Parkinson’s disease (PD) is a progressive disorder characterized by resting tremors, rigidity, slow movement, and gait instability. It affects more than 1%–3% of people over 65 years old in the world. Although levodopa replacement and deep brain stimulation can improve PD symptoms, no treatment can stop disease progression (Pringsheim et al., 2021; Zhang et al., 2022; Livingston and Monroe-Duprey, 2023). Molecular studies have characterized that the pathological hallmark of PD is misfolded protein aggregation (Lewy body) mainly composed of α-synuclein (α-SYN), which is coded via the SNCA gene in dopaminergic neurons in substantia nigra (SN) of PD patients (Dodel et al., 2008; Panicker et al., 2021). PD is divided into sporadic cases with unknown causes and familial cases which are attributed to causative mutations in the genes, such as SNCA and Parkin. Overexpression of SNCA is involved in the pathogenesis of familial and sporadic PD.

Since identification of the mutations in the SNCA gene in familial PD in 1997, >50 PD-related genes and loci have been identified. These include the PD loci of PARK1-15, such as SNCA, Parkin, Dj-1, and LRRK2, and other PD-associated genes, such as MAPT, GBA, SMPD1, and LRP10 (Nalls et al., 2019; Bandres-Ciga et al., 2020; Day and Mullin, 2021; Song et al., 2022). These studies have shown that the expression products of PD-related genes can interact with cellular molecules in lysosomes and mitochondria to induce apoptosis and neurodegeneration in PD (Coppedè, 2012; Yu et al., 2015; Panicker et al., 2021; Song et al., 2022). In addition to the genetic variations in the PD genes, studies have also demonstrated epigenetic regulations, which can change the expression of mRNAs and proteins associated with the pathogenesis of PD. Epigenetic modifications alter the expression of a gene or chromosomal region without changing the DNA sequence of the gene itself (Pavlou and Outeiro, 2017; Bertogliat et al., 2020; Rittiner et al., 2022).

Eukaryotic chromatin is tightly packaged inside the nucleus of the cell. The basic DNA packaging unit in the nucleus is the nucleosome, which is composed of 147 bp DNA wrapped around a histone octamer (two copies of histones H2A, H2B, H3, and H4). Nuclear chromatin regulates various cellular and biological processes by keeping some specific DNA loci in an ‘open’ state where they are accessible to proteins responsible for DNA replication, transcription, and DNA repair, while the other regions of DNA are not accessible. Transcriptionally active chromatin is defined as euchromatin while transcriptionally silent chromatin is referred to as heterochromatin (Grunstein, 1997; Campos and Reinberg, 2009). Some epigenetic alterations have been shown to affect the onset and progression of many degenerative diseases including PD (Jiang et al., 2008; Basavarajappa and Subbanna, 2016). In this review, we discuss the epigenetic mechanisms for PD which includes DNA methylation, the post-translational modification (PTM) of histones, and microRNAs (miRNAs) as shown in Figure 1. We will also summarize some potential epigenetic diagnostic and therapeutic targets for PD.

DNA methylation is the key epigenetic modification by which a methyl group is added to cytosine, converting it to 5-methylcytosine (5-mC) by DNA methyltransferases (DNMTs). This modification occurs on cytosine adjacent to guanines (the CpG islands) and plays a crucial role in regulating gene expression, cellular differentiation, and developmental process. In mammalian cells, DNMT1 is the major regulating enzyme for DNA methylation after replication, while DNMT3a and DNMT3b are essential for the de novo methylation predominantly during early development and gametogenesis (Bird, 2002; Andrews et al., 2023). Past studies have demonstrated that the pathogenesis of PD is mainly attributed to the elevated expression of α-synuclein which induces protein aggregates (Lewy bodies) in the middle brain of patients with PD (Espay et al., 2020). Even though the pathogenesis of familial PD is caused by mutations in the PD genes, such as SNCA, LRRK2, and Parkin, for most sporadic patients, the disease is caused because of a combination of mutations in the PD-associated genes and environmental risk factors. Methylation alterations have been found in the promoter/intron 1 of these PD genes, contributing to PD risk (Cai et al., 2011; Panicker et al., 2021; Sohrabi et al., 2023). A recent review demonstrates the genetic and environmental factors leading to PD risk and summarizes epigenome-wide association studies (EWAS) of PD conducted over the past decade. Variants in the SNCA gene, exposure to environmental factors, and physical activity are suggested to impact the level of CpG methylation in PD risk genes, thereby contributing to PD pathology (Schaffner and Kobor, 2022).

An earlier genome-wide DNA methylation study detected 2908 CpG islands with differential methylation (DM) in the brain and 3897 CpG islands in the blood of patients with PD (Masliah et al., 2013). Aging-related methylation patterns indicated methylation alterations associated with aging and age-related diseases (Noroozi et al., 2021). Further analysis of the top 1,000 genes with the aging-related methylation pattern identified 10 genes with DM are associated with PD. These genes include SLC12A5, ABCA3, FHIT, FAT1, CPLX2, APBA1, MAGI2, CNTNAP2, ATP8A2, and SMOC2, most of which are related to neural degeneration. This study supported the role of epigenetic alterations as a molecular mechanism for PD (Masliah et al., 2013). A recent study comparing DNA methylation and transcriptome changes in blood samples from patients with PD and matched controls identified 9 genes that were downregulated and 21 genes that were overexpressed in patients with PD. Furthermore, 31 differentially methylated regions (DMRs) were identified, of which 13 regions have the CpG islands that were hypermethylated, whereas 18 regions were hypomethylated in patients with PD. The DMRs were mainly found in the genes of NFYA, DDR1, RING finger ubiquitin ligase (RNF5), acetyltransferase AGPAT1, and vault RNA (VTRNA2-1). This study indicated that more genes are hypomethylated in patients with PD, suggesting methylation alteration could be an accurate and non-invasive molecular biomarker for PD (Henderson et al., 2021). Further analysis for 44 DMRs indicated that hypomethylated regions include the promoter region of NFYA and Discoidin domain receptor (DDR1) which codes for a receptor tyrosine kinase. The DDR1 gene was also hypomethylated in patients with PD. A DDR1 inhibitor was shown to degrade misfolded α-synuclein aggregates and increased dopamine release in animal models with neurodegenerative diseases (Pagan et al., 2019). The hypermethylated CpG sites were identified near the promoter regions in the genes of RNF5, AGPAT1, VTRNA2-1, and CYP1A1. Both RNF5 and AGPAT1 have been associated with neurodegeneration. The hypermethylation was found in 15 CpG sites spanning the genomic region of the vault RNA2-1 (VTRNA2-1) gene which has two-fold increased expression in the frontal cortex of late-stage patients with PD. The gene coding for the key metabolizing enzyme, CYP1A1 was found to be hypermethylated, which may influence Parkin expression and thus is associated with progression of PD. In addition, the M1 polymorphism (T6235C transition) in CYP1A1 is also found to be associated with PD risk (González-Barbosa et al., 2019; Henderson et al., 2021). The dopamine transporter (DAT) gene codes for dopamine transport and functions for the dopamine release in dopaminergic neurons in the SN and striatum of the brain. DNA methylation of six CpG sites in the 5′-UTR of the DAT1 gene was detected in peripheral blood mononuclear cells of 101 sporadic patients with PD compared to 59 healthy controls. As a result, significantly higher methylation at specific CpG2, CpG3, and CpG5 of CpG sites was associated with advanced-stage of patients with PD (Rubino et al., 2020).

Hypomethylation in the promoter region of the SNCA gene is usually associated with increased expression. DNA methylation at SNCA intron 1 was shown to affect SNCA transcription initiation. A lentiviral vector is developed by CRISPR-deactivated Cas9 (dCas9) fused with the catalytic domain of DNA-methyltransferase 3A (DNMT3A) for targeted DNA methylation within intron 1 of SNCA. Fusion of the DNMT3A catalytic domain with CRISPR-dCas9 drives DNA methylation. After this vector is delivered to dopaminergic neurons derived from human induced pluripotent stem cells (hiPSC) with the SNCA triplication, the mRNA and protein expression of SNCA were downregulated by increasing DNA methylation at intron 1 of SNCA. DNA methylation at intron 1 decreases the expression of SNCA and reduces cellular protein aggregates of the hiPSC-derived dopaminergic neurons with SNCA triplication, as well as reduced mitochondrial ROS production. This study indicated that the specific DNA target sequence combined with the CRISPR-dCas9 technology can be a novel epigenetic-based therapeutic approach for PD (Kantor et al., 2018).

To identify methylation of α-synuclein (SNCA) and leucine-rich repeat kinase 2 (LRRK2) in leukocytes in PD, bisulfite specific PCR-based sequencing was used to detect methylation of CpG islands in the promoter (CpG-1) and intron 1 (CpG-2) of SNCA and LRRK2. Although there were no methylation differences in CpG-1 of SNCA or the LRRK2 promoter, the second, fourth, and ninth CpG sites of the 14 CpG sites of CpG-2 were significantly hypomethylated in patients with PD. Notably, decreased methylation in SNCA is associated with early-onset PD and increased SNCA mRNA in PD (Tan et al., 2014). The Rep1 polymorphism of SNCA is associated with sporadic PD. A study analyzing the relationship between the methylation status of SNCA intron-1 and Rep1 polymorphism in peripheral blood monocyte cells (PBMCs) from 100 sporadic patients with PD and 95 normal controls found that hypomethylation in SNCA intron-1 was associated with increased SNCA expression. Hypomethylation in the SNCA intron-1 was also identified in postmortem brains from sporadic patients with PD and DNA methylation was associated with the Rep1 polymorphism (Ai et al., 2014). Another study analyzed the methylation status of 72 blood and 24 cerebral cortex samples from patients with PD by detecting CpG islands in intron 1 of SNCA using methylation-sensitive restriction enzyme digestion and quantitative polymerase chain reaction (PCR). Patients with PD showed significant hypomethylation compared with controls (Pihlstrom et al., 2015). An earlier next-generation study using 454 GS-FLX sequencing to analyze methylation in SNCA in 15 patients with Lewy body disease (LBD) and 187 controls indicated that the overall methylation levels in the promoter and intron 1 regions of SNCA were higher in LBD patients (de Boni et al., 2011). A study investigating if altered DNA methylation in SNCA intron 1 is a mechanism that results in PD and dementia with Lewy bodies (DLB) found that Dnmt1 expression is reduced in the nuclear compartment of postmortem brain samples from PD and DLB patients as well as in brain samples from α-synuclein transgenic mouse models. After Dnmt1 is sequestered from the nucleus to cytoplasm, global DNA hypomethylation occurs in human and mouse brains, involving CpG islands in intron 1 of the SNCA, SEPW1, and PRKAR2A genes. Furthermore, reduced levels of nuclear Dnmt1 were partially rescued via overexpression of Dnmt1 in cultured neuronal cells and brains of α-synuclein transgenic mice, suggesting an epigenetic dysregulation mechanism in PD and DLB (Desplats et al., 2011).

Over the past decade, studies have shown that 5-hydroxymethylcytosine (5 hmC) is an epigenetic modifier of various neurodegenerative diseases, including PD. 5 hmC is the oxidation product of 5-methylcytosine (5 mC) ten-eleven translocation (TET) enzymes and generally upregulates gene expression, which is different from the effects of 5 mC (Al-Mahdawi et al., 2014). The 5 hmC level is ∼10% that of 5 mC in the mammalian genome. High levels of 5 hmC are found in the central nervous systems, which is ∼40% as abundant as 5 mC (Sherwani and Khan, 2015). A study that performed sequence analysis for 12 PD genes (including SNCA, PARKIN, and TET) related to DNA methylation and hydroxymethylation in 1,657 patients and 1,394 controls identified TET1 as a major gene in PD, further suggesting that hydroxymethylation might play an important role in PD (Shu et al., 2019). To analyze alterations in DNA methylation (5mcC) and hydroxymethylation (5 hmC) in PD, a recent genome-wide study screened 5 mC and 5 hmC in the substantia nigra of brains from patients with PD and identified 4,119 differentially hydroxymethylated regions (DhMRs) compared to controls. Further analysis revealed that the PD-associated DhMRs are associated with several signaling pathways, including phospholipase D (PLD), cAMP, and Rap1, that regulate neurogenesis and neuronal differentiation. Notably, one of the 5 hmC-modulated genes, PLD1, was found to modulate toxicity of α-synuclein in a Drosophila model of PD. This study demonstrated that 5 hmC may act as an independent epigenetic mechanism contributing to the pathogenesis of PD (Min et al., 2022). A genome-wide study identified elevated hydroxymethylation levels in cytosine modifications at enhancers in patients with PD compared to controls. Patients with PD also showed epigenetic transcriptional upregulation of TET2, which is the main enzyme modifying cytosine in DNA sequences. Moreover, inactivation of Tet2 in mice has been shown to prevent MPTP-induced dopaminergic neuronal loss in the substantial nigra of mouse brains. This study demonstrated that widespread epigenetic dysregulation of enhancers of neurons in PD mice may be mediated via increased TET2 expression, suggesting that downregulation of Tet2 could be neuroprotective and an epigenetic therapeutic target for PD (Marshall et al., 2020). A summary of studies investigating DNA methylation for PD is listed in Table 1.

Histone acetylation is the addition of an acetyl group at lysine residues by histone acetyltransferases (HATs), which is associated with transcriptional activation whereas histone deacetylation usually represses transcription by removing an acetyl group from the acetylated histones, which is catalyzed by histone deacetylases (HDACs) (Barski et al., 2007). The histone deacetylase 3 (HDAC3) is a transcriptional repressor to suppress p53-dependent apoptosis. PINK1 is an autosomal recessive gene for PD and the PINK1 protein with phosphorylase activity can bind to the promoter region of HDAC3 and increase its deacetylase activity through phosphorylation of HDAC3 in neuronal cell lines. PINK1 is also found to increase the binding of phosphorylated HDAC3 to p53 to inhibit p53 acetylation and, thus alleviate p53-mediated neuronal apoptosis. Mutations in PINK1 result in mutant PINK1 protein that cannot phosphorylate HDAC3, which leads to a dysfunction of HDAC3 and increased p53-dependent neuronal apoptosis and neurodegeneration in PD (Choi et al., 2015; Labbé et al., 2016). PD is not only associated with the loss of dopaminergic neurons but also has activated microglia in the SN of the brain. Examining histone acetylation in SN of post-mortem patients with PD and age-matched controls showed a significant increase in acetylated histone residue of AcH3-K9, AcH2A-K5, AcH2B-K15, and AcH4-K5 in patients with PD compared with that of control individuals. However, the histone acetylation in AcH2A-K5, AcH3-K9, and AcH4-K5 was not found to increase in the cerebellar cortex from the patients with PD, indicating that histone acetylation is specifically affecting the SN region (Park et al., 2016). Some studies have demonstrated that alterations in lysine acetylation and deacetylation of histone and nonhistone proteins, which are catalyzed by related lysine acetyltransferases (KATs) and lysine deacetylases (KDACs, respectively, are present in patients with PD and various PD models. An increased level of histone acetylation could be caused because of downregulation of KDACs, including HDAC1, HDAC2, HDAC4, HDAC6, and SIRT1.

A comprehensive analysis of histone acetylation in the SN in the brains of normal individuals, early and late-stage patients with PD, demonstrated that histone acetylation is increased with disease progression in patients with PD, which is likely caused by both degenerating dopaminergic neurons and activated microglia (Harrison et al., 2018). A study has identified transcriptional adapter 2-alpha (TADA2a) as a novel binding partner of α-syn in Lewy bodies. As a p300/CBP-associated factor, TADA2a is associated with the acetylation of histone H3/H4. The A53T mutation in SNCA significantly decreased acetylation of histone H3 in cultured SH-SY5Y cells and histone H3 acetylation is also found to be reduced in the striatum and substantia nigra of PD mice injected with α-synuclein -folded fibrils (PFFs). PFF decreased the expression of TADA2a in the striatum and SN of α-syn PFF-induced PD mice. This study indicated that histone modification may be associated with the α-syn-mediated pathogenesis of PD (Lee et al., 2021). Another study examined the histone modifications in post-mortem brains of patients with PD and controls. As a result, H3K4 me3 was significantly increased at the promoter of SNCA in the SN of patients with PD. To investigate how H3K4 me3 regulates α-synuclein expression, a CRISPR/dCas9-based locus-specific H3K4 me3 demethylating vector was established to reduce H3K4 me3 at SNCA promoter and was shown to decrease mRNA and protein expression of α-synuclein, both in the SH-SY5Y cells of neuronal cell line and dopaminergic neurons derived induced pluripotent stem cells (iPSC) of patients with PD (Guhathakurta et al., 2021). Comprehensive analysis with chromatin immunoprecipitation sequencing (ChIP-seq), immunoblotting, and RNA sequencing identified histone sites with altered acetylation in fresh-frozen brain tissues of 28 patients with PD and 21 control individuals (Toker et al., 2021). This analysis demonstrated a significant acetylation increase in H2BK15ac, H3K9/14 ac, H3K27ac, H3K56ac, and H4K12ac in PD. In addition, H3K27 hyperacetylation was also observed in the PD striatum and cerebellum. The histone acetylation in SIRT1 and SIRT3 genes is also upregulated, but it is not for SIRT2, indicating that sirtuins can modulate the acetylation of this residue. Importantly, this study identified H3K27-hyperacetylated regions in more than 20 genes linked to familial or sporadic forms of PD. H3K27 hyperacetylation is involved in PD-related genes including SNCA, PARK7, PRKN, and MAPT (Dickson, 2018; Lim et al., 2019). The hyperacetylated region was found to be in the enhancer region of the SNCA gene, which was previously shown to be affected by both genetic sequence variation (Soldner et al., 2016) and drug exposure in PD (Toker et al., 2021). A summary of studies investigating histone acetylation for PD is listed in Table 2.

Epigenetic modifications may be involved in the pathogenesis of PD through DNA methylation, acetylation, phosphorylation, and ubiquitylation of histones in chromatins and miRNAs. Epigenetic changes make distinct contributions to neurodegeneration and neural death in PD by regulating gene expression. Although exact epigenetic mechanisms of PD progression are not completely known, current findings indicate critical effects of epigenetic modifications in PD. Some epigenetic alterations can be detected in specific tissues and blood samples from patients with PD and PD animal models, suggesting that epigenetic modulators may be useful biomarkers for PD diagnosis and therapy. Although many studies have revealed different epigenetic mechanisms in PD, epigenetic alterations have not yet been translated into biomarkers suitable for clinical use. One reason may be the difficulty in obtaining postmortem brain tissues from patients with PD. Another reason is that DNA methylation may show different patterns in different brain regions of patients with PD. Thus, more studies are needed to determine specific methylation patterns of blood samples from patients with PD as biomarkers. Future challenges are to identify new epigenetic targets to elucidate the precise epigenetic mechanisms in the different pathological stages of PD and to translate epigenetic modulators into novel biomarkers for PD diagnosis and therapy.