MiR-30a-5p regulates GLT-1 function via a PKCα-mediated ubiquitin degradation pathway in a mouse model of Parkinson’s disease

Background: Glutamate excitotoxicity caused by dysfunctional glutamate transporters plays an important role in the pathogenesis of Parkinson’s disease (PD); however, the mechanisms that underlie the regulation of glutamate transporters in PD are still not fully elucidated. MicroRNAs have been reported to play key roles in regulating the translation of glutamate-transporter mRNA. Methods: We established model of PD 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)-treated mice in vivo and 1-methyl-4-phenylpyridinium (MPP + ) treated astrocyte in vitro. Stereotaxic injection of shRNA in mouse, and miRNA inhibitor/mimic, or antagonist/agonist treated the cell model, Behavioral experiments, glutamic acid uptake, transport activity of synaptosomes, underlying mechanisms and the impact on neuronal survival were assessed. Results We demonstrated that short-hairpin RNA-mediated knockdown of miR-30a-5p ameliorated motor deficits and pathological changes like astrogliosis and reactive microgliosis in a mouse model of PD. Western blotting and immunofluorescent labeling revealed that miR-30a-5p suppressed the expression and function of GLT-1 in MPTP-treated mice and specifically in astrocytes treated with (cell model of PD). Conclusion Both in vitro and in vivo , we found that miR-30a-5p knockdown promoted glutamate uptake and increased GLT-1 expression by hindering GLT-1 ubiquitination and subsequent degradation in a PKCα-dependent manner. Therefore, miR-30a-5p represents a potential therapeutic target for the treatment of PD.

Briefly, mice were placed on the peak of a foam ball (diameter: 2.0 cm) fixed on a pole (diameter: 1.0 cm; length: 50 cm). The time taken to climb from the peak to the bottom of the pole was measured. Each mouse was assessed three times and its mean time was calculated.
Hanging test: Limb movement coordination was assessed using the hanging test, as described previously [48]. Each mouse was suspended three times by its forelimbs hanging on a metal rod (diameter: 1.5 mm) and positioned ~ 30 cm above a box. The time spent hanging was recorded.

Tissue preparation
After the behavioral tests, the mice in each group were euthanized to obtain brain sections. Specifically, the mice were anesthetized using an i.p. injection of 4% chloral hydrate. They then received a thoracotomy and were perfused transcardially with 50 mL 0.9% saline followed by 50 mL 4% paraformaldehyde. The brains were then immediately removed and placed in cold 4% paraformaldehyde at 4°C for 12 h. Next, they were sequentially dehydrated in 20% and 30% sucrose. Finally, the brains were embedded in optimum cutting temperature compound (Sakura Finetek, Torrance, CA, USA) and cut into 10-μm sections with a freezing microtome (Leica, Germany). The sections were placed in a 65°C incubator for 3.5 h and then stored at −80°C for immunohistochemistry and immunofluorescent staining.
For other experiments, mice were euthanized three days after their last MPTP injection.
Specifically, the mice were anesthetized with an i.p. injection of 4% chloral hydrate. and then perfused transcardially with 50 mL of 0.9% saline to remove traces of blood. The midbrain, striatum, cerebral cortex, and hippocampus were removed and stored at −80°C until use.

L-[ 3 H]-Glutamic acid uptake assay
Astrocytes were seeded in 24-well plates at ~5 × 105 cells/cm2. The L-[3H]-Glutamic acid uptake assay was performed as described previously [16]. Briefly, the culture medium was discarded, the astrocytes were washed once with choline solution (150 mM choline chloride, 5 mM KPi, 0.5 mM MgSO 4 , and 0.3 mM CaCl 2 , pH 7.4), and 0.4 μCi/well L-[ 3 H]-Glutamic acid (Perkin Elmer; Boston, MA, USA) was added to each well. The astrocytes were incubated for 10 min at room temperature, and the reaction was stopped by two applications of ice-cold NaCl solution (150 mM NaCl, 5 mM KPi, 0.5 mM MgSO 4 , and 0.3 mM CaCl 2 , pH 7.4). A solution of 1% sodium dodecyl sulfate (SDS) was then added to lyse the cells, and their radioactivity was immediately measured by liquid scintillation counting.
These data are from three separate experiments, each performed in triplicate.

Determining the transport activity of synaptosomes
The midbrain, striatum, cerebral cortex, and hippocampus were individually minced and homogenized with 1:10 (w/v) 0.32 M sucrose solution (0.32 M sucrose, 5 mM HEPES, pH 7.4) and centrifuged at 1,000 × g for 15 min. The supernatant was centrifuged further at 15,000 × g for 30 min at 4°C, and the resulting precipitate was resuspended in 0.32 M sucrose solution to obtain crude synaptosomes. A bicinchoninic acid (BCA) assay was used to measure the protein concentration of each sample, and then the crude synaptosomes were resuspended to a final concentration of 0.5 mg/mL in Kreb's buffer (127 mM NaCl, 3.73 mM KCl, 1.8 mM CaCl 2 , 1.18 mM KH 2 PO 4 , 20 mM NaHCO 3 , 2 mM ATP, 2 g/L D-glucose, pH 7.4). For the glutamate uptake assay, the synaptosome preparations were placed in a 25°C water-bath and 1 μCi L-[ 3 H]-glutamic acid was added for 10 min. The reactions were terminated with 5 mL ice-cold Kreb's buffer. The synaptosomes were washed three times with phosphate-buffered saline (PBS) to remove excess labeled glutamate and then filtered through 0.22-μm pore-size filter paper. The filter papers containing the rinsed synaptosomes were then placed in scintillation vials containing 3 mL scintillation cocktail in the dark overnight. The radioactivity was measured the following day by liquid scintillation counting. The glutamate uptake into the synaptosomes was measured as counts per minute (cpm) per mg protein/min [15]. Data were obtained from three different animals, with three independent cultured synaptosome preparations from each animal, and radioactivity measurements were performed in triplicate.

Reverse transcription real-time quantitative PCR (RT-qPCR)
Total RNA was extracted from total tissue or cells using a Trizol kit according to the manufacturer's instructions (Life Technologies, Carlsbad, CA, USA). Using the PrimeScript RT reagent kit (TaKaRa, Otsu, Japan), RNA was reverse-transcribed to synthesize firststrand cDNA, which was then quantified using the SYBR Premix Ex Taq kit (TaKaRa, Otsu, Japan). For each transcript, a mixture of the following reaction components was prepared to the indicated end-concentration: forward primer (10 μM), reverse primer (10 μM), and SYBR Green Premix Ex Taq. Fluorescence was detected on a Corbett Research RG-6000 real-time PCR machine (Corbett Life Science, Sydney, Australia). Oligonucleotide primers for Slc1a2 and β-actin were designed as follows: Slc1a2 forward 5´-CGATGAGCCAAAGCACCGAA-3´ and reverse 5´-CTGGAGATGATAAGAGGGAGGATG-3´; and βactin forward 5´-GGACTCCTATGTGGGTGACG-3´ and reverse 5´-GTTGGCCTTAGGGTTCAGGG-3´. Slc1a2 mRNA levels were normalized to β-actin expression. Data are from three separate experiments, and each was performed in triplicate.

Immunohistochemistry assay
Immunohistochemistry was performed as described previously [15]. Frozen brain sections were recovered at room temperature and antigen retrieval was performed using 0.01 M citrate buffer for 40 min. The sections were permeabilized with 0.3% Triton X-100 and blocked with 5% bovine serum albumin (BSA) for 10 min. Then, the sections were incubated with primary antibodies for 15-18 h at 4°C. After washing with PBS three times for 5 min, the sections were incubated with secondary antibodies for 2 h at 37°C.
Antibody-peroxidase complexes were revealed by incubating the sections with 3, 3diaminobenzidine peroxidase substrate (Maixin, Fujian, China). Finally, the sections were mounted with resin and images were acquired using an Olympus 1 × 81 FV1000 Laser Scanning Confocal Microscope (Shinjuku, Tokyo, Japan). Image analysis was performed using Image-Pro Plus, version 6.0 (IPP 6.0, Media Cybernetics, Bethesda, MD, USA).

Immunofluorescent double-staining
Frozen brain sections were recovered at room temperature, fixed in 4% paraformaldehyde, and then rinsed with PBS three times for 5 min. Next, the sections were permeabilized with 0.3% Triton X-100 and blocked with 5% BSA for 10 min. For the colocalization experiments, the brain slices were incubated with primary antibodies (anti-GLT-1 and anti-ubiquitin or anti-EEA1) for 15-18 h at 4°C, rinsed with PBS three times, and incubated with FITC-conjugated goat anti-mouse IgG (to bind anti-GLT-1) and TRITCconjugated goat anti-rabbit IgG (to bind anti-ubiquitin) or Cy3-conjugated donkey antigoat IgG (to bind anti-EEA1) for 2 h at 37°C. DAPI was used to stain cell nuclei. Images were acquired using a laser scanning confocal microscope (Zeiss, Germany) and analyzed using Image-Pro Plus, version 6.0 (IPP 6.0, Media Cybernetics, Bethesda, MD, USA). To investigate the different degradation pathways of GLT-1, only GLT-1 fluorescence that colocalized with ubiquitin (for the ubiquitin-proteasome pathway) or that colocalized with EEA-1 (for the endocytosis-mediated pathway) were measured under each condition.

Isolation of membrane-bound proteins by cell-surface biotinylation
Cell-surface expression of glutamate transporters was examined using the membraneimpermeable biotinylation reagent EZ-LinkTM Sulfo-NHS-SS-Biotin (Thermo Scientific, Waltham, MA, USA) as described previously [49]. Astrocyte or brain-tissue samples were washed twice with ice-cold PBS (pH 8.0) and incubated with 2.5 mL EZ-Link Sulfo-NHS-SS-Biotin (0.5 mg/mL in PBS) in two successive 20-min incubations on ice with gentle shaking.
Then, the samples were washed twice with 100 mM glycine to remove non-reacted biotinylation reagent and lysed on ice for 20 min in 750 μL cell-lysis buffer containing a protease inhibitor mixture. After centrifugation at 12,000 rpm for 20 min, debris was removed. Supernatants were transferred to new tubes, and 200 μL streptavidin-agarose beads was added to bind the biotin-labeled cell membrane proteins. The samples were then centrifuged at 3,000 rpm for 1 min at 4°C, and the supernatant was discarded. This supernatant was washed three times with ice-cold lysis buffer, then washed once with icecold PBS and centrifuged for 1 min at 4°C. The precipitate from this centrifugation contains membrane-bound proteins, which was diluted with protein-loading buffer and heated at 55°C for 30 min prior to electrophoresis and western blotting.

Western blotting
First, cells or brain tissues were lysed with radio-immunoprecipitation assay (RIPA) buffer (Beyotime, Shanghai, China) containing 1 mM phenylmethylsulfonyl fluoride (PMSF) (Beyotime, Shanghai, China) for 30 min. Then, the homogenates were centrifuged at 14,000 × g for 15 min and the supernatant was subjected to a BCA assay to determine the protein concentration. Samples of equal protein content were diluted with loading buffer and denatured at 100°C for 10 min. Then, the protein samples underwent SDSpolyacrylamide gel electrophoresis (SDS-PAGE) on a 12% acrylamide gel and were transferred to polyvinylidene difluoride (PVDF) membranes. To block non-specific protein binding, the membranes were incubated in 5% BSA for 1.5 h at room temperature with gentle shaking. Then, the membranes were incubated with primary antibodies at 4°C for 15-18 h, washed with TBST (Tris-buffered saline containing 0.1% Tween-20) three times, and incubated with HRP-conjugated secondary antibodies for 1 h at room temperature.
Finally, the membranes were washed again with TBST and enhanced chemiluminescence (Millipore, MA, USA) was used to visualize the protein bands on a Tanon imager system (Shanghai, China). The bands were quantified using ImageJ software (NIH Image, Bethesda, MD). Actin, GAPDH, or integrin-β1 immunoreactivity was used as the loading control for each protein. The detail information of antibodies were in table 1.

Dual-luciferase reporter gene experiments
To verify the regulation of Slc1a2 by miR-30a-5p binding, dual-luciferase reporter vectors were generated by cloning the entire wild-type 3'-untranslated region (3'-UTR) or a mutant 3'-UTR of Slc1a2 into the psiCHECKTM-2 vector (Promega, WI, USA) at the cleavage sites with NotI and XhoI. Then the luciferase reporter constructs containing the 3'-UTR of mouse Slc1a2 or its seed-sequence mutants were co-introduced with the miR-30a-5p mimic (100 nM) into 293T cells using Lipofectamine 2000 (Invitrogen). Using this setup, Renilla luciferase fluorescence (Rlu) would decrease if the Slc1a2 3'UTR was bound by the miR-30a-5p mimic. Fluorescence from the Firefly luciferase gene (Flu), also present on the vector, would remain the same in all conditions and is used to normalize fluorescence intensity. After 48 h, all target validation assays were performed with the dual-luciferase reporter system (Vigorous Biotechnology, Beijing, China) according to the manufacturer's instructions. Luciferase activity was determined by the Dual-Luciferase® Reporter Assay System (Promega) and presented as the relative intensity of Rlu/Flu.

Fluorescent in situ hybridization (FISH)
MiR-30a-5p FISH probes were obtained from FOCOFISH (Guangzhou, China). The sequence of the miR-30a-5p antisense probe was 5'DIG-CTTCCAGTCGAGGATGTTTACA-3'DIG. Briefly, the astrocyte cultures or 10 μm-thick brain sections were incubated at 37°C with solution A for 20 min, solution B for 20 min, and 4% paraformaldehyde for 15 min. Then, the cells or sections were incubated with 200 μL hybridization buffer in a 55°C water bath for 2 h.
In the meantime, the probes were diluted 50 times, denatured at 85°C for 3 min, balanced at 37°C for 2 min, and then added to each slide or coverslip. Hybridization was carried out in a pre-warmed, humidified box kept in a 42°C incubator for 15-18 h. Finally, DAPI was used to stain cell nuclei and the sections or cells were examined using a laser scanning confocal microscope (Zeiss, Germany).

Statistical analysis
All data are expressed as mean ± standard error of the mean (SEM). Statistical analysis of the data was performed with the Student's t-test to compare two groups or the one-way analysis of variance (ANOVA) with post-hoc multiple comparison and Dunnet testing with SPSS 20.0 software (SPSS Inc., Chicago, IL, USA). Results were considered statistically significant at *p < 0.05 (and **p < 0.01).

SNpc of MPTP-treated mice.
Our previously conducted high-throughput sequencing revealed that miR-30a-5p was upregulated significantly in the midbrain of MPTP-treated mice. Bioinformatics analysis with TargetScan, miRanda, and miTarget identified Slc1a2 (the gene of GLT-1) as a potential target gene of miR-30a-5p. Therefore, we chose miR-30a-5p for further study.
We verified by RT-qPCR that miR-30a-5p was increased in MPP + -treated astrocytes and in the SNpc of MPTP-treated mice We also used a FISH assay to determine miR-30a-5p levels; the results showed that miR-30a-5p was elevated significantly in MPP + -treated astrocytes (Fig 1b, c, p < 0.01) and in the SNpc of MPTP-treated mice (Fig 1d, e, p < 0.01). Taken together, the above results led us to hypothesize that miR-30a-5p might contribute to PD symptoms by regulating the expression and function of GLT-1.

MiR-30a-5p regulated the expression and function of GLT-1 in vitro
In order to test whether miR-30a-5p affects the expression and function of GLT-1 in vitro, we detected the expression of GLT-1 in both total and membrane-bound protein fractions of MPP + -treated astrocytes. Our data showed that MPP + can decrease the expression of total GLT-1 and membrane-bound GLT-1 in astrocytes. Nevertheless, when MPP + -exposed astrocytes were treated with an miR-30a-5p inhibitor, the expression of GLT-1 was significantly increased compared with those exposed to the scramble control miRNA (p < 0.05). Similarly, treatment with an miR-30a-5p mimic exacerbated the reduction of GLT-1 caused by MPP + exposure in both the total and membrane-bound protein fractions (p < 0.05 ) (Fig. 2a-d). We further explored whether miR-30a-5p affected the function of GLT-1 in MPP + -treated astrocytes by measuring glutamate uptake under the different treatment conditions. We found that the miR-30a-5p inhibitor improved the ability of glutamate uptake by MPP + ( p < 0.05), and the miR-30a-5p mimic aggravated the abnormal function of GLT-1 caused by MPP + ( p < 0.05, p < 0.01) (Fig. 2e). Finally, we tested if the above effects were also present at the mRNA level. Similar to our results with the GLT-1 protein, MPP + decreased Slc1a2 mRNA levels but the miR-30a-5p inhibitor was able to reverse this change (p < 0.05) (Fig. 2f). Thus, our findings indicated that miR-30a-5p can regulate the expression and function of GLT-1 at both the mRNA and protein levels.

MiR-30a-5p regulated the expression and function of GLT-1 in vivo
In order to determine whether GLT-1 can be affected by miR-30a-5p in vivo, we used viral-mediated knockdown of miR-30a-5p in MPTP-treated mice. First, we constructed the lentiviral vector (LV) containing miR-30a-5p shRNA and used a FISH assay to test its effect in mice. Our results showed that this LV suppressed the expression of miR-30a-5p by 50% (p < 0.01) (Fig. S1a,b). After lentiviral injection into the SNpc and MPTP administration, we isolated tissue from the midbrain, striatum, cerebral cortex, and hippocampus and retrieved a total protein fraction, membrane protein fraction, complete synaptosome fraction, and total RNA. These different samples were used to detect the GLT-1 protein (p < 0.05, p < 0.01 ) (Fig. 3a-d, Fig. S2a,b ), glutamate uptake (p < 0.01) (Fig. 3e, Fig. S3ac), and mRNA expression levels (p < 0.05) (Fig. 3f). We found that the expression of total and membrane GLT-1, Slc1a2 mRNA levels, and the glutamate-uptake function of GLT-1 were increased significantly in MPTP-treated mice with shRNA/LV-mediated blockade of miR-30a-5p compared with those MPTP-treated mice injected with the scramble shRNAcontaining LV (p < 0.05, p < 0.01). Taken together, these results confirmed our predictions that the miR-30a-5p could regulate GLT-1 in MPTP-treated mice in vivo, and that the expression and function of GLT-1 could be enhanced by knockdown of miR-30a-5p in MPTP-treated mice.

MiR-30a-5p does not bind the 3'-UTR of Slc1a2 mRNA
Our finding that miR-30a-5p can regulate the expression and function of GLT-1 in vitro and in vivo led to the prediction that Slc1a2 may be a target gene of miR-30a-5p (Fig. 4a).
In addition, our preliminary bioinformatics analyses had suggested the same idea. We tested this using dual-luciferase reporter gene experiments, where a decrease in the ratio of Rlu/Flu fluorescence intensity would imply that miR-30a-5p binding caused Slc1a2 mRNA to be degraded rather than translated. However, there was no difference in relative fluorescence between any of the groups, which indicated that there was no direct binding site between miR-30a-5p and Slc1a2 (p > 0.05) (Fig. 4b). This unexpected finding suggested that the regulation of GLT-1 by miR-30a-5p in MPTP-treated mice occurs indirectly via one of the cell's other signaling pathways.
Next, we tested whether miR-30a-5p used these same two pathways to regulate the expression of GLT-1. Western blots for GLT-1 in total and membrane fractions showed that the PKCα antagonist (staurosporine) reversed the decrease in membrane-bound GLT-1 caused by MPP + , but did not affect total levels of GLT-1 (p < 0.01) (Fig. 5a-d). In contrast, the mTOR agonist (MHY1485) was unable to restore any of the GLT-1 expression deficits in MPP + -treated astrocytes; both total GLT-1 (p < 0.01) (Fig. 5e, g) and membrane-bound GLT-1 (p < 0.01) (Fig. 5f, h) levels were still decreased. Taken together, these results suggest that miR-30a-5p can regulate both the function and expression of GLT-1 through the PKCα pathway.

Ubiquitin-mediated degradation of GLT-1 was hindered by miR-30a-5p knockdown or PKCα antagonist (staurosporine)
Since we established that miR-30a-5p can regulate the expression of GLT-1 in vitro and in vivo, we wanted to test if this regulation was carried out by hindering the degradation pathways of GLT-1. To this end, we used immunofluorescent double-staining in MPP +treated astrocytes and of the SNpc of MPTP-treated mice to detect ubiquitin and early endosome antigen (EEA-1). Ubiquitin and EEA-1 would serve as markers for the ubiquitinproteasome degradation pathway and the endocytosis degradation pathway, respectively.
Next, we wanted to test whether miR-30a-5p regulated GLT-1 ubiquitination through PKCα or mTOR. We once again used immunofluorescent double-staining to determine the effect of the PKCα antagonist (staurosporine) or mTOR agonist (MHY1485) on the ubiquitin-mediated degradation pathway. The results showed that staurosporine reversed the increase in ubiquitination caused by MPP + ( p < 0.01) (Fig. 6e, f), but MHY1485 did not (p < 0.01) (Fig. 6g, h). Taken together, our data indicated that miR-30a-5p can regulate the ubiquitin-mediated degradation of membrane-bound GLT-1 through the PKCα pathway.

MPTP-treated mice.
In order to assess the functional outcome of miR-30-5p knockdown in MPTP-treated mice, behavioral tests were performed to determine motor capacity. ShRNA-mediated miR-30a-5p knockdown in MPTP-treated mice shortened their pole-climbing time and prolonged their holding time in the hanging test (p < 0.01) (Fig. 7a, b). These results clearly showed that miR-30a-5p knockdown can alleviate motor deficits in MPTP-treated mice.
Next, we used immunohistochemistry to test whether miR-30a-5p knockdown could also ameliorate the pathological changes observed in MPTP-treated mice. In line with our predictions, the expression of α-Syn was reduced significantly in the SNpc of MPTP-treated mice injected with miR-30a-5p shRNA compared with MPTP and scramble RNA-treated mice (p < 0.01) (Fig. 7c). Furthermore, miR-30a-5p knockdown reversed the MPTP-induced reduction in the number of TH-positive neurons in the SNpc and TH-positive neuron density in the striatum (p < 0.01) (Fig. 7d, e). These results indicated that miR-30a-5p knockdown had a neuroprotective effect in MPTP-treated mice.
Since astrocytic activation and microgliosis are some of the hallmark characteristics of PD [14,15,55,56], we tested the effect of miR-30a-5p knockdown on these phenotypes in the SNpc. We found that the MPTP-induced increases in GFAP expression (a marker of activated astrocytes) and Iba-1 (microglial marker) were ameliorated by miR-30a-5p shRNA (p < 0.01) (Fig. 7f, g). So, we can conclude that miR-30a-5p knockdown attenuated astrogliosis and reactive microgliosis. Subsequently, improved the expression and function of GLT-1 in an MPTP model of PD in vivo were increased.

Discussion
Glutamate transporters can remove extracellular glutamate from the synaptic cleft to prevent glutamate excitotoxicity [57]. Glutamate excitotoxicity mediated by dysfunctional glutamate transporters is a crucial part of the pathogenesis of PD; many studies have shown that increasing glutamate-transporter expression and function can improve the progress of neurodegenerative diseases, including PD [14-16, 58, 59]. In our previous studies, we reported that ceftriaxone improved MPP+-induced dysfunction of GLT-1 in astrocytes by promoting the membrane expression of GLT-1 and attenuated the apoptosis of astrocytes by suppression of the NF-κB/JNK/c-Jun signaling pathway [16]. We also found that rapamycin can preserve the astrocytic expression of glutamate transporters and glutamate reuptake, which is protective against MPTP [ 14 ]. Evidence from our group and others confirm that glutamate transporters play an important role in the development of PD, but its mechanism of regulation had not been well elucidated thus far.
It has been reported that the expression of protein is regulated by microRNAs, and several studies have investigated miRNAs in PD. He Q et al. found that MPP+ is taken up by microglia through organic cation transporter 3 (OCT3) and downregulates miR-7116-5p, which is a miRNA target of tumor necrosis factor alpha (TNF-α) [36]. They reported that this MPP+-induced downregulation of miR-7116-5p elicited the production of TNF-α in microglia [36]. Furthermore, they found an improvement in miR-7116-5p expression in the microglia of MPTP-treated mice where the production of TNF-α and activation of glia was inhibited, which prevented the loss of DA neurons [36]. Kim J et al. found that microRNAs miR-27a and miR-27b can regulate PINK1 expression at the translational level and autophagic clearance of damaged mitochondria [60]. Their data further supported a novel negative regulatory mechanism of PINK1-mediated mitophagy by miR-27a and miR-27b [60].
Other studies have investigated miRNAs in the context of glutamate excitotoxicity. Aw et al. found that miR-263a limited glutamate-receptor levels in a subset of glia by targeting Grik, Nmdar1, and Nmdar2 in drosophila, protecting the cells from excitotoxicity [61].
They also found that loss of miR-263a led to the death of a specific population of ensheathing glia, which was correlated with severe climbing defects in C57BL/6 mice [61].
A different study by Harraz et al. showed that neuronal miR-223 regulated glutamate receptors; specifically, miR-223 can control neuronal excitability in response to glutamate by targeting the AMPA receptor subunit GluR2 and the NMDA receptor subunit NR2B, both of which are glutamate receptors [62]. The presynaptic activity of miR-1000 has also been shown to modulate synaptic plasticity via regulation of vesicular glutamate transporter expression [63].
Despite the number of reports about the role of microRNAs in glutamate excitotoxicity, not many detail the regulatory mechanisms of miRNAs on glutamate transporters. Morel et al. found that exosomes isolated from neuron-conditioned medium contained miR-124a and are directly internalized into astrocytes to increase astrocytic GLT1 levels [64].
Moreover, miR-124a is decreased in the spinal-cord tissue of SOD1 G93A (mouse model of amyotrophic lateral sclerosis [ALS]), and exogenous delivery of miR-124a in vivo prevented further pathological loss of GLT-164. Yang Z et al. found that GLT-1 gene is a target of miR-107, and that miR-107-induced suppression of GLT-1 caused the accumulation of glutamate after ischemic stroke [65]. A recent study from our group provided evidence that miR-543-3p regulated the glutamate transporter GLT-1 in PD [41].
Furthermore, our data showed that inhibiting miR-543-3p in vivo and in vitro improved the expression and function of GLT-1 and improved the motor defects and pathological changes in MPTP-treated mice [41]. Until our current study on miR-30a-5p, miR-543-3p was likely the first and only identified miRNA that targeted GLT-1 in PD.
Our current study investigated the potential role of miR-30a-5p in the progression of PD.
We found that miR-30a-5p regulated the expression and function of GLT-1, and miR-30a-5p knockdown improved the abnormal expression and function of GLT-1 in vitro and in vivo in a PD mouse model. Contrary to our prediction, a dual-luciferase reporter assay showed that Slc1a2 was not a target gene of miR-30a-5p. This led us to the discovery that miR-30a-5p indirectly regulated GLT-1 by the PD-related signaling molecule PKCα. Activation of PKCα appears to be a crucial step for mediating the biological effects of miR-30a-5p, at least partly. PKCα does not contain a binding site for miR-30a-5p, therefore, how exactly it is regulated by miR-30a-5p requires further research. Others have reported that PKC promoted the degradation pathway of GLT-1 [15,[66][67][68]. Specifically, PKC activation induced the ubiquitination of GLT-1 [69], which mediated its interaction with the endocytic machinery for degradation [68]. Our group also used co-immunoprecipitation assays to show that PKC activation can regulate the ubiquitination of GLT-1 [15].
In this study, we used immunofluorescent double-staining to detect the ubiquitinmediated degradation pathway and the endocytosis degradation pathway in MPP + -treated astrocytes and in the SNpc of MPTP-treated mice. We found both degradation pathways of GLT-1 upregulated in MPTP-treated mice, and miR-30a-5p shRNA reversed this upregulated degradation only by the ubiquitin pathway, not the endocytosis pathway (Fig. 6a-d).
Dysfunction of the ubiquitin-proteasome degradation pathway is important in the pathogenesis of PD; an increasing number of reports are implicating ubiquitin in PD. α-Syn inclusions in Lewy bodies can also be ubiquitinated in PD [70]. Parkin (PARK2), another gene mutated in familial PD, is an E3-ubiquitin ligase that plays an important role in selecting proteins for ubiquitination [71,72]. Our group previously found that Nedd4-2mediated ubiquitination hindered glutamate-transporter trafficking between the membrane and cytoplasm, consequently decreasing the expression and function of glutamate transporters in the membrane [15]. Accordingly, Nedd4-2 knockdown promoted glutamate function and glutamate-transporter expression in vitro and in vivo, while also improving behavioral and pathological changes in MPTP-treated mice [15] These parallel lines of evidence from the literature strongly suggest that ubiquitination of GLT-1 is important in PD.
In this study, we showed that miR-30a-5p knockdown ameliorated the movement defects in MPTP-treated mice, did not increased the number of TH-positive neurons in the SNpc, and augmented the TH cell density in the striatum, but prevented the dopaminergic neuronal loss. MiR-30a-5p knockdown also decreased the α-Syn density in the SNpc of MPTP-treated mice. These results suggest that miR-30a-5p knockdown protected the animal from the effects of MPTP toxicity and we conclude that this protection is mediated by miR-30a-5p's regulation of GLT-1.

Conclusion
We demonstrated that miR-30a-5p can regulate the expression and function of GLT-1 by ubiquitination of these glutamate transporters through the PKCα pathway in vitro and in vivo (Fig 8). We also showed that miR-30a-5p knockdown ameliorated movement deficits and attenuated pathological change in MPTP-treated mice via increasing glutamate transporters. We conclude that miR-30a-5p should be considered a promising therapeutic target for the treatment of PD.

Availability of data and materials
All data supporting the conclusions of this study are included in the article.

Ethics approval and consent to participate
The animals used in this study were treated in accordance with protocols approved by the Institutional Animal Care and Use Committee of Southern Medical University, China. Table   Table 1. List of antibodies used in experiment. MiR-30a-5p regulates the expression and function of glutamate transporter GLT-1 in vitro. Primary cortical astrocytes were incubated with miR-30a-5p inhibitor, inhibitor control, miR-30a-5p mimic, or mimic control for 24 h, then incubated with MPP+ for 24 h, after which the following assays were conducted. Western blots of (a) total GLT-1 protein and (b) membrane-bound GLT1. Quantification of the immunoreactivity in (a) and (b) are displayed in (c) for total GLT-1 and (d) for membrane-bound GLT-1 (e) Slc1a2 mRNA expression was detected using RT-qPCR. The effect of miR-30a-5p on signaling protein expression and transporter activity of GLT-1. (a) The coding sequence (CDS) of GLT-1 mRNA is followed by its 3'-UTR sequence, where the predicted target site of miR-30a-5p lies between amino acid 1976 and 1983. Sequence alignment of the Slc1a2 3'-UTR in mouse (Mus.musculus, Mmu) and miR-30a-5p. (b) Dual-luciferase reporter assay to determine binding of the Slc1a2 3'-UTR by miR-30a-5p.293T cells were cotransfected with reporter vectors containing either a wildtype or mutant 3'-UTR of GLT-1 (Slc1a2-3'-UTR and Mut-Slc1a2-3'-UTR, respectively) and a miR-30a-5p mimic. The cells were harvested 24 h later and fluorescence was detected. Empty vector (psicheck2) and scrambled shRNA (150608-NC) were used as negative controls for the luciferase reporter vectors and the miR-30a-5p mimic. C57BL/6 mice were injected in the SNpc with miR-30a-5p shRNA or scrambled shRNA, then received one i.p. injection of MPTP-HCl or saline daily for five days, and were euthanized on day 3 after the final MPTP injection. Midbrain tissue was prepared for western blots of total PI3K, PKCα, STAT3, and mTOR protein, as well as their active, phosphorylated (p-) forms. (c) Images of the total and phosphorylated protein bands under the different conditions: "Ctrl" (scrambled shRNA+saline), "miR-30a-5p shRNA" (miR-30a-5p shRNA+saline), "MPTP" (MPTP+scrambled The potential mechanism of miR-30a-5p regulates the expression and function of GLT-1.

Supplementary Files
This is a list of supplementary files associated with the primary manuscript. Click to download.