1 Introduction
Amyotrophic lateral sclerosis (ALS), also known as “Lou Gehrig’s disease,” is a progressive and destructive neurodegenerative disease that affects upper and lower motor neurons (MNs); it results in paralysis and eventually death generally due to respiratory failure [
1], thereby placing great burdens on families and the society. Riluzole and edaravone are the only drugs approved by the FDA to treat ALS, but they have limited therapeutic effects. Riluzole has a modest benefit on the survival of patients with ALS, whereas edaravone mildly improves the ALS scale score in a small number of patients with ALS [
2,
3]. Although the etiology and pathogenesis of ALS remain unclear, pathogenesis may be associated with gene mutation, cell apoptosis, oxidative stress injury, abnormal immune function, abnormal folding and aggregation of proteins, axonal transport disorders, mitochondrial function damage, excitatory amino acid toxicity, and other factors [
4–
6]. The diagnosis of ALS often relies on clinical evaluation, electromyography, and exclusion of diseases mimicking ALS. Indeed, an average diagnostic delay of 9–15 months from the first symptoms to diagnostic confirmation usually occurs [
7]. Thus, the diagnostic delay should be shortened, and the efficiency of the diagnosis must be improved because of the limited survival of patients with ALS. Therefore, the discovery of specific biomarkers for early diagnosis and monitoring the progression of ALS must be needed.
In this study, exosomes were extracted from the plasma of patients with ALS and healthy controls. The differentially expressed proteins in exosomes were evaluated by proteomic analysis and compared between the two categories of subjects. After functional classification, functional enrichment, cluster analysis, and a literature review, this work focused on coronin-1a (CORO1A) because it is the most significantly different protein. However, the role of CORO1A in ALS is still unknown because no previous studies exploring the relationship between CORO1A and ALS pathogenesis are available.
Coronin protein exists in all eukaryotic species. The different subtypes of the coronin family have a similar structure because they have a WD40 repeating domain containing a β-propeller, which binds F-actin. The members of the coronin family regulate actin dynamics, cell migration, vesicle transport, phagocytosis, and cell division [
8]. The coronin family includes three types of proteins: coronin I, coronin II, and coronin III. Coronin I is one of the best characterized mammalian coronins. Previous studies indicated that coronin I plays a crucial role in protecting pathogenic mycobacteria from lysosomal delivery [
9]. CORO1A is a member of the coronin I family abundantly expressed in neurons, T cells, B cells, macrophages, mast cells, and neutrophils [
10]. Thus, the expression of CORO1A in the plasma of patients with ALS and the spinal cords of ALS mouse models was evaluated to assess whether it increased. Moreover, our previous research showed an overactivation of autophagy in a mouse model of ALS [
11], suggesting a potential relationship between CORO1A and autophagic flux in ALS.
Therefore, in this study, the abnormal expression of CORO1A in the plasma of patients with ALS and the spinal cords of ALS mice was investigated, and the potential mechanisms of CORO1A in neurodegeneration were further explored. Our study is the first to demonstrate the potential role of CORO1A in the pathogenesis of ALS, revealing its value as a plasma biomarker in the diagnosis of this disease.
2 Materials and methods
2.1 Participants and sampling
This study was approved by the Ethics Committee of Ruijin Hospital, Shanghai Jiao Tong University School of Medicine, Shanghai, China. ALS was clinically diagnosed in patients according to the EI Escorial-revised criteria and Chinese guidelines for the diagnosis and treatment of ALS [
12,
13]. The exclusion criteria were the following: (1) familial ALS; (2) a concomitant diagnosis of any type of dementia; (3) history of major psychiatric disorders; and (4) history of major cardiac, renal, liver, or other systemic diseases. In addition, 33 healthy controls without acute and chronic diseases were selected. All participants signed the informed consent form before the collection of the peripheral blood samples. Blood samples were obtained from all participants by venipuncture and immediately centrifuged to isolate the plasma, which was stored at −80 °C until it was further used.
2.2 Proteomic analysis of exosomal proteins
The plasma of three patients with ALS and three healthy controls was used. The blood of the three patients with ALS was collected 6 months after the onset of the first symptoms. The exosomes were extracted using an exosome isolation kit (UR52131, Umibio, China) in accordance with the manufacturer’s instructions. The samples were centrifuged at 3000× g and 4 °C for 10 min and then at 10 000× g and 4 °C for 20 min to remove the cells and debris. The working reagents were added proportionally to the volume of the starting sample. Mixtures were vortexed and incubated at 4 °C for 2 h. Then, they were centrifuged at 10 000× g and 4 °C for 60 min to precipitate the exosome pellets. The pellets were resuspended in PBS and purified using an exosome purification filter and centrifuged at 3000× g and 4 °C for 10 min.
The exosomes were identified by electron microscopy (Fig. S1) and nanoparticle tracking analysis. The exosomal proteins were processed by trypsin digestion and tandem mass tag labeling to acquire tryptic peptides. The tryptic peptides were fractionated by high-pH reverse-phase high-performance liquid chromatography. Then, the peptides were subjected to a nanospray ionization source followed by tandem mass spectrometry (MS/MS) in a Q ExactiveTM Plus (Thermo) system coupled online to an ultra-performance liquid chromatography system.
The acquired MS/MS data were processed using the MaxQuant search engine (v.1.5.2.8). Tandem mass spectra against the human UniProt database concatenated with the reverse decoy database were used for peptide and protein identification. Trypsin/P was specified as a cleavage enzyme allowing up to four missing cleavage events. The mass tolerance for the precursor ions was set at 20 ppm in the first search and 5 ppm in the main search, and the mass tolerance for fragment ions was set as 0.02 Da. Carbamidomethyl on cysteine was specified as a fixed modification, and an acetylation modification and oxidation on methionine were specified as variable modifications. The false discovery rate was adjusted to < 1%, and the minimum score for the modified peptides was set as > 40.
2.3 Animals
The SOD1-G93A transgenic mice expressing the G93A mutant form of the human SOD1 gene (a substitution of glycine to alanine in position 93) are commonly used as ALS animal models. These animals can develop paralysis in one or more limbs because of the loss of motor neurons from the spinal cords, so they are similar to the clinical phenotypes and histopathological features of patients with ALS. The SOD1-G93A mice were purchased from the Jackson Laboratory (stock No. 002726). The colony was maintained by mating the transgenic mice with wild-type mice (NC) from the same background. All mice were kept in a controlled temperature and light/dark cycle (12 h/12 h). All experiments and animal care procedures were conducted in accordance with the Laboratory Animals Care Guidelines approved by the Animal Committee of Shanghai Jiao Tong University School of Medicine. Mice were anesthetized with chloral hydrate and transcardiac perfusion by 100 mmol/L cold PBS was performed at the age of 4 weeks (4W), 8 weeks (8W), 12 weeks (12W), and at the time of death (DIE). The fresh spinal cords were collected. When the mouse lost the ability to straighten within 30 s of being placed on its back, it was considered dead.
2.4 Cell culture and transfection
The mouse motor neuron NSC-34 cell line (Cell Bank of Chinese Academy of Sciences, China) was used in this study. These cells were cultured in high-glucose Dulbecco’s modified Eagle’s medium (DMEM, 11965, Gibco, USA) supplemented with 100 U/mL penicillin, 10% fetal bovine serum, 1 mmol/L L-glutamine, and 100 μg/mL streptomycin. They were maintained in a humidified atmosphere with 5% CO2 at 37 °C. CORO1A overexpression plasmids were constructed as follows: the wild-type CORO1A-CDS was amplified through polymerase chain reaction (PCR) from the NSC-34 cDNA by using the following primers: forward 5′-CGGCTAGCATGAGCCGGCAGGTGGTCCG-3′ and reverse 5′-CGGGATCCCTACTTGGCCTGGACTGTCT-3′. The PCR product was purified and cloned into a pCDH vector (CD513B-1, SBI, USA) by using NheI and BamHI enzyme sites. CORO1A overexpressing vectors and short hairpin RNA (shRNA) vectors (TRCN0000029340) were transfected into the NSC-34 cells. The cells were kept under 2% FBS in the medium for 6 h before transfection by using Lipofectamine 3000 (L3000015, Invitrogen, USA). The cells were kept in an FBS-free medium during transfection; the culture medium was replaced with a fresh whole medium containing 10% FBS and high-glucose DMEM after the transfection. The binding and functional efficiency of the overexpressing or shRNA vector was identified through Western blot (Fig. S2). The cells were divided into three groups: (1) cells transfected with the CORO1A-overexpressing vector (OE group); (2) cells transfected with CORO1A shRNA vector (SH group); and (3) cells without any transfection but treated with the transfection reagent (NC group). The following experiments were performed 24 h after transfection. Cyclosporin A (CsA, S2286, Selleck, China), a calcineurin (CaN) inhibitor, was added to the culture medium of the OE group at 100 nmol/L. The cells were incubated with CsA for 24 h and named the CsA group.
2.5 Western blot
Exosomal proteins were prepared as described above. The total proteins from the plasma, mouse spinal cords, or NSC-34 cells were homogenized and extracted by RIPA lysis buffer. Protein concentration was evaluated using the bicinchoninic acid method. The proteins were loaded and separated on an SDSPAGE gel and transferred to a 0.45 or 0.22 μm polyvinylidene fluoride membrane, which was treated with 5% nonfat milk blocking solution for 1 h and subsequently incubated in a solution with the respective primary antibodies: anti-CORO1A (92904S, Cell Signaling Technology, USA), anti-light chain 3 II (anti-LC3II, L7543, Sigma-Aldrich, Germany), anti-sequestosome-1 (anti-p62, 5114S, Cell Signaling Technology, USA), anti-Beclin-1 (ab207612, Abcam, UK), and anti-CaN (ab3673, Abcam, UK). The membrane was incubated with the respective peroxidase-conjugated secondary antibodies. After the incubation with the chemiluminescent horseradish peroxidase substrate, the membrane was exposed to film. Proteins were quantified using image analysis software.
2.6 Enzyme-linked immunoassay
Sandwich enzyme-linked immunoassay (ELISA) was used to quantify CORO1A according to the manufacturer’s protocol (05657, Y-S Biotechnology, China). Briefly, the reagents, samples, and standards were prepared as indicated in the protocol. Then, 50 μL of the sample or standard was added to the appropriate wells, and 50 μL of the antibody cocktail was added to all wells. The plate was incubated at room temperature for 1 h. The medium was removed from each well, and the wells were washed three times with wash buffer. Then, 100 μL of tetramethylbenzidine development solution was added to each well and incubated for 10 min. Finally, 100 μL of stop solution was added, and OD was read at 450 nm.
2.7 TdT-mediated dUTP nick-end labeling assay
TdT-mediated dUTP nick-end labeling (TUNEL) staining was performed to detect apoptosis by using a one-step TUNEL apoptosis assay kit (C1089, Beyotime, China) in accordance with the manufacturer’s instructions. The cell smears fixed with 4% paraformaldehyde were washed with PBS solution and then incubated with PBS containing 0.3% Triton X-100 at room temperature for 5 min. The cells were washed with PBS and incubated in a freshly prepared working solution at 37 °C for 60 min. Then, the nuclei were stained with DAPI (62247, Life Technologies, USA), a coverslip was placed, and the cells were observed under a fluorescence microscope. The apoptotic cells were identified by TUNEL-positive staining. The percentage of apoptotic cells was calculated as the percentage of TUNEL-positive cells out of the total number of DAPI-positive nuclei.
2.8 Reactive oxygen species activity
The intracellular reactive oxygen species (ROS) level was detected using the ROS assay kit (S0033S, Beyotime, China). The cells were suspended in the diluted dichloro-dihydro-fluorescein diacetate (DCFH-DA) and incubated at 37 °C for 20 min. Then, they were washed with a serum-free medium three times to remove excessive DCFH-DA. Fluorescence was measured at the excitation wavelength of 488 nm and emission wavelength of 525 nm through flow cytometry (Thermo Fisher, USA).
2.9 Immunofluorescence staining
The cells were prepared, fixed with paraformaldehyde, and processed with Triton X-100 solution to expose the antigenic sites. Then, the proteins were blocked with 1% bovine serum albumin solution. The slides were treated with the following primary antibodies and incubated overnight at 4 °C: anti-LC3II (ab243506, Abcam, UK) and anti-lysosomal associated membrane protein 1 (anti-LAMP1, ab25630, Abcam, UK). Then, the primary antibodies were removed, and the slides were incubated with Alexa Fluor dye-conjugated IgG secondary antibody for 2 h. After being sealed with an antifade reagent, the slides were visualized under a fluorescence microscope.
2.10 Statistical analysis
Data were statistically analyzed using SPSS 19.0 (IBM, USA) and subjected to one-way ANOVA. LSD test or Dunnett-t-test was performed when the variance between groups was equal (P > 0.05) or not equal. Data with P < 0.05 were considered statistically significant.
3 Results
3.1 Increased expression of exosomal CORO1A was detected by mass spectrometry
Three male patients with sporadic ALS (sALS) and three male healthy controls were enrolled in this study. The average age of the three patients with sALS was 42 ± 3.06 years, whereas the average age of the controls was 39.33 ± 3.79 years.
Interestingly, the mean diameter of the exosomes from the ALS group was significantly smaller than that from the control group (103.2 ± 67.9 nm vs. 135.6 ± 53.3 nm, n = 30, P = 0.035, Fig. S1). A total of 200 792 spectra by mass spectrometry were obtained from the extracted exosomes, and a spectral matching was observed in 16 283 of them. A total of 5024 peptides were detected, with 4724 unique peptides. Then, 1107 proteins were identified in the exosomes, and 929 were quantitively analyzed. The threshold of the expression change was set to 1.2-fold, and P was < 0.05. The expression of 106 proteins in the patients with sALS was significantly higher than that in the healthy controls, whereas the expression of 86 proteins was downregulated (Fig. S3).
3.2 Increased CORO1A expression in ALS
Among the differentially expressed proteins (Tab.1), the expression of exosomal CORO1A was the one with the most significant difference between the two groups of patients because it was 5.3-fold higher in the patients with ALS than in the controls. Then, the proteins from the plasma of patients with ALS and the spinal cords of SOD1-G93A mice were extracted. Western blot analysis showed an increased CORO1A expression in the plasma of patients with ALS (Fig. 1A and 1C), and ELISA detection also indicated a significantly higher expression of CORO1A in the plasma of 30 patients with ALS (41.50 ± 8.39 years, 18 males and 12 females) than that in 30 controls (39.57 ± 10.08 years, 16 males and 14 females, P < 0.0001; Fig.1). Moreover, ELISA detection and western blot analysis indicated a significant increase in the expression of this protein in the spinal cord of the SOD1-G93A mice compared with that of the WT mice (Fig.1, Fig.1, andFig.1). The CORO1A expression in the spinal cords was also measured at the age of 4W, 8W, 12W, and at the time of death, revealing that the CORO1A expression increased with age (P = 0.53 (4W vs. NC), P = 0.042 (8W vs. NC), P = 0.0062 (12W vs. NC), P = 0.0019 (DIE vs. NC); Fig. 1B and 1D).
Furthermore, the CORO1A expression in the plasma of 10 patients with ALS (5 males and 5 females) was measured at three time points (baseline, 6 months, and 12 months after the onset of the first symptoms). According to the classic subtype of ALS, the first sign of ALS in 10 patients started in the left/right hand and subsequently spread to other parts of the body. The baseline time was defined as 3 months after the onset of the first symptoms. The average age of 10 patients was 39.60 ± 6.42 years. The percentages of the predicted forced vital capacity (FVC%) in all of them were greater than 70% at the last time point. The CORO1A expression in the plasma of 8 patients (5 males and 3 females) increased with the disease progression (Fig.1). The average expression of CORO1A after 6 months and after 12 months was significantly higher than that at baseline (baseline, 2.74 ± 0.58 ng/mL; after 6 months, 3.54 ± 0.77 ng/mL; after 12 months, 4.23 ± 0.81 ng/mL; P < 0.05), and the average level of CORO1A after 12 months was higher than that after 6 months ( P < 0.05).
3.3 CORO1A overexpression led to increased apoptosis and oxidative stress damage in NSC-34 cells
The CORO1A expression was upregulated in the NSC-34 cells by using an overexpression vector, but it was downregulated by shRNA. Western blot analysis revealed that the CORO1A expression in the OE group increased, but it was inhibited in the SH group (Fig. S2), suggesting that the vectors of the CORO1A overexpression and shRNA were constructed correctly.
The changes in the apoptotic level by TUNEL staining demonstrated more TUNEL immunofluorescence in the OE group than that in the SH and NC groups (P = 0.031). Conversely, no significant difference in the apoptotic level was observed between the SH and NC groups (Fig.2). Oxidative stress was measured by the ROS assay kit, and the results revealed that ROS levels in the OE group were higher than those in the NC group. However, the ROS level in the SH group was lower than that in the NC group (Fig.3).
3.4 CORO1A overexpression led to the overactivation of autophagy in NSC-34 cells
The expression of autophagy-related proteins, such as LC3II, p62, and Beclin-1, was also evaluated (Fig.4−Fig.4), and the results showed that it was significantly higher in the OE group than in the NC and SH groups. The expression of p62 and Beclin-1 was lower in the SH group than in the NC group, but the expression of LC3II did not significantly change. The expression of LC3II was further detected by immunofluorescence staining. Quantitative analysis indicated less LC3II positively stained puncta in the SH and NC groups than in the OE group (Fig.4and Fig.4).
The number of autophagosomes and autolysosomes was evaluated by immunofluorescence staining (Fig.5). The LC3II-positive staining was observed only in the autophagic vacuoles (AVs), LC3II- and LAMP1-positive staining was observed in the autolysosomes (ALs). The ALs/AVs ratio was lower in the OE group than in the NC group. However, ALs/AVs ratio was higher in the SH group than in the NC group (Fig.5 and Fig.5).
Our hypothesis was that the effect of CORO1A on autophagy might be exerted by the calcium ion/calcineurin (Ca2+/CaN) pathway. Thus, the expression of CaN was evaluated by Western blot (Fig.4 and Fig.4), and the results revealed that it was significantly higher in the OE group than in the SH and NC groups. Next, the OE group was treated with the inhibitor of CaN (CsA). Under this condition, the expression levels of LC3II, p62, and Beclin-1 in the CsA group were lower than those in the OE group (Fig.6), and the immunofluorescent intensity of LC3II staining also decreased (Fig.5and Fig.5). The TUNEL-positive ratio in the CsA group decreased compared with that in the OE group (Fig.2), whereas the ALs/AVs ratio increased (Fig.5 and Fig.5).
4 Discussion
In recent years, exosomes have been widely explored for their important role in intercellular communication. Proteins are one of the essential components in exosomes. Thus, differentially expressed proteins in the exosomes of the plasma of patients with ALS were assessed. A total of 192 differentially expressed proteins were screened out; among them, CORO1A was the one mostly increased and selected for further research. Moreover, a high expression of CORO1A was found in the plasma of patients with ALS and SOD1-G93A mice. Hence, CORO1A might play a vital role in ALS pathogenesis.
Since the early symptoms of ALS are not typical, the early diagnosis is still difficult, taking months from disease onset to diagnosis. Currently, no definite biomarkers for ALS are available [
14]. Our study indicated that the CORO1A expression in the spinal cord of SOD1-G93A mice increased with disease progression. Its expression in the plasma of patients with ALS was significantly higher than that of the controls, and it increased with the disease progression in 8 of the 10 selected patients. The plasma of the two other patients did not show an evident increase in the CORO1A expression. However, the two patients had a slower disease progression and milder clinical symptoms than the eight other patients during the 12-month follow-up. Therefore, CORO1A may be a potential biomarker for diagnosing ALS and monitoring disease progression. Although many therapies with promising results in different ALS animal models are available, few therapies improved clinical outcomes in patients with ALS [
15] partly because of the lack of useful blood biomarkers, which can help in the screening of individuals at risk of developing ALS and the evaluation of the therapeutic effect. CORO1A may have a potential role as a biomarker of ALS, allowing the design of new therapies.
Coronin I is important for normal immune cell homeostasis and neuronal function [
10]. CORO1A, one member of coronin I, is highly expressed in the nervous system. Genetic studies on ALS have suggested a causal relationship between an abnormal cytoskeleton and the degeneration of motor neurons [
16]. The dysregulation of some cytoskeletal proteins is involved in ALS [
17]. The coronin family is an actin-binding protein family, which regulates actin dynamics [
8]. Therefore, our hypothesis was that the upregulated expression of CORO1A in ALS might be related to the dysregulation of the actin cytoskeleton.
In this study, the overexpression of CORO1A led to the increased expression levels of LC3II, p62, and Beclin-1, suggesting the blockage of autophagic flux in the OE group. Besides, CORO1A shRNA inhibited autophagy by downregulating the expression of LC3II, p62, and Beclin-1 compared with their level in the NC group. Hence, CORO1A played a role in autophagy regulation.
Autophagy begins with the formation of autophagosomes, which can fuse with late endosomes and lysosomes to form autolysosomes. Our previous study revealed autophagic overactivation in spinal motor neurons in a SOD1-G93A mouse model. The expression of autophagic proteins in the spinal motor neurons of ALS mice significantly increased, and the number of autophagosomes significantly increased [
4,
18–
20]. However, no significant change was observed in the number of ALs, resulting in a decrease in the ALs/AVs ratio and abnormal protein aggregation. This result demonstrated that the formation of ALs might be inhibited in ALS. Is there a relationship between CORO1A and the formation of ALs?
Once inside the neurons, CORO1A can bind to endosomes to inhibit the fusion of an endosome with a lysosome; as a result, the degradation of endosomes reduces [
21,
22]. Moreover, previous studies showed that the infection of
Mycobacterium tuberculosis can upregulate the expression of CORO1A, which can inhibit the fusion of phagocytes and lysosomes and protect from the degradation of phagolysosomes [
23,
24]. The increased expression of CORO1A can lead to the inhibition of the fusion of phagosomes/endosomes and lysosomes. Therefore, our hypothesis was that CORO1A might block the fusion of AVs with lysosomes. In our study, the formation of AVs was evaluated by the immunofluorescent staining of LC3II, and ALs were detected by the co-staining of LC3II and LAMP1. The overexpression of CORO1A resulted in more LC3II positive puncta, further indicating the overactivation of autophagy. The ratio of ALs/AVs in the OE group was lower than that in the NC group; conversely, the SH group had a higher ratio than the NC group. This result suggested that the formation of ALs might be inhibited by the overexpression of CORO1A, which led to the blockage of autophagic flux and abnormal protein aggregation. Therefore, the overexpression of CORO1A might be involved in ALS pathogenesis by inhibiting the formation of ALs.
CORO1A in
M. tuberculosis can promote the activation of CaN, leading to the inhibition of the fusion of phagosomes and lysosomes [
25]. CaN plays an important role in the regulation of phagosome–lysosome fusion [
26]. Thus, our hypothesis was that CaN might also have a function in the fusion of autophagosomes and lysosomes, which further affected the autophagic flux. In this study, the expression of CaN increased in the OE group but decreased in the SH group. When the OE group was treated with a CaN inhibitor (CsA), the overexpression of LC3II, p62, and Beclin-1 was reversed. CsA also promoted the synthesis of ALs. The expression of CaN might be positively correlated with the expression of CORO1A and the activation of autophagy. Therefore, CORO1A might block the autophagic flux through the activation of CaN. According to previous studies, CORO1A promotes the release of Ca
2+ and the activation of CaN [
25,
26]. CaN dephosphorylates the transcription factor EB (TFEB) by binding to it. TFEB translocates into the nucleus to activate the transcription of autophagy-related genes, leading to the upregulation of autophagy [
27]. In addition, the increased cytosolic Ca
2+ concentration inhibits the phosphorylation of the mammalian target of rapamycin (mTOR) through the activation of the Ca
2+/calmodulin-dependent protein kinase-kinase beta/adenosine 5ʹ-monophosphate-activated protein kinase (CaMKKβ/AMPK) pathway, potentially resulting in the upregulation of autophagy [
28].
The overexpression of CORO1A triggered apoptosis and oxidative stress, while CORO1A shRNA decreased the apoptosis and oxidative stress level compared with that in the NC group. This evidence further suggested that CORO1A might play an important role in the degeneration of motor neurons in ALS.
Despite the encouraging results, our study had some limitations. ALS symptoms vary greatly from person to person. Our clinical sample size was not large enough, and the rate of disease progression was not the same. The present work provided an idea of the significance of CORO1A. Further studies will be performed using more patients with ALS for a more detailed analysis. Moreover, the detailed mechanism is unclear, additional efforts will be made to provide more evidence on this aspect.
In conclusion, the role of CORO1A in ALS was evaluated for the first time. An increased expression of CORO1A was observed in ALS. The upregulated expression of CORO1A might inhibit the autophagic flux by the activation of the Ca2+/CaN signaling pathway. Therefore, CORO1A might have a potential role as a biomarker and therapeutic target for ALS (Fig.7). However, further confirmation in ALS animal models and more patients should be provided, and further studies will be performed by our research group in the future.
PDF
(5013KB)
