The Relationship Between Serum Levels of lncRNA H19, GAS5, HAR1B, LINC01783 and Clinical Signs of Parkinson's Disease

Betul Ozdilek (  ozdilekbetul@gmail.com ) Istanbul Medeniyet University Faculty of Medicine: Istanbul Medeniyet Universitesi Tip Fakultesi https://orcid.org/0000-0003-1608-9882 Ibrahim Alper Kaya Istanbul Medeniyet University Faculty of Medicine: Istanbul Medeniyet Universitesi Tip Fakultesi Berna Demircan Istanbul Medeniyet University Faculty of Medicine: Istanbul Medeniyet Universitesi Tip Fakultesi Temel Tombul Istanbul Medeniyet University Faculty of Medicine: Istanbul Medeniyet Universitesi Tip Fakultesi Handan Ankarali Istanbul Medeniyet University Faculty of Medicine: Istanbul Medeniyet Universitesi Tip Fakultesi


Introduction
Parkinson's disease (PD) is the most common chronic and progressive neurodegenerative movement disorder after Alzheimer's disease (AD) and clinically present bradykinesia, rigidity, resting tremor and postural instability. In addition to these motor symptoms, non-motor symptoms such as cognitive, psychiatric, sleep and autonomic nervous system dysfunctions are also frequently observed. PD affects approximately 2-3% of people over the age of 65 [1,2]. The prevalence of disease is predicted as nearly double in the following 30 years when the aging of the population is considered especially at the point of the extended life expectancy [2]. Therefore, PD represents a heavy burden on patients and their families and also economic burden on society, so that more effective treatments are urgently required. The dominant pathological features in PD are progressive loss of dopamine-producing neurons in the substantia nigra pars compacta (SNpc) and abnormal deposition of the α-synuclein, Lewy bodies and other proteins [3]. Since the molecular mechanism of neurodegeneration in PD has been mostly unknown, it is clear that further investigations are required. Which molecules can be biomarkers for this neurodegeneration have been studied and long noncoding RNAs (lncRNAs) are thought to be speci c biomarkers for PD diagnosis, clinical stages, prognosis and therapeutic targets [4][5][6][7]. Besides, there are a few studies conducted on the role of various lncRNAs in the pathogenesis of PD [8][9][10][11][12][13][14][15][16][17].
lncRNAs are a class of noncoding RNAs that consist of more than 200 nucleotides in length and represent one of the largest fractions in the human genome [18,19]. Studies have shown that lncRNAs interact with DNA, RNA and protein molecules to regulate gene expression at the epigenetic, transcriptional and posttranscriptional level in cellular homeostasis. Even though studies are still in the preliminary stages, lncRNAs are considered to play a role in development, differentiation, aging and apoptosis activities, immune system and cancer. lncRNAs are highly expressed in various parts of the central nervous system (CNS) and alterations in lncRNA levels have been shown in many neurodegenerative disorders such as AD, PD, Huntington's disease, amyotrophic lateral sclerosis and stroke [20][21][22][23][24][25][26]. Accumulating evidence has shown that lncRNAs play role in the onset and progression of PD and it can be used as a potential therapeutic target for the disease. Somewhat abnormal expression levels of these lncRNAs were detected in the samples of brain tissue, cerebrospinal uid (CSF), blood and even saliva. Serum samples are accessible by minor invasive procedures and offer the possibility of a cheap, fast and quick way of identifying disease-related biomarkers. Whereas the number of the studies on lncRNAs has signi cantly increased in recent years, identi ed lncRNAs' numbers in PD patients have also increased [8-12, 14, 15, 24] We downloaded the experimentally validated disease-to-lncRNA associations according to the lncRNADisease 2.0 database and about 27 of lncRNAs were speci cally linked to PD [26]. When reviewed literature data, we decided to select and analyze four of these lncRNAs; H19, GAS5, HAR1B and LINC01783 (ENST00000415386) which have been implicated in neurodegenerative disorders. In this study, we aimed to analyze serum expression levels of these four lncRNAs in patients with PD and to nd whether those are associated with motor, non-motor symptoms and treatment.

Materials And Methods
This study was conducted in accordance with the ethical standards of the Declaration of Helsinki, and national and international guidelines for researches with human subjects. Protocol was approved Istanbul Medeniyet University (T-GAP-2019-1548) and Goztepe Training and Research Hospital (2019/0223) Institutional Review Boards. All of the subjects recruited to study provided a written informed consent agreeing to participate the project.

Participants
Eighty-three PD patients who were followed up in a movement disorders outpatient clinic and fty healthy controls, matched for age and gender, included in the study. The controls were recruited from the patients' spouses or individuals who wanted to participate in the study. The participants ranged from 40 to 80 years in age and have at least ve years of education.
Patients were diagnosed as having PD based on the Movement Disorder Society Clinical Diagnostic Criteria and regularly followed up by the same experienced neurologist [27,28]. The patients have had this diagnosis for at least 2 years and have been taking antiparkinsonian treatment regularly for the last six months. Demographic information, clinical ndings and scales were recorded in their databank in each follow-up. Data collection included demographic and clinical information such as age, gender, years of education, duration of disease and treatment, positive family history of PD and levodopa equivalent daily dose (LEDD, mg/day) [29]. Exclusion criteria were as follows: (1) presence of other neurological disorders; (2) to be diagnosed with diabetes mellitus, coronary heart disease, ischemic or hemorrhagic stroke, infectious disease, malignant tumor, glaucoma, severe visual and hearing impairment; (3) to receive any anti-in ammatory or immunosuppressive drugs; (4) presence of psychiatric disorder such as moderate to severe depression with a score of 17 or higher on the Geriatric Depression Inventory, psychosis and recent delirium diagnosed with a structured clinical interview; (5) the subjects whose scores less than 24 on the Mini-Mental State test; (6) any history of alcohol and/or substance abuse; (7) to have a brain surgery for PD or another reason. The same exclusion criteria were applied to the controls, and they had negative family history of movement disorders. We only included hypertension comorbidity for the whole participants.

Assessment Of Motor And Non-motor Symptoms In Pd Patients
A complete neurological examination was performed and Turkish versions of the scales were administered to PD patients during "on" periods before collecting blood samples. Onset sign of disease (bradykinesia or tremor) and their lateralization (left or right) and the presence of postural instability and gait dysfunction were noted. Disease severity was measured using the Hoehn and Yahr (HY) staging scale [30]. Scores varied between 1 and 5, higher scores mean higher disease severity. Current clinical symptoms were assessed with the Uni ed PD Rating Scale (UPDRS). It has four parts. Parts I is about mental function disorder and mood, II about motor activities of daily living, III about motor examination, IV about motor complications. PD patients rated items on a Likert scale ranging from 0 normal: symptom not present to 4 severe: symptom present and precludes patient's ability to carry out normal activities or social interactions or to maintain previous standards in personal and family life. A total of these four parts were calculated by total score [31]. All patients were using dopamine agonist therapy. The presence or absence of motor uctuations, dyskinesia, hallucinations, delusions and dopamine dysregulation syndrome was recorded.

Blood Sample Collection
From each participant, ~5 mL venous blood samples were collected into serum vacutainer tubes with gel and clot activator in the morning following 12 hours of fasting. Serum samples were kept at room temperature for one hour, and then centrifuged for 10 min at 4000× g for serum separation. Supernatant serum was stored at -80°C until analysis.

Laboratory Analysis
Total RNA isolation The lncRNAs were isolated from serum samples of PD patients and healthy controls using the miRNeasy serum / plasma kit (Qiagen, Germany) according to the manufacturer's instructions. Brie y, 500 µL of QIAzole lysis reagent was added to 100 µL of the serum sample and incubated the whole reaction mixture for 5 min at room temperature. Then, 100 µL chloroform was added to the lysate tube, vortexed for 15 s and incubated it for 2 min at room temperature. Thereafter, we performed centrifugation at 12,000× g for 15 min at 4°C. ~ 300 µL of the upper aqueous phase was removed and the mixture transferred to a new collection tube, then 450 µL of 100% ethanol was added. Then, 700 µL of the mixture was added to the RNeasy MinElute spin column in a 2 mL collection tube and centrifuged at 8000× g for 15 s at room temperature. After the mixture had completely transferred to the column, we added 700 µl of buffer RWT to each column and centrifuged at 8000× g for 15 s at room temperature. Next, 500 µl buffer RPE was added and centrifuged at 8000× g for 15 s. Finally, 500 µL of 80% ethanol prepared with RNasefree water was added to the column and centrifuged at 8000× g for 2 min. A full speed centrifugation was performed to dry the membrane with an open cover for 5 min. The ltrate and collection tube were discarded at each step. Total RNA was eluted by centrifugation for 1 min at full speed using 14 µL of RNase-free water. Each sample was evaluated by nanodrop for its purity and concentration.
Complementary DNA (cDNA) synthesis After RNA extraction, cDNA was generated using Qiagen cDNA RT2 First Strand Kit. All reverse transcription quantitative polymerase reaction (RT-qPCR) was set up on ice. For each sample, 100 ng of total RNA and 6 µL buffer GE2 completed with RNase-free water to 14 µl of nal volume was incubated at 37°C for 5 min and then kept on ice for 1 min. RT procedure was completed in a total volume of 20 µL with the addition of 6 µL BC5 to each 14 µL of the mixture at 42°C for 15 min and 95°C for 5 min. Thereafter, the cDNA was diluted with 80 µL nuclease-free water for the later use in qPCR.

RT-PCR
Expression levels of serum lncRNAs were analyzed using Rotor-Gene® Q instrument with 2.1.0.9 software and QuantiTech SYBR Green PCR Kit (Qiagen, Germany). qPCR was performed in duplicates, including RT controls to evaluate DNA and nontemplate controls to avoid background signal. The qPCR reaction was set up with minimal changes according to the manufacturer's instructions as follows: 5 µL 2x QuantiTect SYBR Green Master Mix, 1 µL 10x miScript Universal Primer, 1 µL 10x primer assay, 1 µL RNase-free water and 2 µL of cDNA. Reaction mixture was prepared in 0,1 mL strip tubes and caps (Qiagen, Germany) in a total volume of 10 µL for each reaction. The following conditions were used for qRT-PCR to amplify the lncRNAs: 95°C for 10 min, followed by 40 cycles at 95°C for 15 s, 55°C for 30 s and 72°C for 15s. Finally, melt analysis from 55 to 95°C temperature in Rotor-Gene Q instrument with 72 well plate was performed.

Analysis of lncRNA qPCR Data
The relative expression level of each lncRNA was calculated according to the cycle threshold (CT) value by using 2 −∆∆CT method [32]. The lncRNA CT levels were exported appropriately into Qiagen GeneGlobe Data Analysis Center. CT cut off value was set to 40. For normalization, β-actin was chosen appropriately.
Results were obtained using the Qiagen Data Analysis Center web tool.

Statistical analysis
All measured variables were subjected to normality testing using Shapiro-Wilk normality test. Descriptive values were expressed as mean ± standard deviation (SD), median (25-75% quartiles) or count and percent frequencies. Independent samples t-test was used to compare differences between PD patients and control group with regard to numerical variables. The relationships between the categorical variables were evaluated by using Chi-squared test, or Fischer exact test. Multivariate binary logistic regression was performed to identify the signi cant predictors of PD risk. A p-value of less than 0.05 was considered statistically signi cant. Statistical calculations were performed using SPSS (ver. 23).

Demographic and clinical features of the PD patients and healthy controls
A total of 133 participants, including 83 PD patients and 50 healthy controls, were included in the study.
The demographic pro le of the patients and controls and clinical ndings are described in Table 1. No signi cant difference was observed in age (p = 0.283), gender (p = 0.398) and presence of hypertension (p = 0.236) between PD and control groups. But duration of education was found signi cantly longer in control group than PD patients (p = 0.024). All participants were right-handedness and there were no alcohol users. Serum expression levels of lncRNAs H19, GAS5, HAR1B and LINC01783 The normalized value of 3 PD patients and 3 controls could not be obtained from the laboratory analysis so they were estimated by the missing value method. Comparison of normalized values of PD group (27.44 ± 2.893) to control group (27.60 ± 2.027), no signi cant difference was found between the two groups (p = 0.697). This result desired further analysis to be performed. Distribution of normalized values in PD patients and control groups was given in Figure 1. When gure is examined, it is seen that the normalized values in the controls and PD group show a slightly right-skewed distribution.
Relative serum expression levels (2 −∆∆CT ) of four lncRNAs; H19, GAS5, HAR1B and LINC01783 in PD patients and control groups are given in Table 2. No signi cant difference was observed in the levels of these lncRNAs between the two groups. Data are expressed as median (25-75% percentiles).

*: Independent samples t-test
Since the normalized value of 8 PD patients and 3 healthy controls couldn't be obtained from the laboratory analysis, they were estimated by the missing value method. The two groups were compared in terms of these four lncRNAs and no signi cant difference was found (Table 3). There was a signi cant positive correlation between H19 and GAS5 and LINC01783. In addition, a signi cant positive correlation was found between GAS5 and HAR1B and LINC01783. Apart from that, no meaningful relationship was found between other lncRNAs in PD patients (Table 4). Table 4 Correlation analysis between the serum levels of H19, GAS5, HAR1B and LINC01783 in the PD patients. * Signi cance at p < 0.05; ** Signi cance at p < 0.001 The relationship between disease duration, onset sign and lateralization, positive family history, disease severity (HY score) and clinical status (UPDRS score), LEDD in patients and lncRNA H19, GAS5, HAR1B and LINC01783 levels were examined. The results showed that the relative expression of lncRNA GAS5 levels were signi cant negatively correlated with HY and UPDRS II, III and total scores (r = -0.243, p = 0.027; r = -0.286, p = 0.009; r = -0.232, p = 0.035; r = -0.225, p = 0.041, respectively). No signi cant relationship was found between these four lncRNAs and other characteristics of the PD. LINC01783 was found to be signi cantly higher in those with a positive family history (p = 0.047). However, the other 3 lncRNA levels were found in similar amounts in patients with and without family history.
Logistic regression analysis to predict PD risk Considering the age, gender, duration of education and presence of hypertension in PD patients and controls together with lncRNA H19 and GAS5, re-evaluation was performed with a multivariate logistic regression model and the results shown in Table 5. lncRNA HAR1B and LINC01783 were not included in the model as there were many unmeasurable values in the laboratory analysis (25 PD patients and 12 controls for HAR1B, 33 patients and 26 controls for LINC01783). We found that male individuals had signi cantly higher risk for PD. In addition, as the duration of education increases, the risk of PD decreases. lncRNA H19 and GAS5 do not appear to have a discriminating role between patient and control groups. After the individuals with normalized values above 30 were excluded from the data, correlation between the age, gender, duration of education, presence of hypertension and lncRNA H19 and GAS5 were evaluated with the multivariate logistic regression model and the results were given in Table 6. Same results were found. Since there are many unmeasurable values of the HAR1B and LINC01783, they were not included in the model.

Discussion
Although there has been a signi cant progress to understand the mechanisms which lead to PD, it is still challenging to determine speci c biomarkers enabling accurate diagnose of the disease, classi cation and risk factors and predict probable patients [33]. Accumulation of a large number of pathological features α-synuclein, Lewy bodies and other proteins resulted from abnormal proteasome function, mitochondrial dysfunction, oxidative stress, calcium homeostasis, synaptic transmission and neuroin ammation has been shown in multiple studies [4,5,15,34,35]. The combined action of these mechanisms causes autophagy and apoptosis of dopaminergic neurons in the SNpc, promoting the development of PD [12,36]. There has been accumulating signi cant evidence of the important role of lncRNAs in these activities of PD pathogenesis [6, 12,15]. It is generally perceived that diagnosis and treatment of the disease will be facilitated by detecting lncRNAs' pro les that extracted from serum, plasma, CSF, saliva and tissue samples [24]. In this study, we studied to determine the levels of four of these lncRNAs H19, GAS5, HAR1B and LINC01783 in the sera of PD patients.
The ubiquitin-proteasome system (UPS), a nonlysosomal pathway of protein degradation, removes damaged mutant and aberrant proteins in cells, regulates cell cycle DNA damage and repairs apoptosis.
UPS dysfunction causes abnormal accumulation of protein within the cell and gets an important role in the pathogenesis of PD [12,37]. Reduced mitochondrial activity promotes free radical formation and enhances the susceptibility of tissues to oxidative stress, resulting in damage to cellular DNA, lipids and proteins. It has been shown that PD patients have high oxidative stress in the brain, meanwhile dopaminergic neurons don't have the ability to control oxidative stress. The process of apoptosis is highly related to mitochondrial dysfunction and oxidative stress, so plays an important role in death of the dopaminergic neurons in the disease [38]. Autophagy-mediated-protein degradation is a process in which the proteins are engulfed within vesicles that fuse with the lysosomes to degrade proteins. Enhanced autophagy can effectively counter PD [12]. Neuroin ammation is so important in PD. The in ammatory cytokines, affecting the integrity of the blood-brain barrier and synaptic plasticity, accelerate the aging and degeneration of dopaminergic neurons [12]. To clarify the pathogenesis of PD and discover new therapeutic targets, it is clearly important to elucidate lncRNA-mediated regulation of the α-synuclein that linked to some mechanisms such as abnormal modi cation of α-synuclein after translation, aggregation of α-synuclein, toxic effects and degradation [39]. During the inhibition of α-synuclein toxicity by the traditional Chinese medicine Acanthopanax senticosus (AS), a commonly used agent, 341 lncRNAs were differentiated under the stimulation of α-synuclein in microarray expression analysis. 29 of these lncRNAs were involved in the inhibition of α-synuclein neurotoxicity mechanism mediated by AS and 19 were potentially related to α-synuclein neurotoxicity [40]. The potential mechanisms of lncRNAs include the inhibition of PD-linked genes' expressions, the reduction in production of α-synuclein, the maintenance of autophagy system balance, the delay in the apoptosis of dopaminergic neurons, the alleviation of nerve in ammation and so forth. All these ndings indicate that lncRNAs have the potential to become a biomarker and therapeutic target for PD.
Soreq et al in 2014, for the rst time, utilized a whole-transcriptome RNA sequencing to determine all the transcripts that code proteins in leukocyte and lncRNAs in PD patients and controls. However, they couldn't fully elucidate the detailed functions of the identi ed lncRNAs. 13 lncRNAs showed differentiated expression levels and selective PD-induced alteration. These researchers also found 4 of these lncRNAs were reversed after deep brain stimulation treatment. Comparisons of the peripheral blood and brain tissue samples of the PD patients and healthy individuals reveal common differences in lncRNAs, with the same expression trends. In this study it was stated that U1 levels (ENST00000415386 which is LINC01738) were differentiated in PD patients' amygdala and leukocytes [8].
A recent study found that PD-related genes associated with lncRNAs decreased in the SN and cerebellum of patients, just as consistent with the results obtained in peripheral blood cells [41]. lncRNAs in the CSF of patients with PD have a higher frequency compared with controls, corroborating previous reports that various lncRNAs performed essential functions in the regulation of PD progression [42]. Both animal and cell models are commonly used in PD researches. Abnormal expression of approximately 756 lncRNAs was detected in the SNpc of the presymptomatic mice. Although these studies cannot exactly replicate the pathological changes in human PD patients, 87 of lncRNAs had signi cant differentiations in the expression in the SN of PD patients, when compared to the normal tissues [43].
Kraus and colleagues studied lncRNA expression levels in brain tissue of postmortem 20 PD patients and 10 healthy controls [10]. It was identi ed that GAS5 and HAR1B were so abundant and expressed in brain samples and so, they were used as normalizers. They found that only 5 patients had signi cant alterations when compared to controls, out of 90 non-coding transcripts investigated in the study. lncRNA H19 upstream conserved regions 1 and 2 were downregulated signi cantly [10]. Besides, it was shown in recent studies lncRNA H19 can play protective roles against dopaminergic neuronal loss and apoptosis in mice models with PD [13,17]. lncRNA H19 regulates the p53/Notch1 pathway to inhibit neurogenesis in ischemic stroke [44]. H19 is one of the rst lncRNAs, linked to different types of cancer, such as bladder cancer, colorectal cancer, and hepatocellular carcinoma [45]. Moreover, lncRNA H19 regulates neuronal apoptosis in AD, suggesting that the regulation of lncRNA networks have unneglectable in uence on AD pathology, that is they may lighten the unclear etiology of AD and lead current drug therapy which is GAS5, as a member of the lncRNA family, is located on chromosome 1 of the human genome. Studies have shown that GAS5 is abnormally expressed in many tumor disorders and plays an oncogene role by inhibiting apoptosis. Moreover, GAS5 also takes place in the development of in ammation-related disorders [16]. There have been some studies investigating its role in neurological disorders such as ischemic stroke and AD. Acetylcholine release has a related role in the cholinergic nervous system of AD patients [47]. Microglia-induced neuroin ammation plays a signi cant role in PD pathogenesis. In another study, it has been shown that GAS5 can activate microglia and increase the expression level of in ammatory cytokines [48]. All these ndings highly suggest GAS5 involvement in PD development, but further studies are still needed for GAS5 role in PD.
As being con rmed, lncRNAs are expressed in many regions of the brain [9]. The study demonstrated that HAR1 speci cally expressed in Cajal-Retzius neurons in the development human neocortex during a period of 7 to 19 gestational weeks, a crucial time for cortical neuron speci cation and migration [21]. It upregulates reelin. lncRNA LINC01783 (Gene ID: 100132147) (ENST00000415386) locates in the 1p36.13 region of human genome. Underlying its importance in cancer, in a recent study it was reported that LINC01783 had an association with the proliferation, migration, and invasion of cervical cancer cells [25].
No signi cant difference was shown when the expression levels of four candidate lncRNAs, H19, GAS5, HAR1B and LINC01783 in the sera of PD patients were studied in comparison to those of healthy controls. However, it was found that there was solely a signi cant negative correlation between GAS5 and HY stage and UPDRS II, III and total scores when the relationship between the expressed levels of these four lncRNAs were examined in age, gender, duration of education, disease duration, disease onset nding, HY stage, UPDRS scores and treatment doses. Even though this parameter was suggested to have a role in in ammation in other previous studies, in our study we revealed that it was associated with the clinical severity of the disease and there was no correlation with other lncRNAs' levels. Other studies did not report any signi cant relationship in the literature [10,13,15,16]. It was found that LINC01783 lncRNA was signi cantly higher only in PD patients with positive family history.
In conclusion, as we know this is the rst clinical study in order to explore the expression levels of serum lncRNAs in Turkish PD patients. It may help to understand PD pathogenesis better and more in detail to search serum lncRNAs levels, but there have been few studies in the literature so far. It can suggest important contribution to the development of potential biomarkers for PD and also the identi cation of new therapeutic targets, although the mechanisms by which lncRNAs role in complex physiological and pathological cases such as PD have not been elucidated fully yet.

Declarations
Disclosure: The authors report no con ict of interest.