INTRODUCTION

Parkinson’s disease (PD) is the most widespread movement disorder and the second most common degenerative disease. The incidence of PD is 1–2 per 1000 people, increasing with age up to 2% of the population aged over 60 years old [1]. In some cases of PD, it is possible to detect the monogenic nature of the disease related to a mutations in a number of genes. It is believed that the overall share of monogenic forms of PD among all patients is 5–10% of cases, with the most mutations in the LRRK2, SNCA, PARK2, PINK1 genes [2, 3]. The remaining PD cases are considered multifactorial. Among the causes of the disease are ecological factors (exposure to insecticides, herbicides, and etc.); eating habits; oncological diseases (melanoma); brain traumas; genetic polymorphisms, predisposition to the development of the disease, and deregulation of the expression of risk genes [4–7].

Lately, increasing attention has been dedicated to the epigenetic mechanisms of the development of multifactorial diseases, as they can modify gene expression and change under the influence of external factors. The effect of all known major epigenetic mechanisms is studied in PD, including DNA methylation, histone modification, and non-coding RNAs (ncRNAs) [8–10].

NcRNAs are important regulators of gene expression. Currently, investigation of ncRNAs is at the beginning stages, new classes are being discovered, the classification is becoming more detailed, targets are being determined, functional connections and mechanisms of interaction in normal and pathological conditions are being defined, and the potential for application in diagnostics and treatment is being considered. MicroRNAs of 19–24 nucleotides in length are the most well studied class of ncRNAs. Study into the other classes of small RNAs (sRNAs) and long RNAs of over 200 nucleotides are also at the beginning stages. Today, over 5000 different microRNAs have been determined in the human epigenome (http://www.mirbase.org) and this number is constantly rising.

There is a classical mechanism of maturation and implementation of the biological function of microRNA. The mechanism is as follows: pre-microRNA is transcribed from the genes coding microRNA, which is further processed by Drosha ribonuclease into pre-microRNA. Then pre-microRNA is transported to cytoplasm with exportin-5 (EXP-5). In the cytoplasm Dicer RNase mediates the maturation of microRNA from pre-microRNA. Afterwards, Argonaute (ago) protein binds to microRNA forming the RNA-induced silencing complex (RISC) [11]. This complex is able to specifically bind to the 3’ end of complementary messenger RNA (mRNA), resulting in its further degradation and silencing [12, 13]. Furthermore, the RISC complex can control the expression through binding to the ribosomal subunits [14, 15] and sequestration of transcribed microRNA into processing P-bodies [16]. It should be noted that a single microRNA can have up to 200 targets and even minor alterations in microRNA network may lead to systemic functional shifts of homeostasis, including the development of pathological processes [4].

The functions of different microRNAs are studied for many diseases. MicroRNAs showed a significant impact on the development of cancer [17]. Many recent researches are investigating the contribution of microRNA to the development of neurodegenerative disease [18]. In the present review, we discuss the role of microRNAs in the pathogenesis of PD and their possibile use in diagnosis of the disease.

Search methods. The articles were searched for in Medline and Scopus databases using the following key words: “microRNA”, “miR”, “Parkinson’s disease”, “neurodegenerative disease”.

MicroRNA IN STRUCTURES AND REGIONS OF THE BRAIN

To specify the participation of microRNA in the pathogenesis of PD, numerous studies were performed with experimental animals and cells, however, this approach does not fully represent the real mechanisms of neurodegenerative process in the human brain. Therefore, individual autopsies of different brain regions of patients with PD are considered the most significant data for understanding of the role of microRNA in pathogenesis of the disease.

The first study of microRNA in PD was performed by Kim et al. in 2007, and found a significant increase in miR-133b in the mesencephalon of the patients with PD compared to the control group. Rodend models showed the potential importance of this microRNA in the pathogenesis of PD [19]. Nevertheless, further experiments on miR-133b mutant mice did not reveal any substantial effect of this microRNA on the development and functioning of dopaminergic neurons [20]. Moreover, the mesencephalon of five patients with PD was compared with the control group and no significant alteration in miR-133b expression was found [21]. Therefore, the primary finding on microRNA in PD did not correspond with the later studies.

Minones-Moyano et al. in 2011 found a decrease in miR-34b/c in all brain regions of patients with PD [22]. Further studies showed that those microRNAs carried sites of binding to mRNAs transcribed from the SNCA and PARK2 genes. The reduction of mir-34b/c level may in turn induce the elevation of the α‑synuclein level and a decrease in the parkin protein level, i.e., cause known and significant pathogenic mechanisms of the disease [4].

Cardo et al. analyzed 733 microRNAs with the help of a special panel in patients with PD and found that the level of miR-135b was elevated in the substantia nigra. Apart from that, no differences in the expression of other microRNAs were detected in patients, compared to the control group [23].

In 2014, Kim et al. analyzed a wide range of microRNAs and found differences in microRNA profile of dopaminergic neurons in the substantia nigra in patients with PD. Bioinformatic analysis revealed a functional connection between the elevated level of miR-126 between deregulation of insulin-like growth factor 1 (IGF-1) and neurodegeneration, which was confirmed in cell lines of dopaminergic neurons [24].

After analysis of the profiles of microRNAs in the substantia nigra, Briggs et al. discovered multiple microRNAs differing in their expression between the patients with PD and control group. However, a more precise bioinformatic analysis identified significant elevation of only miR-132 and miR-184 [25].

In 2017, McMillan et al. found a decrease in the expression of miR-7 in the substantia nigra of patients with PD. Later, they confirmed these data on animal models, showing that a reduced concentration of miR-7 promoted hyperexpression of α-synuclein [26]. The direct action of miR-7 on degradation of primary fibrils of α-synuclein was also found in experiments with cells which were performed by Choi et al. in 2018 [27].

In PD, the profile of microRNAs may not only be changed in the substantia nigra but also in other brain areas. In 2012, Cho et al. identified an increase in miR-205 in the frontal cortex and striatum of patients with PD which was associated with enhanced expression of the parkinsonian gene LRRK2. In addition, a binding site for miR-205 was discovered on mRNA of the LRRK2 gene, proposing a pathogenic connection between these phenomena [28].

Hoss et al. (2016) discovered 125 microRNAs with a changed level of expression in the prefrontal cortex of patients with PD compared to a healthy patients. After selecting the most significant microRNAs, they elaborated a panel of 29 microRNAs that may be used to distinguish patients with and without PD with a specificity of 93.9% and sensitivity of 96.6% [29].

Nair et al. analyzed 800 microRNAs in the striatum, and for 13 of them the level of expression was changed; interestingly, many of these microRNAs were related to proinflammatory genes. The expression of genes of the inflammatory cascade was shifted towards the proinflammatory factors. The authors designated this phenomenon to both the development of the disease itself and levodopa intake [30].

Tatura et al. in 2016 studied the concentrations of 744 different microRNAs of the anterior cingulate cortex. The concentration of 43 microRNAs was different in patients with PD compared to the control group. Of them, 13 types of microRNAs were separated, which influenced the expression of DJ-1, PARK2, PINK1, LRRK2, SNCA, and HTRA2, which, according to the literature, are genes of monogenic parkinsonism. These microRNAs were reexamined and it was confirmed that the concentrations of miR-144, miR-199b, miR-221, miR-488, and miR-544 were elevated and the concentrations of miR-7, miR-145, and miR-543 were decreased [10].

The analysis of the data from the above studies shows how poorly reproducible the results of microRNA analysis in the brain are at this moment. At least two studies indicated elevated expression of miR-106a and miR-129-5p, while the expression of the other microRNAs differed significantly from the control group in single studies. This fact may be explained by currently limited amount of studies, the usage of different methods (quantitative PCR, next generation sequencing), and small group sizes (less than 12 patients). Moreover, the high instability of microRNA in the brain should be noted [31]. Last but not the least is the fact that PD is a highly heterogenic disease from both the clinical and pathogenic points of view.

Summarizing the changes, it should be noted that a number of the presented studies indicated a connection between microRNA and the expression of monogenic parkinsonian genes (SNCA, PARK2, and LRRK2) and genes of the inflammatory response, i.e. there is a relation to the main components of the PD pathogenesis. However, the data should be interpreted carefully, since microRNAs are expressed specifically in different brain regions as well as in different cell populations [32], therefore, the information from the neurons of substantia nigra should not be compared with the data from other brain regions.

MicroRNA IN THE BLOOD

MicroRNAs may enter the bloodstream from the central nervous system. They have an ability to bind to proteins and encapsulate into microvesicles called exosomes. Due to accessibility of the necessary biosamples, such as whole blood, leukocytes, serum, plasma, or exosomes, these studies seem to be the most promising for searching for the biomarkers of PD and are currently the most common. The search for PD biomarkers is a crucial task and it will help in early diagnostics and differential diagnosis with other parkinsonian syndromes.

Several studies were conducted with the aim of searching for microRNAs applicable as biomarkers. Thus, Kean Khoo et al. in 2012 used a panel of 833 microRNAs and chose 4 of them (a combination of miR-1826/miR-450b-3p, miR-626, and miR-505) to elaborate a panel with a sensitivity of 91% and specificity of 100% [33]. A study by Botta–Orfila (2014) demonstrated after a two-step validation that decreased expression of miR-29c, miR-29a, and miR-19b may serve as a biomarker of PD with a specificity and sensitivity of around 70% [34]. In 2015, Ding et al. carried out an experiment with the help of sequencing and further validation using quantitative PCR. The authors managed to identify five microRNAs (miR-195, miR-15b, miR-221, miR-181a, and miR-185) that showed a high predictive ability for PD according to cluster and ROC-analyses [35]. Furthermore, some researches defined a specific set of microRNAs in the blood that was statistically different in patients with PD from the control group [36–40].

Apart from blood serum and plasma, microRNAs were studied in blood cells and exosomes. Martins et al. (2011) analyzed the leukocytes and found not only the typical microRNAs but also an association between different single nucleotide polymorphisms, reduced concentration of microRNAs, and a defective ubiquitin–proteasome pathway and biosynthesis of glycosphigolipids in patients with PD using bioinfomatic analysis and genome-wide study [41].

An intriguing study was performed by Cao et al. on the level of microRNAs in blood serum exosomes. Among 24 microRNAs, which were previously determined as blood biomarkers, the authors found significant elevation of miR-24 and miR-195 and a decline of miR19b in patients with PD [42].

Several researches have investigated microRNAs during antiparkinsonian treatment. Margis et al. compared the expression of microRNAs in untreated patients with PD, treated patients, and a control group. The untreated group demonstrated a higher level of expression of miR-16-2 and miR-26a2 compared to the two other groups [43]. Alieva et al. found a difference in the expression of miR-7, miR-9-5p, miR-9-3p, miR-129, and miR-132 between the untreated patients with PD and those treated with any medications for PD [44]. Serafin et al. (2015) detected an increase in plasma levels of the microRNAs miR-29a-3p, miR-30b-5p, and miR-103a-3p in patients with PD treated with levodopa compared to patients that did not receive it [45]. A study by Caggiu analyzed the expression of four microRNAs in the peripheral blood. An increase in miR-155-5p was discovered in patients with PD which reduced to the normal level during the treatment with levodopa [46].

Apart from the medications for PD, the influence of deep brain stimulation (DBS) on microRNA profile was studied in patients with PD. Soreq et al. (2013) studied microRNA in blood leukocytes of patients before neurosurgery and under DBS. They found 16 microRNAs with alternated expression in all patients with PD and 11 microRNAs that changed their expression after DBS [47].

Vallelunga et al. examined a small group of patients with PD and multiple system atrophy (MSA) and determined microRNAs that were significantly elevated in both groups (miR-24, miR-233*, miR-324-3p, miR-339-5p) and microRNAs that differed between patients with PD and MSA (miR-24, 34b, 148b) [48]. Jin et al. (2018) studied miR-520d-5p. It was elevated in patients with PD but neither in patients with MSA nor with Alzheimer’s disease, which is important for differential diagnostics [49].

Analysis of the above studies suggests that it is still impossible to choose a single microRNA as a biomarker. A few microRNAs had a proven expression profile in PD in more than two studies (Table 1). Massive systemic reorganization of gene expression occurs in PD, particularly in microRNAs. Chatterjee et al. (2017) made attempts to select microRNA networks that were concomitantly changing in the blood and brain using bioinformatics analysis [50].

Table 1. MicroRNAs in blood
Full size table

MicroRNA IN CEREBROSPINAL FLUID

Due to its stable composition, direct contact with brain tissue, and the absence of various fractions, cerebrospinal fluid (CSF) is a suitable substrate for obtaining biomarkers; however, lumbar puncture is not performed routinely in patients with PD, therefore only few studies used this method.

The research of Burgos et al. was aimed to analyze the differences in microRNA profiles between healthy volunteers, patients with PD, Alzheimer’s disease, and PD with dementia. Many microRNAs with variable expression were detected in the CSF and blood of patients from the studied groups [36]. Gui et al. tried to find the biomarkers of PD in CSF. To select microRNAs, the authors predicted metabolic pathways with the help of a special software and confirmed further using quantitative PCR. A changed level of expression was seen in 27 microRNAs. The combination of miR-153 and miR-409-3p had the highest sensitivity and specificity of all to show a statistical difference between PD and the control group [51].

Marques et al. investigated the expression of 10 microRNAs in CSF of patients with PD and MSA. The patients with PD were characterized by elevated expression of miR-205 and reduced expression of miR-24 [52].

MicroRNA IN INDUCED PLURIPOTENT STEM CELLS

Recent studies investigated the profile of microRNA in PD in induced pluripotent stem cells derived from fibroblasts. Tolosa et al. (2018) showed an enhanced expression level of miR-95p, miR-135a-5p, miR-135b-5p, miR-449a, and miR-449b-5p and a decrease in miR141-3p, miR-199a-5p, miR-299-5p, miR-518e-3p, and miR-519a-3p in patients with PD [53]. This is a novel research direction, hence, comparison of the results from different studies is not yet possible.

CONCLUSIONS

The studies on microRNA in patients with PD and other neurodegenerative diseases have two major directions: discovery of new mechanisms of pathogenesis of the diseases in order to find targets for targeted therapy, and the search for biomarkers for diagnostics and differential diagnostics of PD.

Participation of microRNA in the pathogenesis was studied using cell and animal models, brain tissues of people with PD, blood samples, and CSF of patients with PD. Numerous studies provide various results and yet it is impossible to build an overall organized picture of PD pathogenesis. Several studies reported an influence of microRNA on the functioning of genes responsible for manifestation of monogenic types of PD. The search for targeted therapy is mainly concentrated on these [4]. However, many of the microRNAs related to monogenic parkinsonism do not correspond with the microRNAs seen in the total selection of PD patients, which once again reflects the complexity of the pathogenesis and heterogeneity of PD. Moreover, new studies are appearing that combine the results using bioinformatic methods into united functional networks of microRNAs, though this approach is currently developing and requires further study.

The chances of elaborating the panels of microRNAs for diagnostics of PD seem to be more optimistic. Several authors have already proposed lists of microRNAs useful for diagnostics of the disease with high predictive capacity and differential diagnostics with other parkinsonian syndromes.