Protective role of mRNA demethylase FTO on axon guiding molecules of nigro-striatal projection system in manganese-induced Parkinsonism

Parkinson's disease (PD) is a neurodegenerative disease caused by environmental and genetic factors. Manganese (Mn) exposure is a major environmental cause of PD. Cellular and molecular mechanism of Parkinsonism caused by Mn has not been explored clearly. In addition, patients with Mn-induced Parkinsonism show poor therapeutic response to levodopa. Therefore, there is need to explore the mechanisms underlying neurotoxicity of Mn exposure. Selective ephrin-B2 activation in the striatum rescues defects in motor function. with inhibition

guidance and neural projection. Ephrins promotes developmental axon path nding by modulating actin cytoskeleton dynamics [16]. Researchers report that ephrins are abundantly expressed in the striatum [17]. Sieber et al [18] reported that ephrin signaling is important in normal connectivity and function of midbrain DA neurons. The present study postulated that Mn regulation of the motor function may be associated with the function of ephrins signaling. However, the speci c molecular mechanisms were not elucidated.
Hundreds of RNA modi cations have been reported, and N6-methyladenosine (m 6 A) is one of the most ubiquitous internal changes that regulate gene expression at the post-transcriptional level [19]. m 6 A modi cation regulates several phases of RNA metabolism, such as RNA translation, decay and splicing [20,21]. The latest reported m 6 A demethylase FTO shows that m 6 A modi cations can be dynamically regulated [22]. However, the targets and features of m 6 A modi cation in CNS have not been explored. m 6 A modi cation can regulate neuronal function such as motor function [23], axon regeneration [24], and synapse function [25]. In addition, Hess et al reported that inhibition of FTO gene impairs neuronal activity and motor function [26]. However, the function of m 6 A eraser FTO on axon guiding molecules of nigro-striatal projection system in Mn-induced Parkinsonism pathogenesis has not been fully elucidated.
The ndings of the current study show that FTO is implicated in regulation of motor function and mediation of axon guiding molecules against Mn-induced Parkinsonism in vitro and in vivo. These ndings show the speci c mechanism of Mn-induced motor dysfunction mediated by FTO and ephrins.
Moreover, these ndings show that re-expression of FTO and ephrin-B2 improve motor dysfunction after Mn exposure.

Materials And Methods
Reagents  For AAV5 viral injection, mice were anaesthetized with 5% chloral hydrate and positioned on a stereotaxic apparatus. Bilateral holes were drilled into the cranium at 0.5 mm foreside and 2.8 mm outside the bregma for administration of treatments into the striatum regions. An AAV5 expressing either FTO or ephrin-B2 (AAV5-FTO or AAV5-ephrin-B2) or hrGFP (AAV5-NC) was obtained through stereotactic injection.
A glass cannula stuffed with a virus was placed on the striatum (-3.5 mm), and 0.2 µL virus was administered at a rate of 0.1 µL/min. The scalp was then sealed. Follow-up exposure was carried out 10 days after virus injection, and then behavioral experiments or other tests were carried out.
After the behavioral experiment, all the mice were sacri ced through decapitation under anesthesia. Striatum of 20 mice in each group was separated in an ice bath. 4 out of the 20 mice were used for histological and ultrastructural analysis of the striatum. Among the other 16 mice from each group, 4 were used for the detection of Mn and DA levels. 2 were used for immuno uorescence detection. 2 were used for determination of ephrin-B2 m 6 A mRNA expression, 2 for electrophysiological recording, and 2 mice were used for uorogold retrograde tracing. Striatum for the other mice was used for determination of target mRNA level, protein, and other indicator levels.
Lentiviral generation and infection.

Cell viability analysis
The cells were seeded into 96-well plate at 5×10 3 /well and cultured in 10% FBS containing DMEM for 24h. Then, cells were treated in 100 µL of DMEM according to the above groups. After incubating for 24 h, add 10 µL of CCK-8 reagent to every well, and incubate for extra 4h. Absorbance of every well was measured at 450 nm. Each well has two duplicate wells to con rm the reproducibility.
Automated observation of axon growth with the live cell imager Cells and axon growth trends were observed using Lionheart FX Automated Live Cell Imager (BioTeK, USA). The relevant treatment was given to every group following the study design. The control group was used to adjust the photo parameters. The duration of observation was 24 h, and data were recorded every hour.

Rotarod test
Mice were positioned on a Rotary rod instrument (UGO 47650, Italy) and skilled for 5 min at the 10 rpm speed for 3 days. For testing, mice were positioned on a rod that turned around accelerating velocity beginning with 5 rpm and accelerating up to 30 rpm in 60 s. Experiments were performed three times.

Forced running wheel test
Mice were positioned on a wheel fatigue meter (YLS-10B, Yiyan, China) to explore their staying power capabilities. Mice were skilled at 30 rpm for 10 min, and the fatigue running distance was determined.
The wheel fatigue meter was stopped if mice fail to run within 10 sec after the electric shock (1 mA) and the wheel fatigue meter was started after mice rest for 30 sec. The experiments were performed in triplicates.

Open eld test
The open eld experimental analysis system for mice was provided by China Medical University. The mice were placed in the laboratory to adapt to the environment for 10 minutes. This experiment is performed in a quiet, low-light environment. The indoor temperature was maintained at about 26°C. The mice were positioned in the open eld and the mice activity was observed within 5 min. Observation indicators include total horizontal movement distance and horizontal movement speed. Clean up the labyrinth after the experiment of each mouse, and try to avoid external interference that will cause unnecessary adverse effects on the experimental results.

Gait analysis
In order to enable all mice to walk to the end of the runway at a uniform speed without external stimulation, the mice were trained for 3 days. Data was collected on the 4th day and recorded 3 times for each mouse. The collected data was analyzed with Cat Walk XT Version10.6, and the average value of each index was taken. The related indexes are run duration, average speed, swing speed, and maximum variation.

Determination of Mn levels
The inductively coupled plasma mass spectrometry (ICP-MS, Agilent 7700x, USA) was used for analysis of Mn levels in striatum of mice. Striatum was predigested, and 0.1, 1, 5, 10, 50, and 100µg/L series of Mn standard solutions were prepared for future use. The content of Mn in the striatum was then determined with internal standard method.

Measurement of DA levels
Total DA levels in the striatum were quanti ed using ELISA kits of mice DA (Baolai, Jiangsu, China) following the manufacturer's directions. The sample value within the detectable range was used for analysis. Experiments were performed in triplicates.
Immuno uorescence of FTO, ephrin-B2, TH, DARPP32, and DAT Expression and localization of FTO and ephrin-B2 were determined in striatum and neurons, respectively, through immuno uorescence. In addition, immuno uorescence triple staining of TH, DARPP32 and DAT in striatum was used to explore whether nigro-striatal projection neurons were damaged. Samples were dehydrated with xylene and alcohol, and tissue sections were steamed with sodium citrate for antigen retrieval. Samples were blocked in 10% Goat serum for 30 min at 25°C. Samples were then treated with Then, samples were mounted with mounting medium for further study. Laser scanning confocal microscope (A1R, Nikon, Japan) was used to capture images.

Hematoxylin and eosin (HE) staining
After the mice were perfused with 4% paraformaldehyde, the striatum tissue was removed. The striatum of mice was serially dehydrated in xylene and ethanol. The striatum was sectioned at 5 µm thickness after para n embedded. Then, HE staining was done for morphological observation with high resolution panoramic imaging system (CS2, Leica, Germany).

Golgi-cox staining
Brains of mice were removed quickly from the skull after euthanizing. After washing, the brains were stained with the FD Rapid GolgiStain™ kit (MD, USA). Firstly, the brains were incubated in the impregnation solution (A and B) for 14-15 days. Afterwards, the brains were treated with Solution C for the 48-72 h. Microtome Cryostat (1950, Leica, Germany) was used to cut 200 µm slices. Slices were mounted on a gelatin-coated slides, then dehydrated, stained, and coverslipped. Then, dendritic spines was observed with high resolution panoramic imaging system (CS2, Leica, Germany).

Nissl staining
Tissue sections were sequentially stained with series alcohol dehydration. The sections were then treated with 1% thionine for 1 h, then the samples were washed with distilled water. 70% alcohol color separation was performed for several sec to several minutes. The sections were dehydrated with 70%, 80%, 95%, and 100% alcohol, re-spectively, with 2 min for each. Next, the sections were washed with anhydrous ethanol twice (with 5 min for each time), followed by washing with xylene twice (with 10 min for each time).
Finally, the sections were observed with high resolution panoramic imaging system (CS2, Leica, Germany).

Fluorogold retrograde tracing
Two mice in each group were injected with 2% FG solution into the striatum via brain stereotaxic injection. After 5 days of survival, the mice were anesthetized and perfused with 4% paraformaldehyde solution. The brain was taken out and xed with 4% paraformaldehyde solution for 4 hours. Then the brain tissue was placed in a 20% sucrose solution at 4°C overnight. The next day, freeze sectioning was performed, and the section thickness was 30 µm. The sections were washed with 0.01 mM PBS and xed in 4% paraformaldehyde for 10 min. After washing 3 times with PBS, samples were mounted with mounting medium for further study. Laser scanning confocal microscope (A1R, Nikon, Japan) was used to capture images.

Electron microscopy studies
After the mice were perfused with 4% paraformaldehyde, the striatum was xed with 2.5% glutaraldehyde for 1h. A 1-mm-thick slice was made and xed overnight. Post xation of slice was carried out in 1% OsO4 for electron microscopy observation. Then, tissues were embedded in spur resin after dehydrated in series of acetone. Blocks were photographed in a transmission electron microscope (JEM-1400-FLASH, Japan) after stained with uranyl acetate and lead citrate.

Electrophysiological recording of striatum slices
Spontaneous discharge frequency in the striatum was recorded in 300 µm slices. Slices were cut using vibratome in a modi ed arti cial cerebrospinal uid (mACSF) saturated with 5% CO 2 /95% O 2 . Slices were then placed on an incubation chamber with normal ACSF and incubated at 25°C for at least 1 h before recording. Analysis was performed with a MED64 at microelectrode array system (Alpha of Japan).

Western Blotting analysis
Total protein of striatum was extracted with RIPA buffer and protease inhibitors. BCA reagent was used for protein quanti cation. Proteins samples were transferred onto polyvinylidene di uoride (PVDF) membranes after gel electrophoresis. PVDF membranes were then blocked using 5% bovine serum albumin fraction V for 1h, then the membranes were incubated with primary antibody against FTO (1:1000), ephrin-B2 (1:1000), ephrin-B3 (1:500), and GAPDH (1:5000) overnight at 4°C. In the next day, membranes were incubated with corresponding secondary antibodies for 1 h at 25°C after washing 3 times with Tris-buffered saline with 0.1% Tween 20 (TBST). Analysis was performed with the multifunctional imaging system (AZURE C500, USA). GAPDH was used as an internal standard.

Quantifcation of m 6 A in total RNA
The ELISA-based m 6 A Quantitative Kit (Epigentek, P-9005-96) was used to quantify m 6 A levels of total RNA (200 ng). In accordance to the manufacturer's instructions, quanti cation was performed in triplicates.

Actinomycin D assay
Samples were treated with 2µM actinomycin D (Act-D, Sigma, USA) for 0, 3, and 6 h. Cells were collected and RNA was isolated for RT-qPCR. HPRT1 gene was used as an internal control as HPRT1 mRNA does not bind to YTHDF2, is rarely affected by Act-D, and does not undergo m 6 A modi cations.

Statistical analysis
Results were presented as means ± standard deviations after measured at least three times. The SPSS 22 software was used to analysis all statistical. The one-way analysis of variance (ANOVA) followed by Bonferroni test were used to determined the differences between the means for multiple comparison. Pvalue of < 0.05 or P-value of < 0.01 were considered statistically signi cant.

Effect of FTO on Mn-induced parkinsonism
To explore the role of FTO in Mn-induced striatal motor dysfunction, the level of Mn in normal striatum and Mn-exposed striatum was determined by ICP-MS. The ndings showed that levels of Mn in striatum increased signi cantly after Mn exposure compared with the control group (Fig. 1A). Further, mRNA and protein levels of FTO and other methylases in striatum and neurons were determined. Analysis showed that mRNA and protein levels of FTO were signi cantly downregulated after Mn exposure compared with the control group (Fig. 1B, C, F, and G). Notably, other methylases levels including METTL3/14, ALKBH5, and YTHDF1/2 were not signi cantly different after Mn exposure compared with the control group ( Supplementary Fig. 1A-T). Immuno uorescence was used to determine speci c FTO protein levels in striatum and neurons. Analysis showed that FTO level (Green) was signi cantly decreased after Mn exposure compared with the normal group (Fig. 1D, H). In addition, m 6 A ELISA was used to determine changes in total RNA m 6 A level in striatum and neurons after Mn exposure. The ndings showed that the level of total RNA m 6 A in striatum and neurons increased signi cantly after Mn exposure compared with the control group (Fig. 1E, I). These ndings indicate that FTO is downregulated in striatum and neurons after Mn exposure, resulting in signi cant increase in the level of total RNA m 6 A in striatum. This implies that FTO plays an inhibitory role in Mn-induced motor dysfunction.

Role of FTO in Mn-induced motor dysfunction and morphological analysis
Effect of over-expression FTO or inhibition of FTO activity in the striatum on regulating the symptoms was explored to verify that Mn-induced motor disorders were caused by down-regulation of FTO in the striatum. AAV5 were used to express FTO (AAV-FTO) or hrGFP (AAV-control) in the striatum through stereotactic injection (Supplementary Fig. 2A, B), resulting in selective re-expression in bilateral striatum ( Fig. 2A and B, Supplementary Fig. 2C). In addition, activity of FTO was inhibited by intraperitoneal administration of MA2, a selective inhibitor of FTO. The results showed that treatment of MA2 did not affect FTO expression (Supplementary Fig. 2D and E). Re-expression of FTO in mice exposed to Mn signi cantly decreased total RNA m 6 A levels, with no signi cant effect on levels of Mn in the striatum ( Fig. 2C and Supplementary Fig. 3A). On the contrary, inhibition of FTO activity in mice exposed to Mn signi cantly increased total RNA m 6 A levels, with no signi cant effect on levels of Mn in the striatum (Supplementary Fig. 2F and Supplementary Fig. 3K). Re-expression of FTO in striatum exposed to Mn signi cantly increased their endurance times and fatigue running distance ( Fig. 2D and E). Furthermore, it improved the distance covered and speed in the open eld ( Fig. 2F and Supplementary Fig. 3B-D). In addition, it signi cantly improved mice gait, mean intensity, run duration, average speed, swing speed and maximum variation performances in gait tests ( Fig. 2G and Supplementary Fig. 3E-J). On the contrary, motor function was damaged more severely after inhibition of FTO activity ( Supplementary Fig. 2G-J and Supplementary Fig. 3L-T). These ndings indicate that inhibition of FTO signi cantly reduces motor performances, implying that FTO plays a protective role in Mn-induced motor dysfunction.
Effects of Mn on striatum morphology were determined by HE staining. In the control group, the nucleus of striatum was stained clearly. The nucleus was round and giant with evident nucleoli (Fig. 2H). The 50 mg/kg MnCl 2 group (Fig. 2H) showed signi cant interstitial vacuolation and cell swelling compared with the control group. Re-expression of FTO signi cantly improved neuropathological injury compared with the 50 mg/kg MnCl 2 group (Fig. 2H). Nissl staining in the control group showed that Nissl bodies were clearly stained, arranged neatly and the number was normal (Fig. 2I). Analysis showed signi cant degeneration of Nissl bodies, deletion and lighter staining in the 50 mg/kg MnCl 2 group compared with the normal control group (Fig. 2I). Re-expression of FTO signi cantly alleviated deformation and deletion of Nissl bodies compared with the 50 mg/kg MnCl 2 group (Fig. 2I). Observation under an electron microscope showed that the nucleus of the control group was approximately round and the structure was normal (Fig. 2J). Neuronal nucleus pyknosis, organelle disappearance, swelling and degeneration were observed in the 50 mg/kg MnCl 2 group (Fig. 2J). Notably, re-expression of FTO signi cantly improved neuronal damage compared with the 50 mg/kg MnCl 2 group (Fig. 2J). Changes in dendritic density and/or morphology were investigated to explore the effects of Mn on axonal guidance regulates synaptic plasticity ( Spontaneous discharge activity re ects the function of striatal neurons. Acute brain slices from mice were used to record electrical signals through substrate embedded electrodes to explore the discharge activity of the striatum. A typical electrical signal obtained using a microelectrode is shown in Fig. 2M. Individual spontaneous potential was recorded when the individual spontaneous potential exceeds the threshold (± 0.02 mV). Neurons from the striatum of the control group showed normal spontaneous activity (Fig. 2M). The amplitude and frequency of spike waves decreased after treatment with 50 mg/kg MnCl 2 compared with the control group (Fig. 2M). Re-expression of FTO signi cantly improved the damaged spontaneous discharge frequency compared with the 50 mg/kg MnCl 2 group.
On the contrary, the morphology was severely damaged and electrophysiological signals of striatum were signi cantly reduced after inhibition of FTO activity compared with over-expression of FTO (Supplementary Fig. 2K-P). These ndings indicate that FTO plays a promoting role in Mn-induced striatal injury.

Effect of Mn exposure on nigro-striatal projection system
To explore the role of Mn treatment on nigro-striatal projection system, FG labeling (Fig. 3A), the immuno uorescence triple staining of TH, DARPP32 and DAT (Fig. 3B), the DA levels (Fig. 3C) and neuronal axons growth (Fig. 3D-E) were determined to evaluate the survival rate of nigro-striatal neurons.
In the control group, the boundary of neurons was clear and evident cell body were observed (Fig. 3A). In the 50 mg/kg MnCl 2 group, the boundary of neurons was unclear, the distribution of cells was uneven, the staining was lighter and disappeared signi cantly compared with the control group. (Fig. 3A). Reexpression of FTO signi cantly improved survival rate of nigro-striatal projection neurons compared with the 50 mg/kg MnCl 2 group (Fig. 3A).
TH, DARPP32 and DAT are speci c markers of nigro-striatal projection neurons, and their expression can re ect damage of nigro-striatal projection neurons. After treatment with 50 mg/kg MnCl 2 , the nigrostriatal projection neurons marked with TH, DARPP32 and DAT were signi cantly reduced, and the projection system was severely damaged compared with the control group (Fig. 3B). Re-expression of FTO signi cantly restored the damage to the projection system compared with the 50 mg/kg MnCl 2 group (Fig. 3B). After treatment with 50 mg/kg MnCl 2 , DA levels decreased signi cantly compared with control group (Fig. 3C). DA levels increased signi cantly after re-expression of FTO compared with the 50 mg/kg MnCl 2 group. Neuronal axons grew slowly and shrank in the 50 mg/kg MnCl 2 group compared with control group (Fig. 3D and E). Neuronal axon length increased signi cantly after re-expression of FTO compared with the 50 mg/kg MnCl 2 group. On the contrary, the substantia-striatal projection system was signi cantly damaged after inhibition of FTO activity compared with that after over-expression of FTO (Fig. 3F-J).

Role of FTO function in neuron and SH-SY5Y cells
To explore the role of FTO in neuron and SH-SY5Y cells, the levels of FTO and total RNA m 6 A and performed cytotoxicity were determined using neuron and SH-SY5Y cells (Fig. 4A-H). Over-expression of FTO ( Fig. 4A-B, and Supplementary Fig. 4A) in Mn-induced FTO-low neuron and SH-SY5Y cells decreased the levels of total RNA m 6 A, and cytotoxicity ( Fig. 4C-D). Inhibition of FTO activity in neurons and SH-SY5Y cells signi cantly increased the levels of total RNA m 6 A and cytotoxicity (Fig. 4E-H). These ndings indicate that FTO alleviates Mn-induced cytotoxicity in vitro.

Role of m 6 A in FTO functions in neuron and SH-SY5Y cells
To

Identi cation of key target genes for FTO function
Axon guiding molecules are expressed in the nigro-striatal projection system, whereas ephrins are mainly expressed in the striatum [17,27]. Inactivation of the FTO gene impairs DA receptors, controls activity and behavior of neurons in a dose-dependent manner. Therefore, analysis was performed to determine if ephrins are key targets of FTO after Mn exposure. Analysis showed showed that Mn exposure signi cantly reduces levels of ephrin-A4/5 and ephrin-B2 compared with the control group. Notably, levels of other ephrins of Mn exposed group showed no signi cant difference with those in the control group ( Fig. 6A-B, and Supplementary Fig. 5A-D). Analysis showed that over-expression of FTO signi cantly increased ephrin-A5 and ephrin-B2 levels ( Fig. 6C-H, and Supplementary Fig. 5E-F). In addition, overexpression of FTO signi cantly decreased mRNA m 6 A levels of ephrin-A5 and ephrin-B2, with no signi cant effect on expression of ephrin-A4 (Fig. 6I). On the other hand, downregulation of FTO activity signi cantly decreased ephrin-A5 and ephrin-B2 levels, and increased mRNA m 6 A levels of ephrin-A5 and ephrin-B2 (Supplementary Fig. 5E-G). Notably, over-expression of FTO showed no effect on expression of other methylases (Supplementary Fig. 5H-K) implying that Mn affects expression of ephrins only by regulating FTO. These ndings indicate that FTO inhibits expression of ephrin-A5 and ephrin-B2 after Mn exposure.
Role of YTHDF2 on function of FTO in neurons m 6 A regulates gene expression through YTH domain-containing family proteins [28]. In the present study, a gain function test was performed to explore mechanism of regulation of gene expression by Mn via determining m 6 A RNA methylation (Fig. 7A) of YTHDF1/2. Over-expression of YTHDF1 showed no effect on the mRNA levels of ephrin-A5 and ephrin-B2 in the control group and FTO over-expression cells ( Fig. 7B-E). However, over-expression of YTHDF2 signi cantly decreased levels of ephrin-B2 in both control and FTO over-expression cells, with no signi cant effect on expression of ephrin-A5 (Fig. 7F-I). In addition, over-expression of FTO signi cantly increased the stability of ephrin-B2 mRNA (Fig. 7J). Overexpression of YTHDF2 decreased the stability of ephrin-B2 mRNA (Fig. 7K). Immuno uorescence double staining of FTO and ephrin-B2 showed that Mn did not change the position of FTO and ephrin-B2. However, it decreased expression level of FTO and ephrin-B2. Notably, over-expression of FTO antagonized the effect of Mn (Fig. 7L). Grb4 and GIT1 are downstream target genes of ephrins, and play key roles in regulating axonal growth. The ndings of the present study showed that Mn treatment decreased mRNA expression of Grb4 and GIT1 (Fig. 7M). Over-expression of FTO and activation of ephrin-B2 signi cantly increased expression of Grb4 and GIT1 ( Fig. 7N-P). These ndings show that YTHDF2 modulates expression of FTO target genes by mediating RNA decay.

Role of FTO and ephrin-B2 in Mn-induced motor dysfunction
The relationship between ephin-B2 and FTO-mediated motor dysfunction was explored. Forced activation of ephrin-B2 in FTO gene suppressor mice signi cantly increased performance of mice in rotating rod, fatigue meter, open eld and gait tests (Fig. 8A-D and Supplementary Fig. 6A-I). Similarly, forced activation of ephrin-B2 in FTO suppressor mice signi cantly alleviated Mn-induced striatal morphological and functional damage (Fig. 8E-J). In addition, forced activation of ephrin-B2 in FTO suppressor neuronal cells signi cantly improved Mn-induced nigro-striatal projection system damage ( Fig. 8K-O). These ndings show that ephrin-B2 is a potential target of FTO in the striatum. In summary, the ndings of the current study indicate that FTO plays a critical role in inhibiting effect of ephrin-B2 in Mn induced nigrostriatal projection system dysfunction, implying that activation of FTO can promote expression of ephrin-B2 and then antagonize Mn-induced motor dysfunction. Disscussion PD is a neurodegenerative disease characterized by bradykinesia. In the top 10 most populated countries in the world, the number of people with PD is projected to be between 8.7 and 9.3 million by 2030 [29]. Several studies report that PD is associated with disruptions in metallic ion homeostasis-mainly Mn [30,31]. High levels of Mn results in parkinsonian-like symptoms known as manganism. This is mainly because neurodegeneration in PD occurs mainly in the DA neurons of the substantia nigra whereas Mn toxicity affects striatum. Both types of damage can decrease DA levels [30,31]. Ephrins are important axon guidance molecules which play important roles in regulating motor function [32]. However, mechanisms of Mn in regulating motor function though axon guidance molecular ephrins has not been explored clearly.
Posttranscriptional regulation is implicated in axon guidance, and several RNA-binding proteins are involved in these mechanisms [13,33]. The ndings of the current study show signi cant downregulation of FTO in Mn treated group in a dose-dependent manner. FTO inhibits striatal nerve and motor dysfunction caused by manganism in vivo. In addition, FTO inhibits toxicity of manganism cells in vitro.
Notably, at the molecular level, FTO over-expression decreases total RNA m 6 A level and upregulates key manganism inhibitory genes ephrin-A5 and ephrin-B2, and increases their mRNA stability. mRNA decay of ephrin-B2 gene in manganism cells is regulated by m 6 A reader YTHDF2. On the contrary, pretreatment with FTO selective inhibitor MA2 aggravates Mn-induced neurological dysfunction. FTO is the RNA demethylase of m 6 A [34], which has the opposite effect on the stability of mRNA. The ndings of this study showed that regulation of ephrin-B2 is partly regulated through the m 6 A/YTHDF2 mechanism due to the following reasons: (1) FTO over-expression increased mRNA stability and level of ephrin-B2 gene, and decreased the abundance of m 6 A; (2) YTHDF2 and METTL3/14 over-expression decreased mRNA stability and level of ephrin-B2 gene, and reversed the effect of FTO over-expression. These ndings imply that in Mn-induced parkinsonism, increase in m 6 A level leads to RNA degradation of ephrin-B2 gene through a YTHDF2-dependent mechanism. However, several other undiscovered additional effects of FTO-regulated demethylation should be explored.
Mn neurotoxicity is correlated with impaired DA, glutamatergic and GABAergic transmission and neuroin ammation. Mn toxicity causes extrapyramidal motor dysfunction due to preferential accumulation of Mn in the striatum [35]. Previous studies report that ephrin signaling is important for normal connectivity and function of midbrain DA neurons [36,37]. To verify that the observed defects resulted from the loss of FTO and ephrin-B2 mainly in the striatum, the role of ephrin-B2 activation in the striatum of mice in improving motor dysfunction was explored. The ndings showed that increase in expression level of ephrin-B2 signi cantly improved their motor function. On the contrary, pretreatment with MA2 damaged motor function. These ndings imply that deregulation of FTO and ephrin-B2 are signi cant mechanisms underlying Mn-induced neurotoxicity in mice. Future studies should be conducted to further explore anti-Mn effect of ephin-B2.
Striatum is associated with several neurodegenerative diseases such as PD [38] and Huntington's Disease [39]. Effect of Mn linked on motor function in the striatum should be explored further. The ndings of the current study show that Mn exposure causes pathological effects on the striatum characterized by cell loss, unusual shape and decrease in the number of neurons. These ndings are consistent with previous ndings by Song et al [40]. The ndings of the current study showed that increase in ephrin-B2 signi cantly alleviates neuropathological damage and ultrastructural changes caused by manganism. Notably, the neuropathological injury was severe in the MA2 pretreatment groups.
These ndings show the effects of FTO and ephin-B2 on ultrastructural changes of Mn-induced neurotoxicity. Studies report that axon guide molecules perform an important role in regulating synaptic formation. Notably, axon guide molecules determine the location of synapses and regulate neuronal function [41][42][43]. In the current study, increased expression of ephrin-B2 antagonized Mn-induced dendritic spine density dysfunction. These ndings show for the rst time that ephin-B2 plays an important role in Mn-induced neuronal damage.
Brain DA neurons play important role in motor function control and related disorders by transmitting electrophysiological signals [44][45][46][47]. In addition, spontaneous ring activity is responsible for placing baseline DA ranges in the striatum [48], which regulate cognitive and motor function [49,50]. Therefore, determinants of the spontaneous ring activity in vivo play important roles in regulation of DA signaling and motor function. Notably, Mn-induced disappearance of spontaneous discharge was improved by ephrin-B2 activation. Spontaneous discharge frequency was damaged in MA2 pretreatment groups.
These ndings indicate a novel mechanism for Mn regulation of spontaneous ring rate of DA neurons in manganism.
Mn induced Parkinsonism is mainly caused by decrease in DA content in the striatum [12]. DA is derived from release of DA neurons from the substantia nigra after projecting to the striatum. This pathway is known as the nigro-striatal projection pathway of DA neurons [51]. To explore the effect of Mn exposure on snigro-striatal projection system, FG labeling, immuno uorescence triple staining of TH, DARPP32 and DAT, DA levels and neuronal axons growth were determined to evaluate the survival rate of nigro-striatal neurons. The ndings showed that Mn treatment signi cantly damaged the nigro-striatal projection system, whereas re-expression of FTO and ephrin-B2 signi cantly increased survival rate of the projection neurons. Notably, substantia-striatal projection system was severely damaged after inhibition of FTO activity. These ndings show that down-regulation of FTO and ephrin-B2 signi cantly affect morphology and function of the nigro-striatal projection system, implying that FTO mediated axon guidance promotes Mn-induced dysfunction of the nigro-striatal projection system, resulting in the failure of DA to be transmitted to the striatum, which may eventually cause striatal motor dysfunction.

Conclusions
In summary, the ndings of the current study show that FTO over-expression inhibits Mn-induced parkinsonism and promotes expression of ephrin-B2 gene, partially through inhibits ephrin-B2 mRNA degradation, which is mediated by YTHDF2. In addition, promoting the expression of ephrin-B2 antagonizes Mn-induced parkinsonism in mice. These ndings provide a basis to explore the speci c mechanisms of Mn-induced motor dysfunction and provides a novel approach for treatment of Mn-