Transgenic human SOD1-G93A mice and their non-transgenic (NTG) littermates were produced by breeding female B6SJLF1 hybrids with male hemizygous carriers [B6SJL-Tg (SOD1-G93A) 1Gur/J], which were purchased from the Jackson Laboratory (Bar Harbor, ME, USA). The SOD1-G93A mouse genotyping was performed as described previously (Li et al., 2017). All animals were group-housed (4–5 mice/cage) under identical conditions (12 h light/dark cycle with free access to food and water). Since the clinical phenotype of SOD1-G93A mice, characterized by an adult onset of motor symptoms in the hind limbs, was first manifested around 12 weeks of age, and pathological changes in spinal cord begin to be observed around 60 days which progresses continuously to end stage at 17–20 weeks (Weydt et al., 2003; Ludolph et al., 2007), SOD1-G93A mice at three predefined stages were used in this study: (1) Pre-symptomatic: 40 days, postnatal, no signs of motor deficit; (2) Onset: 80 days, postnatal, initial signs of motor deficit; (3) End: animals can no longer right themselves 30 s after being placed on their backs or sides. The experiments were conducted in accordance with the regulations of laboratory animal management promulgated by the Ministry of Science and Technology of the People’s Republic of China, and approved by the Ethics Committee of the Second Hospital of Hebei Medical University. All mice were narcotized under 1.0–3.0% (vol/vol) isoflurane (RWD, Shenzhen, China), and we made all efforts to minimize suffering.

The mouse neuroblastoma/spinal cord hybrid cell line (NSC34) is a motoneuron-like cell line through the fusion of embryonic mouse spinal cord motor neurons and mouse neuroblastoma cells. NSC34 cells stably transfected with GFP-empty vector (E), GFP-human SOD1 wild type (hSOD1WT), or GFP-human SOD1G93A (hSOD1G93A) were established as previously described in our laboratory (Bai et al., 2021). The cell lines were incubated in Dulbecco’s modified Eagle’s medium (DMEM), including 10% heat-inactivated fetal bovine serum (FBS) and 1% penicillin streptomycin. To maintain the stable cell lines, 0.2 mg/mL Geneticin 418 sulfate (G418) was used. Cells were cultured at 37^°C under a 5% CO₂ humidified atmosphere, refreshing the medium every 1–2 days. When 80–90% confluence was reached, cells were passaged using a 0.25% trypsin/EDTA solution for subsequent experiments. The cells were starved in serum-free DMEM for 24 h, and then incubated in medium containing ET-1 (1 nM, 10 nM, 100 nM, 1,000 nM) for 24 h or with a fixed 100 nM ET-1 concentration for different lengths of time (12 h, 24 h, 48 h). For rescue experiments, 100 nM ET-1 in combination with BQ123 or BQ788 was added to the cells and then incubated for 24 h.

The measurement of cell viability was performed by a cell counting kit-8 (CCK-8) assay (BOSTER Biological Technology Co., Ltd., Wuhan, China) (Liu et al., 2014). In short, after a 24 h exposure to the respect treatment, we added 10 μL of CCK-8 solution to each well and incubated the cells at 37^°C for 1 h in the dark. Then, a microplate reader was used to measure the absorbance at a wavelength of 450 nm (Tecan Spark 10 μM, Switzerland). All experiments were performed at least three times independently.

The SOD1-G93A mice and NTG mice at different disease stages were perfused with 0.1 M ice-cold phosphate buffered saline (PBS), followed by perfusion with 4% paraformaldehyde (PFA). The spinal cords were collected and lumbar segments (L3-5) were post-fixed in 4% PFA/PBS at 4^°C for 24 h. The following day, the tissues were cryoprotected in 30% sucrose at 4^°C overnight and then embedded in optimum cutting temperature (OCT) compound. The lumbar segments (L3-5) were sliced into coronal sections (25 μm thick) with a Leica cryostat (CM1850). Sections were treated with 3% hydrogen peroxide (H₂O₂) for 10 min and then washed in PBS with gentle agitation for 10 min. Afterward, the sections were permeabilized/blocked with 5% goat serum and 0.3% Triton X-100 in PBS and then co-immunostained for primary antibodies diluted in blocking solution at 4^°C overnight. The primary antibodies include: rabbit anti-Endothelin-1(Abbiotec, 250633, 1:100), rabbit anti-ETA (Bioss, bs-1757R, 1:300) and rabbit anti-ETB (Abcam, ab117529, 1:500). After incubation, the tissues were washed in PBST (0.2% Tween 20 in PBS) three times, and then incubated with a corresponding biotin-conjugated secondary antibody (1:1,000). Next, after washing three times in PBS, incubating in Vectastain ABC Reagent (Vector Laboratories, Burlingame, CA, USA, PK-6100) for 40 min, color development was performed to the sections by the immPACT DAB Peroxidase Substrate Kit (Vector, SK-4105). At last, the slices were loaded onto slides and dried appropriately. The slides were soaked in anhydrous ethanol for 5 min, xylene for 10 min and finally sealed with Permount TM Mounting Medium (ZSGB-BIO, ZLI-9516). All images were photographed under an Olympus BX51 microscope equipped with a DP72 digital camera system (Olympus, Tokyo, Japan).

For immunofluorescence, spinal cord sections (25 μm thick) were pretreated with 1% Triton X-100 in PBS for 30 min, and blocked with solution containing 5% donkey serum, 0.3% Triton X-100 and 0.2% dry milk in PBS for 30 min. The sections were then incubated with primary antibodies in blocking buffer overnight at 4^°C. Primary antibodies include goat anti Iba-1 (Wako, 019-19741, 1:250), mouse anti-GFAP (Millipore, MAB360, 1:400), mouse anti-NeuN (Millipore, MAB377, 1:100), mouse anti-APC (Millipore, OP80, 1:400), rabbit anti-SOD1 (Immunoway, YT4364, 1:100), rabbit anti-cFOS (Abcam, ab222699, 1:100), rabbit anti-MAP1B (Proteintech, 21633-1-AP, 1:100), mouse anti-SMI32 (Biolegend, 801701, 1:100), rabbit anti-GFP (Life tech, G10362, 1:100), rabbit anti-Endothelin-1 (Abbiotec, 250633, 1:100), rabbit anti-ET-A (Bioss, bs-1757R, 1:300) and rabbit anti-ET-B (Abcam, ab117529, 1:500). Thereafter, the sections were washed for 30 min followed by incubation with the appropriate secondary antibody (Alexa Fluor 488, 594 or 647, Invitrogen) at room temperature for 2 h. After further rinsing in PBS for 30 min, sections were stained with mounting medium containing DAPI (VECTOR, VECTASHIELD H-1200) and sealed with nail polish.

Quantitative assessment of motor neuron labeling was performed as follows: Three representative images of the lumbar ventral horns from each tested animal were used for quantitative measurements of motor neurons. Motor neurons were defined according to the following criteria: (1) neurons were located in the anterior horn of the spinal cord and were NeuN⁺ and (2) neurons with a diameter of 20 μm or larger had a distinct nucleolus (Manabe et al., 2003). The images from immunofluorescence staining were captured using an Olympus FV1000 confocal microscopy or Zeiss Vert. A1, German confocal microscopy. The numbers of motor neurons in the ventral horn area and quantification of fluorescence intensity were analyzed using ImageJ software (Java 8, National Institutes of Health, USA) (Jensen, 2013).

Total protein from frozen tissues was extracted using a protein extraction kit and phenylmethylsulfonyl fluoride protease inhibitors (Beijing Solarbio Science and Technology Co., Ltd., China). 30 μg of protein lysates from each sample were electrophoresed through a 10% or 12% SDS–PAGE and then transferred onto polyvinylidene fluoride (PVDF) membranes. The membranes were blocked with 5% non-fat milk, and, respectively incubated with primary antibodies overnight at 4^°C, including rabbit anti-SOD1 (Immunoway, YT4364, 1:1,000), rabbit anti-GFP (Life tech, G10362, 1:1,000), rabbit anti-Endothelin-1 (Abbiotec, 250633, 1:300), rabbit anti-ET-A (Bioss, bs-1757R, 1:1,000), rabbit anti-ET-B (Abcam, ab117529, 1:4,000), and rabbit anti-GAPDH (Proteintech, 10494-1-AP, 1:5,000). After incubation, the membranes were washed three times (each for 10 min), and incubated with relevant fluorescence-conjugated secondary antibodies for 2 h at room temperature. Finally, the membranes were detected by an Odyssey Infrared Imaging System (LI-COR, Lincoln, NE, USA), and quantitatively analyzed by ImageJ (Java 8, National Institutes of Health, USA).

The detailed experimental methods for sample preparation before LC-MS/MS analysis can be found in Supplementary Files. For whole-cell proteome analysis, the tryptic peptides were loaded on an EASY-nLC1200 UPLC system (ThermoFisher Scientific). Tryptic peptides were dissolved in solvent A (0.1% formic acid, 2% acetonitrile/in water), separated with solvent B (0.1% formic acid in 90% acetonitrile) via a gradient from 8 to 23% (0–68 min), 23–32% (68–82 min) and climbing to 32–80% (82–86 min) then holding at 80% (86–90 min). Separation was analyzed in Orbitrap Exploris™ 480 (Thermo Fisher Scientific) using a nano-electrospray ion source. The electrospray voltage applied was 2.3 kV and the full MS scan range was set from 400 to 1,200 m/z. Then, more than 25 most abundant precursors were selected for further MS/MS analyses by 20 s dynamic exclusion. The HCD fragmentation was operated at normalized collision energy (NCE) of 27%. The resolution for fragments were 15,000 in the Orbitrap and fixed first mass was set as 110 m/z. Automatic gain control (AGC) target was set at 100%, along with an intensity threshold of 2E4 and a maximum injection time of Auto.

The resulting MS/MS data were processed by Proteome Discoverer (v2.4.1.15) search engine (v.1.6.15.0). Tandem mass spectra were searched against the Mus musculus SwissProt database (20,387 entries) and concatenated with reverse decoy database. The detailed parameter settings are shown in Supplementary Files.

To further analyze hierarchical clustering based on functional classification of differentially expressed proteins (such as: GO, Domain, Pathway), all the categories obtained after enrichment along with their P-values were collated, and those categories that were at least enriched in one of the clusters with P < 0.05 were filtered. We transformed the filtered P-value matrix by the function × = –log10 (P-value). At last, these × values were z-transformed for each functional category, and then clustered by one-way hierarchical clustering (Euclidean distance, average linkage clustering) in Genesis. A heat map with the “heatmap.2” function from the “gplots” R-package can visualize the cluster membership finally.

Parallel reaction monitoring (PRM), a technique based on high-resolution and high-precision mass spectrometers (such as Q Exactive™ Plus) can screen the remaining samples of LC-MS/MS proteome for further verification. The sample was slowly added to the final concentration of 20% v/v TCA to precipitate protein, then vortexed to mix and incubated for 2 h at 4^°C. The precipitate was collected by centrifugation at 4,500 g for 5 min at 4^°C. The precipitated protein was washed with pre-cooled acetone for 3 times and dried for 2 h. The protein sample was then redissolved in 200 mM TEAB and ultrasonically dispersed. Trypsin was added at 1:50 trypsin-to-protein mass ratio for the first digestion overnight. The sample was reduced with 5 mM dithiothreitol for 30 min at 56^°C and alkylated with 11 mM iodoacetamide for 15 min at room temperature in darkness. The tryptic peptides were dissolved in 0.1% formic acid (solvent A), and were mounted to a home-made reversed-phase analytical column directly, with a gradient of 6–20% solvent B (0–40 min), then 23–35% (14 min), 20–30% (40–52 min), 30–80% (52–56 min) then 80% (56–60 min) on an EASY-nLC1000 UPLC system using the flow rate of 500 nL/min constantly. Later, risk assessment was performed in the proteins with fold change no less than 1.2 and P < 0.05, according to peptide hits, coverage, signal strength and number of second order maps. Risk was divided into high, medium and low risk of three gradients. After grading the peptide residues, they were ionized into the NSI ion source and then analyzed by Q ExactiveTM Plus mass spectrometry. The resulting MS data were processed using Skyline (v.21.1). Each protein in the experiment applied one, two or more unique peptides for quantitative analysis (Chai et al., 2020). Additional Methods showed the full details in Supplementary Files.

Unpaired t-test was used to calculate significant differences between two groups and one-way ANOVA was used to assess significant differences within multiple groups. Data were analyzed by SPSS-23 software (IBM SPSS Statistics 23, USA) and GraphPad Prism software (5.0.0; GraphPad Software, USA) were applied to graphing. Statistical significance was defined as P < 0.05. Data were expressed as the mean ± SEM.

To evaluate whether ET-1/ET-Rs signaling axis prompt neuron degeneration in ALS, we first observed the expression of ET-1, ET-A, and ET-B receptors in the spinal cords of SOD1-G93A mice at different disease stages using immunohistochemistry (Figure 1). An obvious increase of ET-1 positive cells, which were clearly glia-like cells (Figure 1A, arrows), was observed in the lumbar spinal cord of SOD1-G93A mice, especially at the end stage of disease. In NTG mice, ET-A receptor was observed mainly in nucleus, both neurons (Figure 1A, arrowheads) and glial cells (Figure 1A, arrows). With disease progression, ET-A-positive motor neurons reduced, while the total number of ET-A-positive nuclei increased. In addition, we found that ET-B receptor was mainly located in the cytoplasm of neurons (Figure 1A, arrowheads), and an obvious decrease of ET-B expression was observed with disease progression (Figure 1A). Consistent with morphological findings, western blotting results revealed an increase of ET-1 expression and decrease of two ET-Rs expression with disease progression (Figure 1B). Apparently, significant expression changes of ET-1 and ET-B receptor were observed at the end stage of SOD1-G93A mice (Figures 1C,E). For ET-A expression, although a decreasing trend was observed, statistically significant difference was not reached even at the end stage (Figure 1D). Together, above-mentioned results indicate that both glial cells and neurons play a vital role in ET-1/ET-Rs pathway thus contributing to ALS pathogenesis.

To further test the cellular localization of ET-1 and ET-Rs, we then performed double immunofluorescence staining. ET-1 positive cells were mainly observed in proliferating GFAP-positive astrocytes in lumbar spinal cord of ALS mice (Supplementary Figure 1), which was in line with the previous opinions (D’haeseleer et al., 2013; Hostenbach et al., 2016). In normal spinal cord of NTG mice, ET-A expression was primarily located in the nucleus of neurons and glial cells, especially motor neurons. In the spinal cord of ALS mice at end stage, ET-A-positive motor neurons decreased obviously (Figure 2A, arrowheads). However, obvious expression of ET-A was identified in mature oligodendrocytes (APC-positive) (Figure 2A, arrows). Furthermore, just as reported (Ranno et al., 2014), increased ET-A expression was also notably observed on proliferating astrocytes in SOD1-G93A mice compared with NTG mice. In addition, certain ET-A positive microglia was also observed in the spinal cord of SOD1-G93A mice (Figure 2A, arrows).

Similarly, we detected the cellular localization of ET-B and found that ET-B was expressed primarily in the cytoplasm of motor neurons, which was in line with the previous study (Michinaga et al., 2013; Figure 2B, arrowheads). Sparse expression of ET-B in APC was also observed in the lumbar spinal cords of SOD1-G93A mice (Figure 2B, arrows). Interestingly, colocalization of ET-B with GFAP-positive astrocytes was occasionally seen in SOD1-G93A mice (Figure 2B, arrows) and no obvious colocalization of ET-B with Iba1-positive microglia was detected in our experiments either in SOD1-G93A mice or NTG mice (Figure 2B).

Since motor neurons are the main cells expressing both ET-A and ET-B receptors in the normal NTG spinal cord and they degenerate and loss with disease progression, we further ask whether the decrease in ET-A and ET-B expression reflects the general loss of motor neurons. Quantitative analysis showed that the loss of NeuN⁺ motor neurons was parallel with the reduce of both ET-A and ET-B expressions in the spinal cord of SOD1-G93A mice at end stage (Figures 2C–E), which to a great extent explains that the reduce of ET-A and ET-B is largely due to the degeneration of motor neurons in SOD1-G93A mice. However, expression of ET-A and ET-B in proliferating glial cells may partly compensate this reduce.

Based on the above results, we proposed that astrocytic ET-1 might have an effect on motor neurons through ET-Rs. To assess the role of ET-1 on motor neurons, we used NSC34 cells stably expressing human SOD1, either wildtype (NSC34-hSOD1WT) or G93A mutated form (NSC34-hSOD1G93A), and NSC34 cells expressing only GFP (NSC34-E) was used as control and conducted a series of experiments (Figure 3A). In order to avoid interference of GFP fluorescence at channel 488, the appropriate secondary antibody (Alexa Fluor 647 or 594, Invitrogen) was used in the cell model.

First, reliability of the cell line model of ALS was confirmed. Western blotting showed that both NSC34-hSOD1 and NSC34-hSOD1G93A expressed GFP-human SOD1 (GFP-hSOD1) fusion proteins, while only GFP was observed in NSC34-E cells (Supplementary Figure 2). Successful transfection was further confirmed by tracing GFP-positive NSC34 cells through immunofluorescence staining for GFP and DAPI. The proportion of GFP-positive NSC34 cells is on average 61 ± 3% (NSC34-E), 66 ± 4% (NSC34-hSOD1WT), and 55 ± 9% (NSC34-hSOD1G93A) (Supplementary Figure 2). CCK-8 assay showed significantly lower cell viability in NSC34-hSOD1 cells, especially NSC34-hSOD1G93A cells compared with NSC34-E cells (Figure 3B). In addition, NSC34-hSOD1G93A cells displayed worse morphological differentiation characterized by less extending neurites which were clearly distinguishable after 24 h compared with NSC34-hSOD1WT or NSC34-E cells (Figure 3C).

Second, we examined the expression of MAP1B, which had been proved to influence the organization and dynamics of microtubules in growing and regenerating axons and growth cones (Vaz et al., 2015; Bora et al., 2021; Strohm et al., 2022), to evaluate the ability of differentiation of different cell lines. Even though the three cell groups extended neurites after 24 h, this characteristic was damaged in NSC34-hSODG93A cells. NSC34-E and NSC34-hSOD1WT cells revealed evident increase of neurites upon differentiation, while the neurites of NSC34-hSOD1G93A cells were not obvious (Figure 4A). By the concentric circle (Sholl’s) analysis (Yang et al., 2016), we found that the neurite arborization of NSC34-E and NSC34-hSOD1WT cells had no significant changes while NSC34-hSODG93A cells were poorly differentiated (Figure 4D). The above results demonstrated that the constitutive expression of hSOD1G93A damaged the differentiate ability of NSC34 cells. Meanwhile, it was reported that cFOS can be activated after different stimuli in the CNS. For example, seizure induction caused expression changes of cFOS in the brain (Yang et al., 2019). Simultaneously, cFOS and c-jun take effect in the pathogenesis of AD, which might reflect the initiation of a cell death program in some neurons (Herrera and Robertson, 1996). Therefore, we wonder whether transfection of mutant SOD1 cause expression change of cFOS. Indeed, a remarkably high expression of cFOS was observed in NSC34-hSODG93A cells suggesting that mutant hSOD1G93A may be a kind of stimuli to cells (Figures 4B,E). We further investigated the expression of GFP-hSOD1 via immunofluorescence. Different from the uniform hSOD1 distribution in NSC34-hSOD1WT cells, dots indicating protein aggregates was observed in NSC34-hSOD1G93A cells (Figures 4C,F). In addition, consistent with that observed in vivo, double-labeling revealed that ET-A and ET-B receptors were both expressed in NSC34 cells (anti-SMI-32 labeled), but no significant expression difference of the two receptors was found among the three cell groups (Figures 4G,H).

ET-1 in the brain is believed to be deeply involved in the central autonomous control and the subsequent cardiac respiratory homeostasis. It may play a role as a neuromodulator or hormone in the way of autocrine/paracrine locally, or be widely transmitted through cerebrospinal fluid (CSF). The concentration of ET-1 in CSF is higher than that in plasma. Through literature review, we have learned about the physiological, the pathological range of ET-1 in the brain, or the comparison of both (Tu et al., 2015). In a previous study, ET-1 exerted a toxic effect on the motor neuron cultures in a time- and concentration-dependent manner, with an exposure to 100–200 nM ET-1 for 48 h resulting in 40–50% cell death, which can indirectly illustrate the sensitive concentration range of neurons for ET-1 (Ranno et al., 2014). To further examine the effect of ET-1 on MNs survival in our experiment, cells were treated for different lengths of time and with different concentrations of the peptide. An exposure to 1, 10, 100, 1,000 nM ET-1 for 24 h revealed a concentration-dependent toxic effect on the three cell groups (Figure 5A). The CCK-8 results showed that the viability of NSC34-hSOD1G93A cells as well as NSC34-hSOD1WT cells decreased as the ET-1concentration increased, and the same ET-1 intervention caused a slight but no statistically significant decrease of cells viability in NSC34-E group. The relatively low concentrations (100 nM) which had been used frequently (Ranno et al., 2014) to study the toxic effects of ET-1 was also approximate near the median inhibitory concentration (IC50) in our experiment, therefore, 100 nM was chosen for further experiments. Furthermore, an exposure to 100 nM ET-1 for 12, 24, and 48 h revealed a time-dependent toxic effect to MNs reaching a plateau at 24 h (Figure 5B). To validate the involvement of ET-A and ET-B in the process of MNs death, the ET-A competitive receptor antagonist BQ123 and ET-B competitive receptor antagonist BQ788 were used. A treatment of BQ123 or BQ788 together with ET-1 rescued the decrease of cell viability caused by ET-1 (Figure 5C) suggesting the involvement of ET-A and ET-B receptors. The ET-1 effect was obviously reversed by BQ123 (2 μM) and BQ788 (2 μM). Then, we asked whether ET-1 influenced the expression of ET-Rs on the cell model. Double-labeling revealed that ET-1 impaired ET-A and ET-B expression in NSC34-hSOD1G93A cells. When BQ123 or BQ788 was added, ET-Rs expression was greatly preserved (Figures 5D,E). For further quantification, western blot analysis was performed. No significant difference of ET-A or ET-B expression was observed among NSC34-E, NSC34-hSOD1WT and NSC34-hSOD1G93A cells. ET-1 treatment on NSC34-hSOD1G93A cells caused significant decrease of ET-B expression while no significant effect was reached on ET-A expression. In addition, BQ788 rescued the ET-B reduction caused by ET-1 treatment (Figures 5F–H).

The present results suggested that pathologically increased ET-1 might have a potential risk of inducing neuronal death and via using pharmacological approaches could potentially reverse the harmful effect. However, caution needs to be paid to further investigate the changes of downstream signaling pathways and function-specificity behind ET-1/ET-Rs signaling axis.

To investigate the underlying pathogenesis of ET-1/ET-Rs signaling axis, we performed label-free LC-MS/MS proteomics analysis on NSC34-hSOD1WT (WT) cells, NSC34-hSOD1G93A (C) cells and NSC34-hSOD1G93A treated with 100 nM ET-1 (S) cells (Figure 6A). The above three groups were analyzed, and heatmap graphs of the hierarchical cluster were constructed to visualize the overall broad modulations of proteomes (Figure 6B). For confirmation of differentially expressed proteins (DEPs), the cutoffs for the fold change of abundance and P-value were set to 1.5 and 0.05, respectively. In this experiment, compared with WT cells, there were 322 up-regulated and 206 down-regulated DEPs in C cells (Figure 6C). 110 proteins (54 up-regulated proteins and 56 down-regulated proteins) were significantly modulated in S cells vs. C cells (Figure 6D). Interestingly, between the two comparable C/WT and S/C groups, 5 common up-regulated proteins (Figure 6E) and 2 common down-regulated proteins (Figure 6F) were identified. However, when the opposite tendency was analyzed, 11 common DEPs were found between the two comparable C/WT and S/C groups (Figures 6G,H). A total of 29 common proteins were identified in the C/WT and S/C groups and listed in the table (Supplementary Table 1). Subsequently, we found that some of the common proteins-ATP synthase protein 8 (Mtatp8) (Supplementary Figure 3), BCL2/adenovirus E1B 19 kDa protein-interacting protein 3-like (Bnip3l) (Figure 7D), ATP synthase-coupling factor 6, mitochondrial (Atp5pf) (Figure 7D), Valine-tRNA ligase, mitochondrial (Vars2) (Supplementary Figure 3) and Protein KIBRA (Wwc1) (Figure 7D) were strongly enriched in reactive oxygen species (Figure 7D), oxidative phosphorylation (Supplementary Figure 3), mitophagy-animal (Figure 7D), Alzheimer disease (Figure 7D), Parkinson disease (Figure 7D), Huntington disease (Figure 7D), Prion disease (Figure 7D), aminoacyl-tRNA biosynthesis (Supplementary Figure 3), ABC transporters (Figure 7D), Hippo signaling pathway (Figure 7D) and so on through Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis, which revealed that ET-1 treatment might play a role in these signaling pathways.

To investigate the associated functions of the identified DEPs in the cells model based on ALS disease, we obtained distinctive KEGG pathway and Gene Ontology (GO) terms analysis between the two comparable groups of C/WT and S/WT (Supplementary Figure 3 and Figure 7). The KEGG pathway is used for analyzing the correlation among known molecular functions such as metabolic pathways, formation of complexes, and biochemical reactions (Wang et al., 2019). As it may be seen in Supplementary Figure 3 of C/WT group, KEGG pathway analysis suggested that the up- and down-regulated DEPs were enriched in oxidative phosphorylation, reactive oxygen species, various neurodegenerative disease such as Parkinson disease and Alzheimer disease. For proteins that enriched in the S/C group, the KEGG analysis showed a strong interaction with a series of pathways related with neurodegenerative disease, such as Hippo signaling pathway-multiple species (Gogia et al., 2019), ABC transporters (Pahnke et al., 2021), oxidative phosphorylation, and ErbB signaling pathway (Supplementary Figure 3; Sun et al., 2020). In summary, these results suggest that ET-1 could induce a series of events in the cells model. In the part of GO enrichment analysis, C/WT group was highly enriched in extracellular region and synaptic cleft based on cellular component while S/C group was related to Gul4-RING E3 ubiquitin ligase complex and nuclear euchromatin (Figure 7A). In C/WT comparable group, significant changes occurred in some molecular functions like protein tyrosine kinase inhibitor activity, sphingomyelin phosphodiesterase activity while more DEPs were associated with xenobiotic transmembrane transporter activity, chromatin DNA binding and MAP kinase activity in the S/C group (Figure 7B). According to biological process enrichment, in C/WT group, mitochondrial membrane organization, mitochondrial transmembrane transport and organophosphate biosynthetic process were greatly enriched, while cellular response to laminar fluid shear stress, mitotic cell cycle phase transition and drug export were significantly involved in the S/C group (Figure 7C). KEGG pathway analysis showed that DEPs are located in important pathways such as allograft rejection and prion disease in C/WT group, however, in S/C group, DEPs were specifically enriched in Hippo signaling pathway-multiple species, ErbB signaling pathway and so on (Figure 7D).

The results of proteomic analysis exhibited that compared with C cells group (as positive control), there were 54 upregulated and 56 downregulated proteins after ET-1 intervention (Figure 8A). Gene ontology (GO) functional classification analysis showed that the DEPs in S/C comparable group mainly fell into cellular process, biological regulation, metabolic process, response to stimulus, multicellular organismal process, developmental process, localization and other biological process (Supplementary Figure 4). The main results of the GO enrichment analysis are shown in Figures 8B–E. In particular, the DEPs based on cell components were enriched mainly in nuclear euchromatin and proton-transporting ATP synthase complex (Figure 8B). Molecular functions were enriched mainly in terms related to xenobiotic transmembrane transporter activity, MAP kinase activity, translation repressor activity and chromatin DNA binding (Figure 8C). Then, the DEPs were enriched in drug export, regulation of intestinal absorption, cellular response to laminar fluid shear stress, positive regulation of necrotic cell death, C2-steroid hormone metabolic process, hormone biosynthetic process, hippo signaling and other biological process (Figure 8D). Meanwhile, KEGG analysis showed that the DEPs were primarily enriched in Hippo signaling pathway-multiple species and ABC transporters (Figure 8E). By combining the results, we found that ET-1 significantly modulated protein metabolic process, signal transduction and response to stimulus. The subcellular localization analysis showed that most of the DEPs were distributed in the nucleus (27.46%, 30 proteins), cytoplasm (23.3%, 27 proteins), mitochondria (17.42%, 19 proteins), extracellular (12.88%, 14 proteins), and plasma membrane (10.98%, 12 proteins). The data revealed that except for the nucleus, the largest proportion of DEPs was located in cytoplasm, followed by mitochondria and extracellular (Supplementary Figure 4).

To further confirm the cells proteome results related with ET-1 treatment, we performed PRM assay rather than the traditional western blot, which selected specific peptides or target peptides and then made the target protein quantified (Xu et al., 2018). According to risk assessment and limited to characteristics and abundances of 110 DEPs in the S/C comparable group, 16 target proteins were evaluated and then quantified in the experiment. At present, a total of 16 DEPs which had not previously been clearly reported to be relevant with ET-1 intervention on NSC34-hSOD1G93A cells was verified, including 9 up-regulated proteins and 7 down-regulated proteins identified in the basis of label-free quantitative analysis. The fragment ion peak of one peptide and corresponding proteins were displayed in Supplementary Table 2. 16 of identified proteins presented the same changing trend and the former 13 proteins that were identified significantly altered, demonstrating the consistency between PRM and LC-MS/MS proteome (Supplementary Table 3). These results suggested that the related effects of ET-1 on the NSC34-hSOD1G93A cells may closely connected with the functions of these proteins. Thus, ET-1 might influence kinds of signaling pathways to impose MNs damaging effects.

Among these verified differential proteins, we noticed that Cdkn1b (Figures 9A,B) and Eif4ebp1 (Figure 9C), significantly down-regulated when treated with ET-1, were particularly enriched in ErbB signaling pathway (pathway ID# mmu04012, Supplementary Figure 5). ErbB pathway dysregulation has been proposed to play a pathogenic role in ALS, especially ErbB4, a member of the epidermal growth factor (EGF) subfamily of receptor tyrosine kinases (RTKs) (Casanovas et al., 2017). These results suggest that the effects of ET-1 on NSC34-hSOD1G93A cells are closely concerned with the functions of mentioned proteins above and ET-1 can exert the related effects mainly through cell cycle and protein synthesis regulation. These proteins are anticipating targets for the study of ALS pathogenesis (Figure 10).