Plasma hsa‐mir‐19b is a potential LevoDopa therapy marker

Abstract Parkinson's disease (PD) is the second most common neurodegenerative disorder among the elderly, the diagnostic and prognostic of which is based mostly on clinical signs. LevoDopa replacement is the gold standard therapy for PD, as it ameliorates the motor symptoms. However, it does not affect the progression of the disease and its long‐term use triggers severe complications. There are no bona fide biomarkers for monitoring the patients’ response to LevoDopa and predicting the efficacy of levodopa treatment. Here, we have combined qPCR microRNA array screening with analysis of validated miRs in naïve versus Levodopa‐treated PD patients. We have identified plasma miR‐19b as a possible biomarker for LevoDopa therapy and validated this result in human differentiated dopaminergic neurons exposed to LevoDopa. In silico analysis suggests that the LevoDopa‐induced miR‐19b regulates ubiquitin‐mediated proteolysis.

with inhibition of Monoaminoxidase (MAO) and Catechol Omethyltransferases (COMT) activity in the periphery (carbidopa, benserazide). 3 Long-term LevoDopa usage associates almost invariably with additional motor side effects and cognitive and neuropsychiatric events, while it does not affect or change the natural progression of cell degeneration towards non-dopaminergic neural circuits. 4 LevoDopa's altered metabolism in PD patients governs the deleterious feedbacks between the main pathogenic axes, including mitochondrial dysfunction, toxic metabolites, proteasome alteration and alpha-synuclein aggregation, thus turning the PD cure into a curse in disguise. 5 MicroRNAs (miRs) are a class of endogenous, small non-coding RNAs able to post-transcriptionally modulate gene expression by binding to complementary sequences in target messenger RNAs (mRNAs). One single miR may alter the stability of hundreds of targets, while several miRs may cooperatively interact with one single mRNA and thus regulate the expression of over one-third of the coding genes in humans. 6 Due to their outstanding stability in various body fluids, miRs are ideal biomarker candidates in various diseases, including PD. 7 Interestingly, in PD patients, most of the miRs were shown to be deregulated, with decreased levels in total blood, serum, and white blood cells, while most of the plasma miRs are upregulated. [8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24] However, there is a large discrepancy among these results, presumably due to several differences in their experimental setups. Moreover, many patients are treated with LevoDopa for several years, which could also affect the miRs levels. To shed light on these discrepancies, we aimed to investigate the levels of miRs in the plasma of LevoDopa-treated PD patients.
In our study, we made use of the sensitivity and specificity of the qRT-PCR technique to quantify in a two-step (discovery and validation) approach the mature miR species that are present in the plasma of LevoDopa-treated PD patients. Analysis of the validated miRs in naïve vs. LevoDopa-treated PD patients identified plasma miR-19b as a sensitive biomarker for LevoDopa therapy, a result further validated in mouse hippocampi and human differentiated dopaminergic neurons exposed to LevoDopa.

| MATERIAL AND ME THODS
The present study has been performed in concordance with the Declaration of Helsinki Code of Ethics and has been reviewed and approved by the local institutional ethics review board.  Table 1.
Peripheral blood was collected between 9 AM and 12 AM in EDTA-coated vacutainers by puncture of the cubital vein; next, the blood was refrigerated at 4°C, processed by centrifugation within the next 2 h, and the plasma fraction stored at −80°C until further use. Plasma samples with signs of hemolysis, turbidity, hyperlipidemia or hyperbilirubinemia have been discarded. After thawing, all plasma samples were again centrifuged for 5 min at 1500 g and 4°C to eliminate precipitates and cell debris contaminants.
Total RNA was extracted from 200 μl of plasma (100 μl in case of Harvard lot) using the miRNeasy Serum/Plasma kit (Qiagen), with Caenorhabditis elegans miR-39 miRNA mimic as a spike-in control for external normalization.

| LevoDopa treatment of mice
The experiments involved two lots (experimental and control) of six one-year-old wild-type mice (three males, three females) of mixed genetic background, housed in Udel ® polysulphone cages, on a 12-h light-dark cycle and fed ad libitum. L-DOPA (Sigma) and Benserazide hydrochloride (Sigma) solutions were prepared freshly (30 min before injections) in physiological saline and administered for 5 days, by two 12-h spaced intraperitoneal injections at 20 mg/kg and 12 mg/kg, respectively. The control lot was injected using the same protocol with equal amounts of Benserazide hydrochloride. The mice were sacrificed on the 6th day, and the hippocampi dissected

| Dopaminergic cells culture
LUHMES cell differentiation to dopaminergic neural cells was performed as previously described. 26 Briefly, LUHMES cells were cul-    Kit followed by individual qRT-PCR quantification using inventoried TaqMan™ Assays (Thermo Fisher). All qRT-PCR reactions were performed in triplicate. The fold changes were calculated by ∆∆CT method of relative quantification using hsa-RNU-44 (for miR calculations) and hsa-Gapdh (for mRNAs calculations) for normalization.

| Statistics
In the discovery lot, differentially expressed miRs were identified by two tails, heteroscedastic Student's t-test and ranked according to their corrected p-values (p = 0.05 as cut-off).
In the validation lot, after checking the normalized Ct values for normality of distribution, we evaluated the differential miR expression's statistical significance between patient and control groups using either a 2-tailed t-test with Welch correction (for normal distribution) or a 2-tailed Mann-Whitney U test for non-normal distribution (statistical cut-off: 0.05).
The differential miR expression's statistical significance between naïve and post-therapy patients' groups was assessed using the 2-tailed Wilcoxon signed-ranks test (p < 0.05) for the Harvard lot.
The statistical significance of the differentially expressed miR in the cultured dopaminergic neural cells exposed to LevoDopa was calculated using a two-way anova test with Dunnett correction. The statistical significance of the differentially expressed target genes in the cultured dopaminergic neural cells exposed to LevoDopa was calculated using a two-way, unpaired, heteroscedastic t-test with Welch's correction (p < 0.05).
The statistical significance of the differentially expressed miR in the mouse hippocampi challenged with LevoDopa was calculated using a two-way, heteroscedastic, unpaired t-test (p < 0.05).
All statistical calculations were performed using Prism 8 for The diagnostic test parameters for the three miRs were estimated using the MedCalc online software (https://www.medca lc.org/calc/diagn ostic_test.php).

| Bioinformatics analysis
Target predictions (against 3′UTR, 5′UTR and CDS regions) for miR- Parkinson's disease model. 28 The miR-19b targets of the differentially expressed genes were subjected to Gene Ontology analysis on the DAVID platform using adjusted FDR of 0.05 as cut-off. 29 3 | RE SULTS

| MicroRNA profiling
Out of the 1008 miRs profiled in the discovery group, only 15 were expressed in all plasma samples and exhibited corrected p-values below the 0.05 threshold, with fold changes (FC) varying between 1.5 and 5.1 ( Table 2).
Next, we asked whether the normalized Ct of the five miRs cor-  (Table S2).

| Mir-19a, miR-19b and miR-195 as diagnostic biomarkers
To evaluate the validated miRs' ability to discriminate between PD and controls, we performed ROC analyses for all five miRs (Table S3 and Figure 1). This analysis revealed that the five miRs exhibit relatively modest performances, with miR-19a, miR-19b and miR-195 having the highest areas under the curve and the lowest p values.   (Table 4 and Figure S2). 29

| DISCUSS ION
There is a bewildering lack of consensus between the circulating miRs expression reports, regardless of their origin: plasma, serum or PBMCs of Parkinson's patients 32 ; this might reflect not only differences in tissue source, cohort sizes, experimental design (from RNA isolation to validation) but also, in the light of our data, differential individual responses to different LevoDopa therapy dosages. 33 F I G U R E 3 miR-19a, mir-19b and miR-195 expression levels in mouse hippocampi exposed to LevoDopa. *p < 0.05; (unpaired, heteroscedastic Student's t-test)  39 Interestingly, global analysis of plasma exosomes did not identify differences in miR-19b expression, thus excluding exosomes as possible contributors to miR-19b PDassociated changes. 40 It is also worth noting that with very few exceptions, the brain-enriched miRs investigated so far show increased plasma levels in PD patients, suggesting that the actual miR source might influence the plasma levels in pathological conditions. 15,24 Very few studies investigated the miR response to LevoDopa therapy in Parkinson's disease, all using a targeted approach and focusing on miR expression in peripheral blood cells. However, the results show the same disconcerting lack of consensus: Alieva et al. and miR-19b are considered 'male-specific' miRs, while miR-195 is rather female-specific. 45 An aspect worth investigating would be the mechanism leading to mature miR upregulation upon LevoDopa exposure. While the human plasma and mouse hippocampi show concordant changes in mature miR-19b and miR-19a levels, consistent with a common, transcriptional regulation of the miR-17~92 cluster, the cultured dopaminergic neurons' response diverges as the intensity of LevoDopa exposure increases, suggesting that different regulatory events might also be at work. Given that miR-19b is less accessible to Drosha processing from pri-miR-17~92 than miR-19a, it is tempting to speculate that LevoDopa post-transcriptionally alters the miR expression ratios within this cluster. 46 Whether this LevoDopa-induced change in miR-19a/miR-19b ratio scenario also applies to the miR-106a-303 cluster (the chromosome X source of miR-19b) and which is its significance for dopaminergic neurons' biology remains to be established.
Analysis of PD transcriptome data shows a striking lack of consensus at the level of differentially expressed genes across all tissue sources; however, despite intra-and inter-experimental cell popula- alteration of this pathway in PD. 47 The proteasome is essential for neuronal survival, and a reduction of the proteasome activity has been correlated with brain ageing and neurodegenerative diseases. 48 Furthermore, a large body of evidence shows that Dopamine can induce neural cell death, an effect observed in a dose-and timedependent manner and associated with alteration of proteasome activity. [49][50][51][52][53] Although restricted to transcript level analysis, our data suggest that this alteration might actually reflect the response to LevoDopa therapy through modulation of miR-19b levels.
The relationship between miRs and UDP has been explored in multiple in vivo and ex vivo experimental and clinical setups, including the heart, osteosarcoma cells and neurodegenerative disease. 54,55 Various components of the UDP system were described as miR targets: for example E2 isoforms UE2A, UBE2B, UBE2D3