DNA methylation, a process in which a methyl group is added to the C5 position of cytosine, thereby forming 5-methylcytosine [1], controls gene expression by attracting proteins that suppress gene transcription or preventing the attachment of transcription factors [1]. During development, the DNA methylation pattern in the genome undergoes changes via dynamic processes, including de novo DNA methylation and demethylation. After development, differentiated cells acquire persistent and distinct DNA methylation patterns that regulate the transcription of specific genes in each tissue [2]. However, the patterns of DNA methylation can change following development, a process possibly influenced by cellular activity or the microenvironment [3]. In the central nervous system, particularly in the hippocampus, precise control of DNA methylation is crucial for normal learning and memory function. DNA methylation changes due to congenital mutations or acquired environmental factors can lead to various psychiatric and neurological disorders, often causing mental impairment [4].

Alzheimer’s disease (AD) is the most prevalent neurodegenerative disorder with symptoms such as deterioration in memory, cognitive, and executive abilities, as well as alterations in personality [5]. The primary pathological features of AD are the accumulation of amyloid beta (A$\beta$) plaques and formation of neurofibrillary tangles (NFTs) that occur throughout the brain regions, particularly in the cerebral cortex and hippocampus [6, 7]. The accumulation of A$\beta$ and tau proteins disrupts the transmission of signals and substances between neurons, ultimately resulting in neuronal death [8]. Recent studies have demonstrated that not only genetic factors but also epigenetic imbalance may attribute to the abnormal expression of genes associated with synaptic plasticity and cognition in AD [9, 10, 11]. However, the relationship between DNA methylation in genetic control studies of AD and the gross pathology of AD remains unclear [12, 13, 14].

In this study, we investigated the integrated analysis of global DNA methylation and gene expression in the hippocampi of 5xfamiliar Alzheimer’s disease (5xFAD) transgenic mice and explored the role of epigenetically regulated genes in the development and neuropathology of AD.

In this study, we performed an in-depth combined analysis of DNA methylation and gene expression in the hippocampal tissues of 5xFAD tg mice, a mouse model of AD. Cognitive impairment, including working and reference memory, was detected in 9-month-old, but not in 3-month-old, 5xFAD tg mice, and the age-dependent memory deficits are similar to the behavioral phenotypes reported in other studies using the Morris water maze and contextual fear conditioning paradigms [30, 31]. Furthermore, A$\beta$ production was significantly increased in the hippocampus of 9-month-old 5xFAD tg mice. Therefore, we selected the hippocampi of 9-month-old wild-type and 5xFAD mice as samples to confirm DNA methylation-regulated gene expression. A total of 13,777 DMRs, including 4484 hyper- and 9293 hypomethylated regions, were identified using DNA methyl-seq, and 101 DEGs, including 52 up- and 49 downregulated genes, were identified using RNA-seq analysis. Although hypomethylated regions were much more abundant than the hypermethylated regions, the distribution of hypomethylation in genes was largely in non-functional regions, which might have affected similar numbers of up- and downregulated genes. Finally, in the combined analysis, three genes, namely Thy1, Myo6, and Filip1l, were selected as candidates that potentially play epigenetic roles in the progression or aggravation of AD symptoms and histopathological lesions.

Given that acquired environmental factor or aging influences the development of neurological illnesses, previous studies have employed epigenome-wide association studies to identify changes in DNA methylation linked to the characteristics of diseases [32, 33]. Over the past few decades, the significance of epigenetic pathways in the development of neurological disorders has been increasingly recognized [32]. AD, marked by a gradual deterioration in cognitive function and neuronal death, is the most prevalent of the late-onset neurodegenerative diseases [34]. Increasing evidence emphasizes the involvement of epigenetic variation in AD development since gene mutations account for only approximately 5% of all patients with AD [35]. Aging is the primary risk factor for the development of AD; recent studies have indicated that both aging and AD are linked to significant alterations in the epigenetic control of gene expression, particularly DNA methylation [36, 37]. DNA methylation analysis has been conducted in several AD mouse models. Cong et al. [38] investigated genome-wide CpG methylation in the cortex of 11-month-old APP/PS1 mice, showing that 2346 CpG sites involving 485 genes could be critically associated with AD. The study suggested that epigenetic alternation of the gene for transforming growth factor beta 1 (TGF-$\beta$1) may have a pathological role in the early stages of AD [38]. Moreover, Sanchez-Mut et al. [14] analyzed DNA methylation changes in the frontal cortex of 12-month-old APP/PS1 and 18-month-old 3xTg mice through a genome-wide promoter DNA methylation array. Three hypermethylated candidate genes (thromboxane A2 receptor (TBXA2R), sorbin and SH3 domain-containing 3 (SORBS3) and spectrin beta 4 (SPTBN4)) were identified via an integrative analysis of DNA methylation changes and confirmed in the two mouse models and in AD patients, suggesting that cAMP response element-binding protein (CREB) signaling and the axon initial segment, which are associated to the three silenced genes, could contribute to the pathobiology of AD [14]. Herein, we identified three candidate genes related to the symptoms and pathological progression of AD through a combined analysis of DNA methyl- and RNA-seq in the hippocampus of 9-month-old 5xFAD tg mice. First, Thy1 encoding DNA was hypomethylated, and mRNA expression was upregulated in the hippocampus of 5xFAD tg mice. Thy1 is a highly conserved glycoprotein connected to glycophosphatidylinositol and found on the surface of several cell types such as thymocytes, fibroblasts, endothelial cells, mesangial cells, and neurons [39]. As regulators of cell-cell or cell-matrix interactions, the physiological effects of Thy1 vary based on cell and tissue type [39]. Thy1, mainly expressed by neurons in the brain, makes up 2–7% of the total surface protein, making it one of the most common surface glycoproteins in the nervous system [40]. Thy1 is specifically found at the ends of axon terminals in neurons and induces actin cytoskeleton contraction, resulting in neuronal retraction [41, 42]. Neurite outgrowth is important for the treatment of brain injury and neurodegenerative diseases, including AD [43]. In addition to its effects on neurite outgrowth, Thy1 is also directly associated with synaptic physiology. The absence of Thy1 in mice disrupts long-term potentiation in the dentate gyrus under specific conditions [44], suggesting that Thy1 plays a role in the regulation of synaptic neurotransmitter release [45, 46]. Hence, we propose that epigenetically increased expression of Thy1 in the hippocampus of 5xFAD tg mice may have detrimental effects on axonal regeneration and synaptic transmission associated with the development and progression of AD.

The second DNA methylation-regulated gene confirmed in this study was Myo6, which was hypermethylated; its mRNA levels decreased in the hippocampus of 5xFAD tg mice. Myo6, unlike other myosins, exhibits atypical retrograde movement toward the minus point of actin filaments [47] and is abundantly expressed in the brain, particularly at synapses [48]. Osterweil et al. [48] discovered that the hippocampi of mice lacking Myo6 showed notable impairments in stimulation-induced endocytosis of the alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid receptor (AMPAR, a glutamate receptor). Furthermore, the phosphorylation of Myo6, which is associated with mitochondrial recruitment and presynaptic filament anchoring, is involved in energy supply and calcium removal in the presynaptic region during intense synaptic activity [49]. Deficits in this process induce local energy depletion and intracellular calcium overloading. Consequently, deficiency in Myo6 expression levels or function leads to unusual dendritic spines, synapse reduction, impairment of synaptic efficacy, long-term depression, and astrogliosis [48], all of which are characteristics of hippocampal lesions in patients with AD [50]. Therefore, we propose that epigenetic downregulation of Myo6 expression can affect synaptic plasticity, which is associated with pathological lesions and impaired learning and memory in AD.

The last methylation-affected gene investigated in this study was Filip1l, whose DNA was hypermethylated, and mRNA level was decreased in 5xFAD tg mice. To date, Filip1l has been studied in the field of cancer and reported to be associated with epithelial-mesenchymal transition, chemoresistance, and extracellular matrix migration in cancer cells [51]. Filip1l is highly similar to Filip1, with ~50% amino acid similarity [52, 53]. The intracellular location and functions of Filip1l are similar to those of Filip1, suggesting that Filip1l can be a Filip1 family protein that binds to filamin A and induces the degradation of filamin A, an actin-binding protein [53]. Therefore, Filip1l may exert its actions via the inhibition of filamin A. Filamin A modulates the rearrangement of the actin cytoskeleton by interacting with integrins, transmembrane receptor complexes, and secondary messengers [54]. Mutations in filamin A generally lead to aberrant neuronal migration and connective and vascular tissue abnormalities [55]. Followed by senile A$\beta$ plaques and neurofibrillary tangles (NFTs), a recent study suggested that a third proteopathy, an altered conformation of the scaffolding protein filamin A, could be intimately connected to the amyloid and tau pathologies in AD [56]. Changes in filamin A function facilitate the continuous activation of toll-like receptor 4 by A$\beta$ plaques, thereby inducing neuroinflammation [56]. Therefore, the epigenetically decreased Filip1l expression in the hippocampus of 5xFAD tg mice might be involved in the development and pathogenesis of AD, possibly connected to the function of filamin A. Future investigations remain warranted to determine the physiological binding partner of Filip1l and analyze the potential link between Filip1l and filamin A.

The DMRs of the candidate genes are located inside the gene body. Methylation of promoter or control regions located outside of the transcribed sequences (e.g., enhancer and insulator) is increasingly being recognized as functionally significant in inverse gene expression [57]. The roles of DNA methylation in gene body have been disregarded, but some recent studies have emphasized its importance. A whole-genome study has shown that exons are more highly methylated than introns, and changes in the degree of methylation occur at the exon–intron border, possibly suggesting a role for methylation in the regulation of splicing [58]. Shukla et al. [59] has suggested that binding of the transcription repressor CCCTC-binding factor (CTCF), which can be regulated by DNA methylation, could stop the progression of RNA polymerase II and, since this would affect splicing, this might provide evidence regarding the connection between DNA methylation and splicing. Furthermore, Anastasiadi et al. [60] clearly demonstrated the important, preserved role of methylation at the level of the first intron, and its inverse correlation with gene expression irrespectively of tissue and species. These observations suggest a previously uncharacterized role for DNA methylation of gene body at the transcriptional level, resulting in changes in mRNA and protein expression via alternative splicing. However, further studies are needed to collect additional evidence for a direct correlation between DNA methylation and the expression of the candidate genes presented in this study.

Although we described three new candidate genes unknown to date through a combined analysis of DNA methylation and RNA-seq, the sequence analyses were performed on a small-scale, which is a limitation of the present study. Moreover, it remains unclear whether the confirmed epigenetic alterations are causal factors or are consequences of AD-like symptoms in 5xFAD mice. Several epigenetically regulated genes, including ABCA7, ANK1, APP, BACE1, BIN1, CDH23, DIP2A, DUSP22, homeobox, MAPT, PSEN2, RHBDF2, RPL13, SERPINF1, SERPINF2, SLC24A4, and SORL1, have been verified so far to be linked to AD neuropathology, including A$\beta$ plaques and tau aggregates [61, 62, 63, 64, 65]. Although extensive research has been conducted to understand the significance of DNA methylation in the progression of AD, the application of this information in clinical practice is still in its early stages. This could be attributed to the challenge of reaching a specific brain region of interest and to the fact that biomarkers can in general only be assessed after death. To overcome these obstacles, it is necessary to conduct prospective or retrospective studies on the progression of AD on a large scale, using blood as a substitute for brain tissue. Nevertheless, we believe that the current preclinical investigation contributes to expanding the knowledge and the pool of genes potentially associated with AD epigenetics, and provides the possibility of developing novel drugs based on DNA methylation.

Our study identified three epigenetically regulated genes associated with the neural cytoskeleton and synaptic plasticity in AD neuropathology through a combined analysis of DNA methyl- and RNA-seq in the hippocampus of wild-type and 5xFAD tg mice. DNA methylation may have a significant impact on the preservation of normal hippocampal functions, such as learning and memory [66], hence alterations in DNA methylation could be linked to the disease activity, progression, and clinical prognosis of AD [32]. Therefore, although no treatment based on DNA methylation has been developed yet, counteracting epigenetic changes may facilitate the development of new therapeutic interventions to delay or ameliorate the symptoms of AD.

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Conceived and designed the experiments: CK and JSK. Performed the experiments: SL and HJL. Analyzed the data: JMC, BJ, JK and CM. Wrote the paper: SL, HJL and JSK. All authors contributed to editorial changes in the manuscript. All authors read and approved the final manuscript. All authors have participated sufficiently in the work and agreed to be accountable for all aspects of the work.

The protocols in this study were authorized by the Institutional Animal Care and Use Committee of the Chonnam National University (CNU IACUC-YB-2023-49) and the animals were cared for in accordance with The National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Not applicable.

This work was supported by the Application Development of Standardized Herbal Resources from the Korea Institute of Oriental Medicine (KSN1822320), and Regional Innovation Strategy through the National Research Foundation of Korea funded by the Ministry of Education (2021RIS-002).

The authors declare no conflict of interest. Changjong Moon is serving as one of the Editorial Board members/Guest editors of this journal. We declare that Changjong Moon had no involvement in the peer review of this article and has no access to information regarding its peer review. Full responsibility for the editorial process for this article was delegated to Gernot Riedel.