Involvement of microRNA-135a-5p in the Protective Effects of Hydrogen Sulfide Against Parkinson's Disease

Parkinson's disease is the second most common neurodegenerative disease of high prevalence among people elder than 65 years of age [1]. Parkinson's disease is characterized with selective loss of dopaminergic (DA) neurons in the substantia nigra and a deficiency of dopamine in the striatum [2,3,4,5]. The precise etiology of Parkinson's disease remains unclear. The presently clinically used drug for treating Parkinson's disease is L-DOPA, which only ameliorates the symptoms but could not reverse the process of DA neuronal degeneration [6,7,8,9]. Therefore, development of effective therapeutic drugs for Parkinson's disease is in great need.

Nitric oxide, carbon monoxide and hydrogen sulfide (H₂S) are established gaseous mediators [10]. Endogenous H₂S is primarily produced by cystathionine ɤ-lyase and cystathionine β-synthase, two members from the pyridoxal-50-phosphate-dependent enzyme family [11]. H₂S have a variety of biological functions. In particular, recent reports suggest that H₂S acts as a neuromodulator in brain, in which it is involved in the regulation of the processes of neurodegenerative diseases, including Alzheimer's disease and Parkinson's disease [11]. In line with these evidence, a reduction in endogenous H₂S in the striatum appears to be responsible for the onset of Parkinson's disease and that exogenous H₂S treatment results in attenuation of DA neuronal degeneration in a mouse model of Parkinson's disease [12,13,14,15]. These findings suggest a neuroprotective effect of H₂S during the neural injury of Parkinson's disease. However, the underlying mechanisms remain ill-defined.

Rho-associated protein kinase 2 (ROCK2) is a key factor in the mediation of repulsive environmental signals in the central nervous system, as well as in the regulation pathway for initiating neuro-degeneration. ROCK2 is strongly expressed in the brain and the spinal cord and its expression increases with age. The most well-defined downstream target of ROCK2 is LIM domain kinase, which activates phosphatase and tensin homolog deleted on chromosome ten (PTEN) or Fas. Recently, ROCK2 has been shown to be a major regulator of axonal degeneration, neuronal death and axonal regeneration in the central nervous system, but the regulation of ROCK2, especially by H₂S, has not been reported.

MicroRNAs (miRNAs) are non-coding small RNAs that regulate protein translation through their interaction with 3′-untranslated region (3′-UTR) of the mRNAs of the target genes [16,17,18,19,20,21,22,23,24,25,26,27,28]. Among all miRNAs, miR-135 has been shown to be involved in regulation of cell differentiation, regeneration, and carcinogenesis [29,30,31,32,33,34,35,36]. However, a role of miR-135 in the protective effects by H₂S against Parkinson's disease has not been shown previously.

In the current study, we addressed these question.

All the experimental methods in the current study have been approved by the research committee at the First Affiliated Hospital of Jinzhou Medical University. All the experiments have been carried out in accordance with the guidelines from the research committee at the First Affiliated Hospital of Jinzhou Medical University. All mouse experiments were approved by the Institutional Animal Care and Use Committee at the First Affiliated Hospital of Jinzhou Medical University (Animal Welfare Assurance). Surgeries were performed in accordance with the Principles of Laboratory Care, supervised by a qualified veterinarian.

Male C57BL/6J mice at 12 weeks of age (weighing 25-28 g) obtained from the SLAC Laboratory Animal Co. Ltd (Shanghai, China) were fed on standard pellet chow and water. The mice were randomly assigned into three groups: saline-only treated group (n = 10), which received s.c. injection of saline daily from day 4 to day 8 and i.p. injection of saline daily from day 0 to day 12; MPTP group (n = 10), which received s.c. injection of MPTP (Sigma-Aldrich, St. Louis, MO, USA) daily from day 4 to day 8 at 20mg/kg and i.p. injection of saline daily from day 0 to day 12; MPTP+NaHS group (n = 10), which received s.c. injection of MPTP daily from day 4 to day 8 at 20mg/kg and i.p. injection of NaHS (Sigma-Aldrich) daily from day 0 to day 12 at 5.6mg/kg. At day 12, mice were subjected to rational behavior test and then sacrificed for analyzing brain tissue.

Apomorphine (Sigma-Aldrich) was dissolved in sterile 0.02% ascorbic acid saline solution, and applied at a concentration of 0.5 mg/ml. Since apomorphine is rapidly oxidized when exposed to light at room temperature, preparation was done in a dark environment and covered with aluminum foil. At the end of experimental day 12, the animals' tendency to rotate in response to apomorphine was examined. Briefly, the rats were placed in rotometer bowls and secured to count rotation. Once they were acclimated for at least 10 min, apomorphine (0.5 mg/kg) was given subcutaneously at the back of neck. After 10 min, the number of net rotations (360° contralateral turns) was continuously recorded with a video camera for 1 hour. Only those animals showing at least 7 turns per min in both tests were considered to be successful model creation.

Animals were perfused with 4% paraformaldehyde (PFA) after the experiments were completed. Brain samples were collected and post-fixed in 4% PFA at 4°C overnight, transferred to 30% sucrose in phosphate-buffered saline (PBS) overnight, until the brain had sunk to the bottom of the tube. The brain tissues were then sectioned on a cryostat in 8µm sections. The sections were stained with ABC method (Sigma-Aldrich) and the primary antibody was rabbit-anti-mouse antiglial fibrillary acidic protein (GFAP; Millipore, Billerica, MA, USA). Hematoxylin counterstaining has been performed after GFAP staining, which has been used to realize nuclei of the cells. For quantification, 5 sections that were 50µm away from each other were counted at a magnification of 20X. These sections well covered the entire striatum. Cell counting was performed manually. For Western blot, Protein was extracted from the lysates from the dissected striatum or the cultured cells with RIPA lysis buffer (Sigma-Aldrich) on ice. Protein concentration was determined using a BCA protein assay kit (Bio-rad, China). After transfer, the membrane blots were first probed with a primary antibody, and then incubated with a horseradish peroxidase-conjugated second antibody, after which enhanced chemiluminescent system was applied to visualize the protein antigen. Primary antibodies were rabbit anti-mouse antityrosine hydroxylase (TH, Sigma-Aldrich), anti-ROCK2 and anti-β-actin (Cell Signaling, San Jose, CA, USA). Secondary antibody is HRP-conjugated anti-rabbit (Jackson ImmunoResearch Labs, West Grove, PA, USA). β-actin was used as a protein loading control. The protein levels were first normalized to β-actin, and then normalized to the experimental controls.

HCN-1A is a human cortical neuron cell line, and was purchased from ATCC (ATCC, Rockville, MD, USA), and were cultured in Dulbecco's Modified Eagle's Medium (DMEM) supplemented with 20% fetal bovine serum (Invitrogen, Carlsbad, CA, USA) in a humidified chamber with 5% CO₂ at 37°C. For modification of miR-135a-5p levels in HCN-1A cells, miR-135a-5p or antisense of miR-135a-5p (as-miR-135a-5p) was expressed under a CMV promoter in a plasmid that was used to transfect the cells. A plasmid carrying a null sequence (null) was used as a control for transfection. All the plasmids were obtained from Origene (Beijing, China). The plasmids were transfected into HCN-1A cells at a concentration of 50nmol/l using Lipofectamine-2000 (Invitrogen), receiving 90% transfection efficiency, based on a GFP reporter in the construct. Therefore, no flow cytometry based sorting was performed to enrich the transfected cells.

Total RNA was extracted from clinical specimens or from cultured cells using a miRNeasy mini kit (Qiagen, Hilden, Germany). Complementary DNA (cDNA) was randomly primed from total RNA using the Omniscript reverse transcription kit (Qiagen). Quantitative PCR (RT-qPCR) were performed in duplicates with QuantiTect SYBR Green PCR Kit (Qiagen). All primers were purchased from Qiagen. Data were collected and analyzed using ^2-△△Ct method for quantification of the relative mRNA expression levels. Values of genes were first normalized against β-actin, and then compared to experimental controls.

MiRNAs targets were predicted with TargetSan (https://www.targetscan.org). Luciferase-reporters were successfully constructed using molecular cloning technology. The ROCK2 3'-UTR reporter plasmid (ROCK2 3'-UTR) and ROCK2 3'-UTR reporter plasmid with a mutate at the miR-135a-5p binding site (ROCK2 3'-UTR mut) were purchased from Creative Biogene (Shirley, NY, USA). HCN-1A cells were co-transfected with ROCK2 3'-UTR/ROCK2 3'-UTR mut and miR-135a-5p/as-miR-135a-5p/null by Lipofectamine 2000 (5×10⁴ cells per well). Cells were collected 24 hours after transfection for dual-luciferase reporter assay (Promega, Beijing, China), according to the manufacturer's instructions.

All statistical analyses were carried out using the SPSS 17.0 statistical software package. All values are depicted as mean ± SD and are considered significant if p < 0.05. All data were statistically analyzed using one-way ANOVA with a Bonferroni correction, followed by Fisher's Exact Test for comparison of two groups.

The mice were randomly assigned into three groups: saline-only treated group (n = 10), which received s.c. injection of saline daily from day 4 to day 8 and i.p. injection of saline daily from day 0 to day 12; MPTP group (n = 10), which received s.c. injection of MPTP daily from day 4 to day 8 at 20mg/kg and i.p. injection of saline daily from day 0 to day 12; MPTP+NaHS group (n = 10), which received s.c. injection of MPTP daily from day 4 to day 8 at 20mg/kg and i.p. injection of NaHS, a H₂S donor, daily from day 0 to day 12 at 5.6mg/kg. At day 12, mice were subjected to rational behavior test and then sacrificed for analyzing brain tissue (Fig. 1).

MPTP is converted to 1-methyl-4-phenylpyridinium by monoamine oxidase-B in astrocytes and then transported into DA neurons by dopamine transporter to create neurotoxicity. Thus, we determined the degree of astrocyte activation by MPTP with or without H₂S treatment. We found that MPTP induced significantly increases in the number of antiglial fibrillary acidic protein (GFAP)-positive astrocytes in the mouse striatum, by representative images (Fig. 2A), and by quantification (Fig. 2B). H₂S significantly decreased the effects of MPTP on GFAP-positive cells (Fig. 2A-B). There appeared to be a visual, qualitative change in the morphology of the astrocytes. Characteristic swelling of the cell body and astrocytic arms were detected in the MPTP treated animals, while addition of the NaHS with MPTP did appear to decrease the appearance of swollen GFAP+ cells (Fig. 2A). These data suggest that H₂S alleviates MPTP-induced astrocytic activation in the mouse striatum.

A consistent neurochemical abnormality in Parkinson's disease is degeneration of dopaminergic neurons in substantia nigra pars compacta, leading to a reduction of striatal dopamine (DA) levels. Tyrosine hydroxylase (TH) catalyzes the formation of L-dihydroxyphenylalanine (L-DOPA), the rate-limiting step in the biosynthesis of DA, using co-factors such as O2 and iron. Hence, PD is often considered as a TH-deficiency syndrome of the striatum, and loss of TH has been applied as an indication of neurodegeneration [37,38]. Thus, we examined the levels of TH in mouse brain. We found that MPTP induced significantly decreases in TH, and H₂S significantly decreased the effects of MPTP on TH levels (Fig. 2C). These data suggest that H₂S may alleviate the low dopamine levels induced by MPTP, a characteristic for neurodegeneration.

Next, we examined the effects of H₂S and MPTP on the behavioral symptoms of the mice in an apomorphine-induced rotational behavior test. While MPTP significantly increased the turns of the mice in the test, H₂S significantly reduced the increases in the turns (Fig. 3). Thus, these data suggest that H₂S improves the apomorphine-induced rotational behavior of the mice.

Apomorphine is a short-acting D1 and D2 receptor agonist that can induce rotational behavior in rodents. The authors need to state in the text how they quantified this behaviour. What counted as a rotation? Is a full 360 degree turn necessary? The authors should explain what apomorphine is, how it is used to induce the behaviour, provide appropriate historical references, and explain how they performed the procedure and quantification in the text.

We studied the underlying mechanisms. We specifically found that the levels of ROCK2 were significantly increased in the brain from MPTP-treated mice (Fig. 4A). Moreover, H₂S abolished the effects of MPTP on ROCK2 (Fig. 4A). However, the mRNA of ROCK2 was unaltered (Fig. 4B), implying a possible mechanism of post-transcriptional regulation of ROCK2.

Posttranscriptional control of protein includes different mechanisms. Among all mechanisms, miRNAs play an essential role. We did bioinformatics analysis on the microRNAs that target ROCK2, and selected those that changed levels in the setting of MPTP and NaHS. Specifically, we found miR-135a-5p as a candidate, and MPTP decreased miR-135a-5p levels, which were significantly attenuated by NaHS (Fig. 5A). A specific miR-135a-5p binding site was detected on the 3'-UTR (from 545th to 552th base site) of the ROCK2 mRNA (Fig. 5B). Next, we analyzed whether this binding of miR-135a-5p to ROCK2 mRNA may alter the ROCK2 levels. Thus, we either overexpressed miR-135a-5p, or inhibited miR-135a-5p in a human neuronal cell line HCN-1A, by a miR-135a-5p-expressing plasmid, or a plasmid carrying miR-135a-5p antisense (as-miR-135a-5p), respectively. The HCN-1A cells were also transfected with a null plasmid, to be used as a control for miR-135a-5p modification (Nul). First of all, the alteration of miR-135a-5p levels in HCN-1A cells was confirmed by RT-qPCR (Fig. 5C). HCN-1A cells were then co-transfected with miR-135a-5p-modified plasmids and plasmids carrying a luciferase reporter for 3'-UTR of ROCK2 mRNA or a luciferase reporter for 3'-UTR of ROCK2 mRNA with mutate at the miR-135a-5p binding site (mut). The luciferase activities were determined in these cells, and our data showed that ROCK2 3'-UTR plus miR-135a-5p had the most repression for ROCK2, and the 3'-UTR ROCK2 mutant plus miR-135a-5p had much lower repression. In addition, the 3'-UTR in the presence of as-miR-135a-5p restored expression of ROCK2 (Fig. 5D). These data demonstrate that miR-135a-5p may target 3'-UTR of ROCK2 mRNA to inhibit its translation in neuronal cells. Together, our data suggest that the protective effects of H₂S against Parkinson's disease are partially through miR-135a-5p-regulated ROCK2 (Fig. 6).

The role of the Rho/ROCK/LIMK pathway in neurodegeneration has been acknowledged in recent studies, but the specific regulation of ROCK2, the key factor in this pathway, is unknown. In the current study, we linked ROCK2 levels with the effects of H₂S on neuroprotection in a mouse model for Parkinson's disease. This MPTP model is so far the most widely used rodent model. MPTP is a lipophilic protoxin that rapidly crosses the blood-brain barrier following systemic injection. In brain, MPTP is taken up by the astrocytes and is catalyzed by monoamine oxidase-B to convert to the intermediary, 1-methyl-4-phenyl-2, 3, dihydropyridinium, which is then rapidly and spontaneously oxidized to the toxic moiety, 1-methyl-4-phenylpyridinium to be taken via dopamine transporter into dopaminergic neurons to exert is neural toxic effects. This model mimicked many features of the biological and pathological changes in Parkinson's disease in humans. Using this model, we confirmed the protective effects of H₂S on neurodegeneration.

Moreover, we found that the activated ROCK2 here was significantly inhibited by H₂S. The differential changes in protein and mRNA suggest presence of a posttranscriptional control of ROCK2. Here, we found that miR-135a-5p was involved in the regulation of ROCK2. When a miRNA molecule is attached as a perfect match to a target mRNA, it causes the mRNA degradation therefore the mRNA levels would be diminished by a RT-qPCR, which was not the case in the current study. Our data suggest that a partial interaction between the miR-135a-5p and the 3'-UTR of the ROCK2 gene. In case that the miRNA does not form a perfect match with the mRNA from the target gene, the translation process stops at that point and hence the protein production is reduced.

It has been well established that microglial ROCK leads to phagocytosis of dopaminergic neurons in these models and is upregulated by MPTP induction. In addition, CaMKK-beta-dependent AMPK activation causes NaHS suppression of neuroinflammation [39]. Hence, it is likely that the model shown here may result from NaHS and ROCK's effects upon activated microglia, but not the dopaminergic neurons directly. Then, the recovery after NaHS administration may indicate that there is a feedback loop between CaMKK-beta and ROCK, which would be seemingly resulting from the fact that they both regulate LIM downstream.

To the best of our knowledge, it is the first study to show the involvement of microRNA and ROCK2 in the protective effects of H₂S against Parkinson's disease. Our study should provide new insights into the mechanisms, and should be a basis for generating novel medicine.

This work is supported by Liaoning Province Natural Science Fund Program (no: 2013022032) and Scientific Research Foundation of the First Affiliated Hospital of Liaoning Medical University (no: FYK201206).

The authors have declared that no competing interests exist.