ (Fig. 6C). These findings suggest that m6A modification plays a crucial role in facilitating microglia-mediated phagocytosis and clearance in AD pathogenesis.

Fig. 6: Downregulation of Cebpβ or CD9 inhibits microglial phagocytosis.
figure 6

A Primary microglia were transfected at 7 days in vitro (DIV) with pAAV plus shCD9 or shCebpβ. Confocal images (20×) were collected at 9 DIV. Scale bar: 50 μm. Representative images of Iba1 (green), Zymosan (red), fluorescence-labeled Aβ (purple), and DAPI (blue) in transfected primary microglia. B Colocalization analysis of fluorescence intensity for PE-labeled Zymosan bead uptake in the PBS and LPS groups or fluorescence-labeled amyloid-beta (Aβ) uptake in the Aβ treatment group following knockdown of candidate genes. C Quantitative analysis of normalized fluorescence intensity for either PE-labeled Zymosan beads or fluorescence-labeled FITC-Aβ (from 3 independent experiments; 2-way ANOVA, *p < 0.05, **p < 0.01).

Full size image

Discussion

As the first confirmed dynamic and reversible mRNA modification [33, 34], m6A is highly abundant in the brain [10, 35]. This modification plays critical roles in neurogenesis, neurodevelopment, immune response, and neurological disorders [5, 10, 19]. In this study, we demonstrated that m6A modification serves as an epitranscriptomic signature during the progression of AD. Epitranscriptomic profiling of mRNA m6A modifications in the hippocampus in APP/PS1 mice revealed abnormal m6A methylation patterns associated with immune system processes, potentially modulating inflammatory immune responses. We further investigated the impacts of suppressing the m6A-modified candidate genes, Cebpβ and CD9, in vitro, examining changes in microglia-mediated phagocytic activity following either LPS or Aβ stimulation. These results underscore the essential role of m6A alterations in modulating microglial function and immune response in AD pathogenesis.

Previous studies on epitranscriptomic regulation have revealed alterations in m6A modification patterns and their effects on cellular dysfunction Notably, these modifications are abundant in the adult mammalian nervous system and tend to increase with aging and in disease conditions [36, 37]. Moreover, aberrant m6A signaling is associated with multiple brain deficits, including impairment of synaptic transmission, spatial learning and memory, synaptic plasticity, and axonal guidance as well as reduced brain volume and tumorigenesis in glioma malignancy [38,39,40]. These findings collectively underscore the importance of m6A-mediated regulatory mechanisms in neurological diseases. Elevated m6A modification is linked to neuroinflammation and the accumulation of pathological proteins, including Aβ aggregation and tau tangle formation, which are crucial markers of AD [21, 41]. Nevertheless, research on the contributions of m6A modifications to AD remains limited.

In the current study, we observed that mRNA m6A modifications were highly enriched in the hippocampus in 6-month-old AD mouse models during AD progression. The accumulation of m6A methylation in the brain is reported to induce widespread alterations in the expression of genes related to neuronal function, synaptic plasticity, and immune responses in AD mouse models [42]. Our results highlight the m6A-modified transcripts in AD mice in multicellular organismal processes and metabolic processes, including cellular communication, immune responses, inflammatory pathways, tissue homeostasis, and energy metabolism, which are crucial for maintaining overall cell health and function. These results indicate that alterations in m6A-modified transcripts impact neuronal and immune signaling by altering the stability and translation of key transcripts involved in AD pathogenesis. The relationship between m6A modification and AD progression is complex. Certain AD risk genotypes are linked to increased m6A abundance [43, 44]. Meanwhile, other studies suggest that reducing m6A modification can attenuate neuroinflammatory responses [45]. Disrupted m6A homeostasis affects mRNA stability and the translation of genes critical for both immune responses and neuronal function, corroborating its regulatory role in AD [5, 40]. Differentially methylated transcripts involved in immune responses are highly enriched in microglia. Furthermore, single-cell and microglia-specific m6A profiling confirm that these disease-associated m6A signatures are predominantly localized to microglia [14, 46, 47]. This altered m6A methylation reshapes microglial function, promoting a disease-associated phenotype that contributes to neurodegeneration [14]. While prior studies link m6A dysregulation to immune pathways, our work is the first to directly associate m6A modification with microglial phagocytic activity in AD.

Microglia, the brain’s resident immune cells, play crucial roles in initiating defense responses against neuroinflammation in AD [48,49,50]. Recent evidence indicates that abnormalities in RNA modification, particularly m6A methylation, contribute to neurotoxicity and neuroinflammation, driving AD pathogenesis and progression [42]. Our study demonstrates that alterations in m6A modification are linked to the activation of the immune response in microglia, the regulation of microglia-mediated phagocytic activity, and the promotion of inflammatory cytokine-induced inflammation. In brain diseases, activated microglia release cytotoxic mediators, including pro-inflammatory cytokines and chemokines [51]. These factors trigger a robust inflammatory response, increasing the number and activity of phagocytic microglia to clear apoptotic cell fragments and other harmful substances, thereby contributing to AD progression [14, 49, 51]. Conversely, decreasing the number of microglia or inhibiting microglial phagocytic activity reduces early synapse loss and neuronal death, which are crucial components of neuroinflammation [52]. Here, we observed that m6A-modified transcripts are enriched in pathways related to inflammatory cytokine signaling and microglial activation. Notably, the expression of these modified genes correlates with chemotaxis and cell migration. Microglia secrete various chemokines and receptors to promote phagocytosis in different brain disease models. Therefore, the enrichment of m6A modifications in AD risk genes within microglia highlights the critical role of immune mechanisms in AD development and progression [53,54,55].

We identified 2 m6A-modified genes, Cebpβ and CD9, which are involved in the phagocytic regulation of antigen-presenting cells [56, 57]. These genes are highly expressed in microglia and are relevant across multiple cell types in the brain. Cebpβ, a member of the CCAAT-enhancer-binding protein (Cebp) family, is highly expressed in myelomonocytic cells and antigen-presenting cells, including microglia in the central nervous system. It regulates cellular functions such as proliferation, differentiation, and immune response coordination [56, 57]. Cebpβ is involved in neuroinflammation by regulating Aβ deposition, inflammatory cytokine production, synaptic dysfunction, and neuronal cell loss [58]. In the AD brain, it is upregulated [59, 60] and exhibits enhanced interaction with the NF-κB p65 subunit, leading to increased transcription of proinflammatory cytokines [61]. Inhibition of the Cebpβ–NF-κB pathway reduces inflammatory mediator-induced neuronal degeneration and decreases Aβ secretion [61]. Furthermore, Cebpβ knockout diminishes Aβ pathologies in AD mouse models, whereas overexpression of Cebpβ accelerates AD pathologies [59, 62]. Therefore, the upregulation of m6A-modified Cebpβ may exacerbate AD progression. Accordingly, our results suggest that inhibition of Cebpβ expression impairs the activation of microglia by suppressing their phagocytic activity.

Similarly, CD9 (TSPANA29), a member of the tetraspanin family, plays critical roles in the immune system by regulating cell adhesion, migration, antigen presentation, phagocytosis, and inflammatory responses. CD9 contributes to immune surveillance and clearance by helping immune cells recognize and engulf pathogens and apoptotic cells [63, 64]. A recent study identified CD9 as a potential marker of disease-associated microglia in an age-dependent manner along disease progression in AD mouse models [65]. The CD9-specific program in AD in both mice and humans involves elevated phagocytosis of Aβ [66, 67]. Therefore, we examined whether CD9-mediated microglial activation is associated with enrichment of intracellular phagocytic particles. Accordingly, our findings show elevated CD9 gene expression in the brain in AD mice. Moreover, our in vitro results show that CD9 knockdown led to decreased primary microglial activation and reduced phagocytic activity, highlighting the negative association between CD9 expression and microglial phagocytosis in AD.

Conclusion

In summary, our study provides novel insights into the epitranscriptomic landscape of AD by demonstrating that m6A modification plays a functional role in AD progression and pathogenesis. First, we show that alterations in m6A modification are not only broadly associated with crucial pathological processes, such as metabolic dysfunction, disruption of synaptic transmission, and immune activation, but also specifically influence gene expression and phagocytic function in microglia. Notably, we identified m6A-modified transcripts, including CD9 and Cebpβ, which are closely linked to microglia-mediated Aβ clearance, suggesting a direct regulatory mechanism through which m6A impacts neuroinflammation in AD. Nevertheless, further research is required to fully elucidate the molecular mechanisms through which m6A modification of mRNA regulates neuroinflammatory processes in AD. As such, future studies will provide additional epigenetic evidence to elucidate the role of m6A modification in AD pathogenesis.