Encyclopedia
Please note this is a comparison between Version 1 by Aymeric Lanore and Version 5 by Beatrix Zheng.
Parkinson’s disease (PD) is a disorder characterized by a triad of motor symptoms (akinesia, rigidity, resting tremor) related to loss of dopaminergic neurons mainly in the Substantia nigra pars compacta. Diagnosis is often made after a substantial loss of neurons has already occurred, and while dopamine replacement therapies improve symptoms, they do not modify the course of the disease. Although some biological mechanisms involved in the disease have been identified, such as oxidative stress and accumulation of misfolded proteins, they do not explain entirely PD pathophysiology, and a need for a better understanding remains. Neurodegenerative diseases, including PD, appear to be the result of complex interactions between genetic and environmental factors. The latter can alter gene expression by causing epigenetic changes, such as DNA methylation, post-translational modification of histones and non-coding RNAs. Regulation of genes responsible for monogenic forms of PD may be involved in sporadic PD.
- Parkinson’s disease
- Parkinson’s and related diseases
- epigenetic
- neurodegeneration
- DNA methylation
- histone modification
- genetic
- RNA-based gene regulation
1. Introduction
Parkinson’s disease (PD) is a neurodegenerative disease, characterized by progressive degeneration of the dopaminergic neurons of the Substantia nigra pars compacta. Its pathology is multifactorial; influenced by both environmental and genetic determinants [1]. Several pathogenic mutations have been linked to autosomal dominant or recessive forms of PD. The discovery of these genes allowed a new insight into the pathophysiology of this disorder [2]. Pathological hallmarks of PD include the presence of cytoplasmic inclusions, called Lewy bodies (LB), mainly composed of aggregated αsynuclein (α-Syn) [3]. Multiplication and point mutations of SNCA encoding α-Syn are now recognized to cause autosomal dominant PD and they are suspected of promoting α-Syn aggregation. The PINK1, PARKIN and DJ1 genes encode for proteins required by mitochondria, which are essential components of neurons for ATP synthesis, calcium storage, lipid metabolism and neuronal survival [2][4]. The fact that mutations in these genes lead to PD is a strong argument that mitochondrial dysfunction is involved in the pathophysiology of PD.
Parkinson’s disease (PD) is a neurodegenerative disease, characterized by progressive degeneration of the dopaminergic neurons of the Substantia nigra pars compacta. Its pathology is multifactorial; influenced by both environmental and genetic determinants [1]. Several pathogenic mutations have been linked to autosomal dominant or recessive forms of PD. The discovery of these genes allowed a new insight into the pathophysiology of this disorder [2]. Pathological hallmarks of PD include the presence of cytoplasmic inclusions, called Lewy bodies (LB), mainly composed of aggregated αsynuclein (α-Syn) [3]. Multiplication and point mutations of SNCA encoding α-Syn are now recognized to cause autosomal dominant PD and they are suspected of promoting α-Syn aggregation. The PINK1, PARKIN and DJ1 genes encode for proteins required by mitochondria, which are essential components of neurons for ATP synthesis, calcium storage, lipid metabolism and neuronal survival [2,4]. The fact that mutations in these genes lead to PD is a strong argument that mitochondrial dysfunction is involved in the pathophysiology of PD.
Genetic mutations account for only 10% of patients with PD, and therefore environmental exposure seems to play a major role in PD [2]. Epigenetic modulation of gene expression by environmental factors is increasingly studied. Epigenetic regulation involves different mechanisms such as modification of the histones of chromatin and DNA methylation changes [5]. Chromatin is a dynamic scaffold and modification of its main components, the histones, can modulate gene expression [6]. The effect of histone modification is mediated either by directly affecting the structure of chromatin, by disrupting the binding of proteins that associate with chromatin or by attracting certain effector proteins to chromatin [5]. Histone acetylation decreases the compression of chromatin and promotes gene transcription. Methylation of histone H3 lysine 4 (H3K4), H3K36 and H3K79 are marks of transcriptional activation, whereas methylation of H3K9, H3K27 and H4K20 are repressive modifications of transcription, involving the recruitment of methylating enzymes and HP1 to the gene promoter [7].
Direct DNA methylation is also a key epigenetic mechanism regulating gene expression. It is a reversible modification of DNA, which consists of the addition of a methyl group to the fifth carbon position of a cytosine, converting it to 5-methylcytosine (5mC). The transfer of a methyl group is carried out by DNA methyltransferase (DNMT) enzymes [8]. This epigenetic mark is frequently found within a 5’-Cytosine-phosphate-Guanosine sequence, called a CpG site [9]. DNA methylation is not homogeneously distributed in the genome, CpG sites are clustered in sequences called CpG islands (CGI). Methylation of promoter-associated CGIs can impair transcription factor binding or recruit repressive binding proteins, thus reduce gene expression [10]. Cytosine methylation is mediated by DNMTs, which can be classified as de novo (DNMT3A and DNMT3B) and maintenance (DNMT1) [11].
Epigenetic regulation is also closely linked to non-coding RNAs. Non-coding RNAs are classified according to their size, with small RNAs less than 200 nucleotides long and long non coding (LncRNAs) longer than 200 nucleotides [12]. Among the small non-coding RNAs, microRNAs (miRNAs) are the most studied. Mature miRNAs bind to complementary sequences of the target messenger RNA (mRNA), often in the 3′ untranslated region (3’UTR), and can increase mRNA degradation but also inhibit translation without reducing mRNA expression [13][14]. The regulation of gene expression by Long non-coding RNAs (LncRNAs) is not yet fully understood. However, they appear to be important genomic regulators, from the epigenetic to the post-translational level [15].
Epigenetic regulation is also closely linked to non-coding RNAs. Non-coding RNAs are classified according to their size, with small RNAs less than 200 nucleotides long and long non coding (LncRNAs) longer than 200 nucleotides [12]. Among the small non-coding RNAs, microRNAs (miRNAs) are the most studied. Mature miRNAs bind to complementary sequences of the target messenger RNA (mRNA), often in the 3′ untranslated region (3’UTR), and can increase mRNA degradation but also inhibit translation without reducing mRNA expression [13,14]. The regulation of gene expression by Long non-coding RNAs (LncRNAs) is not yet fully understood. However, they appear to be important genomic regulators, from the epigenetic to the post-translational level [15].
Epigenetic mechanisms influencing the development of sporadic PD have yet to be identified. Genes involved in monogenic forms of PD could be over- or under-expressed in the sporadic form compared to the general population without PD (controls). In the context of altered expression of these genes in sporadic forms, it can be assumed that epigenetic mechanisms may be involved in this dysregulation (
Table 1
).
Table 1.
Expression profile and epigenetic changes observed for genes involved in monogenic forms of PD.
| Studies | Tissues Analyzed | Proteins in Controls vs. sPD | mRNA in Controls vs. sPD | DNA Methylation in Controls vs. sPD | MiRNA Expression in Controls vs. sPD | Reference |
|---|
| SNCA | |||||||
| Grundemann et al., 2008 | Brain: DA neurons of SN | Increase | Increase | - | - | [16] | |
| Jowaed et al., 2010/Matsumoto et al., 2010 | Brain | - | - | Hypomethylation | - | [17][18] | [17,18] |
| Pihlstrom et al., 2015/Ai et al., 2014/Tan et al., 2014 | Blood immune cells | - | - | Hypomethylation | - | [19][20][21] | [19,20,21] |
| Minones-Moyano et al., 2011 | Brain | - | - | - | Decrease miR-34b/c | [22] | |
| LRRK2 | |||||||
| Cho et al., 2013 | Brain: frontal cortex/striatum | Increase | No difference | - | Decrease miR-205 | [23] | |
| Cook et al., 2017 | Blood | Increase | - | - | - | [24] | |
| Tan et al., 2014 | Blood immune cells | - | - | Hypomethylation | - | [21] | |
| PRKN | |||||||
| Beyer et al., 2008 | Brain | - | Increase in variant TV3 and TV12 | - | - | [25] | |
| Cai et al., 2011 | Blood immune cells | - | - | No difference | - | [26] | |
| De Mena et al., 2013 | Brain | - | - | No difference | - | [27] | |
| Eryilmaz et al., 2017 | Blood immune cells | - | - | Hypomethylation | - | [28] | |
| Ding et al., 2016 | Plasma | - | - | - | Decrease miR-181a | [29] | |
| Xing et al., 2020 | Brain | - | - | - | Decrease miR-218 | [30] | |
| Serafin et al., 2015 | Plasma | - | - | - | Increase miR-103a-3p | [31] | |
| PINK1 | |||||||
| Muqit et al., 2006 | Brain | Increase ∆1-PINK1 | - | - | - | [32] | |
| Blackinton et al., 2007 | Brain: SN | - | No difference | - | - | [33] | |
| Navarro-Sanchez et al., 2018 | Brain: SN | - | - | No difference | - | [34] | |
| Fazeli et al., 2020/Dos Santos et al., 2018 | PBMC/CSF | - | - | - | Decrease miR-27a | [35][36] | [35,36] |
| DJ1 | |||||||
| Kumaran et al., 2009 | Brain | Decrease | Decrease | - | - | [37] | |
| Tan et al., 2016 | Blood immune cells | - | - | No difference | - | [38] | |
| Chen et al., 2017 | Plasma | - | - | - | Increase miR-4639-5p | [39] | |
| GBA | |||||||
| Murphy et al., 2014 | Brain | Decrease | No difference | - | - | [40] | |
| Moors et al., 2019 | Brain | - | No difference | - | - | [41] | |
| Eryilmaz et al., 2017 | Blood immune cells | - | - | No difference | - | [28] | |
sPD: sporadic Parkinson’s disease; PBMC: peripheral blood mononuclear cell; CSF: cerebrospinal fluid; DA: dopaminergic; SN: Substantia nigra; controls: general population without PD.
sPD: sporadic Parkinson’s disease; PBMC: peripheral blood mononuclear cell; CSF: cerebrospinal fluid; DA: dopaminergic; SN: Substantia nigra; controls: general population without PD.
2. Current Insights
The expression of genes involved in familial forms of PD seems to also be dysregulated in sporadic PD patients. However, it remains unclear whether the change in gene expression corresponds to a pathophysiological mechanism, a marker of degeneration or a protective effect. For SNCA, its overexpression seems to be a prerequisite for aSyn aggregation leading to Lewy body (LB) formation [42]. Hypomethylation of SNCA could be involved in its overexpression. A pathophysiological mechanism suggested is the sequestration of cytoplasmic DNMT1 by the aggregates leading to the dysregulation of aSyn homeostasis [43]. The decrease in the brain level of miR-34b/c could also participate in the overexpression of SNCA [44]. The increase in LRRK2 kinase activity mediated by the G2019S and other disease-causing mutations raises the question whether LRRK2 is overexpressed in sporadic PD [45]. Increased expression of LRRK2 was found in the brains of PD patients, without increased transcripts [23]. A step in the pathophysiology could be a decrease in miR-205 expression, but the causes of this dysregulation is not clear [23]. Furthermore, the regulation of LRRK2 expression by immune stimulation in blood tissue suggests the implication of neuro-inflammatory mechanisms [46] and could represent an accessible epigenetic biomarker. Due to alternative splicing and multiple splice isoforms, it is more challenging to understand the relationship between Parkin expression and its pathophysiology in sporadic PD [47]. Epigenetic mechanisms could be involved in the modification of the PRKN expression profile. Mitochondrial damage could induce overexpression of PINK1 [48]. However, the role of increased PINK1 expression is unclear. It could be either protective or participating in neuronal death. Inflammation in PD seems to be most intense at the beginning, just after clinical diagnosis, attenuating in later stages [49][50]. It has been reported that miR-27a is downregulated in macrophages after stimulation [51]. The decrease in miR-27a observed in early PD may be involved in the inflammation-induced upregulation of PINK1. Decreased transcription and translation of DJ-1 in sporadic forms of PD may also be involved in the pathophysiology of the disease [37]. Oxidative stress leads to the expression of miR-494-3p [52]. The induction of miR-4639-5p expression by oxidative stress remains to be explored. The researchers can hypothesize that miRs could be sensors of oxidative stress and contribute to the cellular response by downregulating DJ1. Decreased GCase protein levels and activity may also be involved in the development of sporadic PD [40]. MiRNA and lncRNA mechanisms may be involved in the regulation of GBA expression but their change in expression level in sporadic PD patients is not clear. The mechanisms regulating miRNA expression are not well elucidated. An interesting approach would be to explore DNA methylation patterns or histone modifications at the transcription sites of these miRNA-regulating genes involved in monogenic forms of PD. Furthermore, epigenetic modifications involve tissue-specific processes. The observation of these mechanisms in pathologically relevant cell types and access to these cells is complex. Analysis of native brain tissue allows the researchers to observe epigenetic changes that take place in the brains of PD subjects. However, these analyses are limited to the post-mortem brain and longitudinal studies cannot be performed to assess the dynamics of epigenetic mechanisms. The collection of more accessible peripheral tissues such as blood or CSF allows an analysis in living individuals over time. However, since the methylation profile is tissue-specific, observations in these tissues do not confirm identical modifications in neurons. Since LRRK2 expression seems to play a role in the inflammatory response, the study of the regulation of its expression in the blood immune cells of PD patients could be interesting. The use of animal models of PD for the analysis of epigenetic alterations presents some difficulties since epigenetic mechanisms are species-specific, and these models may not be representative of epigenetic alterations in humans. However, they can allow for the study of specific neuronal populations [53]. Human iPSC-derived neurons have been developed, which allows for the generation of specific neuron lineages such as dopaminergic neurons derived from PD patients [54][55]. However, the generation of these cells requires cellular reprogramming based on epigenetic modifications resulting in “epigenetic rejuvenation” [56] and may bias the study of epigenetic alterations associated with age in PD. More recent direct conversion techniques make it possible to obtain induced neurons (iNs) with the same epigenetic age as their original fibroblasts. This conversion leaves the most age-related epigenetic marks intact but nevertheless leads to a reorganization of large parts of the epigenome [57]. On the other hand, significant epigenetic changes are largely cell type specific and the value of maintaining epigenetic marks in adult fibroblasts is controversial. Although these in vitro models have limitations, they offer the opportunity to directly study putative effects of epigenetic modifications on gene expression. Moreover, they allow the development of 3D brain organoids or spheroids (cerebral organoids) with a better reproduction of the cerebral environment. This enables for better understanding of the epigenetic modifications that take place in neurons and other cell lineages such as astrocytes and microglia, also suspected in the pathophysiology of PD [55]. In addition to providing an understanding of the pathophysiology, epigenetic modifications could also allow for the development of biomarkers for the diagnosis, prognosis and monitoring of PD [58]. The demonstration of the causality of epigenetic mechanisms in the onset or progression of the disease could allow for the emergence of new therapeutic targets. Epigenomic identity may also be mediated by chromosome folding [59]. Recent studies reveal that the 3D organization of the genome correlates with epigenetic modifications and that these modifications predict the structure of chromatin [59][60]. Moreover, changes in the 3D architecture of chromosomes have been observed in cancer cells [61]. This new field of analysis could allow a better understanding of the involvement of epigenetic modifications in PD.The expression of genes involved in familial forms of PD seems to also be dysregulated in sporadic PD patients. However, it remains unclear whether the change in gene expression corresponds to a pathophysiological mechanism, a marker of degeneration or a protective effect.
For SNCA, its overexpression seems to be a prerequisite for aSyn aggregation leading to Lewy body (LB) formation [45]. Hypomethylation of SNCA could be involved in its overexpression. A pathophysiological mechanism suggested is the sequestration of cytoplasmic DNMT1 by the aggregates leading to the dysregulation of aSyn homeostasis [59]. The decrease in the brain level of miR-34b/c could also participate in the overexpression of SNCA [75].
The increase in LRRK2 kinase activity mediated by the G2019S and other disease-causing mutations raises the question whether LRRK2 is overexpressed in sporadic PD [88]. Increased expression of LRRK2 was found in the brains of PD patients, without increased transcripts [23]. A step in the pathophysiology could be a decrease in miR-205 expression, but the causes of this dysregulation is not clear [23]. Furthermore, the regulation of LRRK2 expression by immune stimulation in blood tissue suggests the implication of neuro-inflammatory mechanisms [101] and could represent an accessible epigenetic biomarker.
Due to alternative splicing and multiple splice isoforms, it is more challenging to understand the relationship between Parkin expression and its pathophysiology in sporadic PD [117]. Epigenetic mechanisms could be involved in the modification of the PRKN expression profile.
Mitochondrial damage could induce overexpression of PINK1 [130]. However, the role of increased PINK1 expression is unclear. It could be either protective or participating in neuronal death. Inflammation in PD seems to be most intense at the beginning, just after clinical diagnosis, attenuating in later stages [141,142]. It has been reported that miR-27a is downregulated in macrophages after stimulation [143]. The decrease in miR-27a observed in early PD may be involved in the inflammation-induced upregulation of PINK1.
Decreased transcription and translation of DJ-1 in sporadic forms of PD may also be involved in the pathophysiology of the disease [37]. Oxidative stress leads to the expression of miR-494-3p [144]. The induction of miR-4639-5p expression by oxidative stress remains to be explored. We can hypothesize that miRs could be sensors of oxidative stress and contribute to the cellular response by downregulating DJ1.
Decreased GCase protein levels and activity may also be involved in the development of sporadic PD [40].
MiRNA and lncRNA mechanisms may be involved in the regulation of GBA expression but their change in expression level in sporadic PD patients is not clear. The mechanisms regulating miRNA expression are not well elucidated. An interesting approach would be to explore DNA methylation patterns or histone modifications at the transcription sites of these miRNA-regulating genes involved in monogenic forms of PD. Furthermore, epigenetic modifications involve tissue-specific processes. The observation of these mechanisms in pathologically relevant cell types and access to these cells is complex.
Analysis of native brain tissue allows us to observe epigenetic changes that take place in the brains of PD subjects. However, these analyses are limited to the post-mortem brain and longitudinal studies cannot be performed to assess the dynamics of epigenetic mechanisms. The collection of more accessible peripheral tissues such as blood or CSF allows an analysis in living individuals over time. However, since the methylation profile is tissue-specific, observations in these tissues do not confirm identical modifications in neurons. Since LRRK2 expression seems to play a role in the inflammatory response, the study of the regulation of its expression in the blood immune cells of PD patients could be interesting. The use of animal models of PD for the analysis of epigenetic alterations presents some difficulties since epigenetic mechanisms are species-specific, and these models may not be representative of epigenetic alterations in humans. However, they can allow for the study of specific neuronal populations [145].
Human iPSC-derived neurons have been developed, which allows for the generation of specific neuron lineages such as dopaminergic neurons derived from PD patients [146,147]. However, the generation of these cells requires cellular reprogramming based on epigenetic modifications resulting in “epigenetic rejuvenation” [148] and may bias the study of epigenetic alterations associated with age in PD. More recent direct conversion techniques make it possible to obtain induced neurons (iNs) with the same epigenetic age as their original fibroblasts. This conversion leaves the most age-related epigenetic marks intact but nevertheless leads to a reorganization of large parts of the epigenome [149]. On the other hand, significant epigenetic changes are largely cell type specific and the value of maintaining epigenetic marks in adult fibroblasts is controversial.
Although these in vitro models have limitations, they offer the opportunity to directly study putative effects of epigenetic modifications on gene expression. Moreover, they allow the development of 3D brain organoids or spheroids (cerebral organoids) with a better reproduction of the cerebral environment. This enables for better understanding of the epigenetic modifications that take place in neurons and other cell lineages such as astrocytes and microglia, also suspected in the pathophysiology of PD [147]. In addition to providing an understanding of the pathophysiology, epigenetic modifications could also allow for the development of biomarkers for the diagnosis, prognosis and monitoring of PD [150]. The demonstration of the causality of epigenetic mechanisms in the onset or progression of the disease could allow for the emergence of new therapeutic targets.
Epigenomic identity may also be mediated by chromosome folding [151]. Recent studies reveal that the 3D organization of the genome correlates with epigenetic modifications and that these modifications predict the structure of chromatin [151,152]. Moreover, changes in the 3D architecture of chromosomes have been observed in cancer cells [153]. This new field of analysis could allow a better understanding of the involvement of epigenetic modifications in PD.
