Abstract Amyotrophic lateral sclerosis (ALS) is a multifactorial motor neuron (MN) disease, characterized by several cellular dysfunctions, many of which are shared by different neurodegenerative diseases. Here, we investigated whether a stressful lifestyle might exacerbate the altered mechanisms and affect the disease progression in ALS-predisposed conditions. To model stress in vivo, SOD1^G93A mice underwent a chronic unpredicted mild stress protocol. This resulted in a significant impairment in body weight gain and motor performance, in a gender-specific manner. Moreover, the gene expression of Col1a1, Col1a2 and Il6 was strongly dysregulated in motor cortex and/or spinal cord of stressed mice. To assess the direct impact of stress on MNs, NSC-34 hSOD1^G93A cells underwent oxygen and glucose deprivation. Compared to NSC-34 hSOD1^WT, mutated MNs exhibited a reduced capacity to cope with stress. By performing gene expression, protein-protein interaction, gene ontology and pathway enrichment analyses, we also revealed the pivotal role of the PI3K/Akt and focal adhesion pathways (triggered by Gsk3b, Il6, Igf1 and/or collagen) in mediating stress response. Similar results were observed in stressed human iPSCs-derived TARDBP^G298S MNs. In conclusion, our results suggest that the PI3K/Akt and focal adhesion pathways play a crucial role in stress response across different ALS-predisposed models: the study paves the way for novel therapeutic targets and highlights the relevance of a healthy lifestyle. Supplementary Information The online version contains supplementary material available at 10.1038/s41598-025-02167-9. Keywords: Bioinformatic analysis, Exposome, Molecular mechanisms, Neuromuscular disease, Stressor Subject terms: Cell biology, Neuroscience, Diseases Introduction Amyotrophic lateral sclerosis (ALS) is a motor neuron (MN) disease due to the degeneration of upper and lower MNs. The disease is characterized by progressive skeletal muscle weakness and atrophy, until respiratory failure 2–3 years after diagnosis. About 5–10% of patients are affected by the familial form of ALS (fALS), while the remaining 90–95% present a sporadic form (sALS)^[38]1,[39]2. More than 30 genes have been associated with ALS aetiology, including SOD1, TARDBP, FUS and C9orf72, which are responsible for about 60% of fALS cases. Mutations of the same genes are also frequently found in sALS cases^[40]3. Interestingly, environmental factors can also play a role in the development of the disease^[41]4. Several molecular mechanisms characterize ALS pathophysiology, including oxidative stress, excitotoxicity, glial dysfunctions, neuroinflammation, mitochondrial alterations, dysregulated nucleocytoplasmic and vesicle transport, impaired protein homeostasis and endoplasmic reticulum stress, axonopathy, impaired DNA damage repair and aberrant RNA metabolism^[42]1. However, several “stressors”, i.e. any kind of physical or psychological disturbance that can affect the optimal functioning of the organism, could also influence ALS pathogenesis, by exacerbating the above-mentioned altered mechanisms^[43]5. Indeed, the stress response can act on the sympathetic-adreno-medullary axis and hypothalamus-pituitary-adrenal (HPA) axis: the first releases noradrenaline and norepinephrine, while the second secretes glucocorticoids. Both cascades result in short- or long-term consequences throughout the body, including the CNS^[44]6. Acute stress triggers a rapid response to facilitate and accelerate adaptive responses: the reduction and/or increase in excitatory or inhibitory activity, in specific CNS areas, is modulated to restore physiological balance and increase resilience^[45]6,[46]7. However, a prolonged exposure to stressful events, typical of chronic stress, increases ROS, triggers neuroinflammation and cell death, alters cellular morphology and decreases axon activity^[47]8. Unfortunately, stress is increasing in our modern society due to the global and lifestyle changes, potentially affecting the onset and progression of neurodegenerative diseases in particular in predisposed individuals, while healthy habits could counteract the acceleration of these pathologies^[48]9,[49]10. In this regard, the present study aimed to investigate the stress impact on ALS pathogenesis and to identify the specific molecular mechanisms and genes involved. We used both in vivo and in vitro models: first, the stress effects were evaluated in vivo using hSOD1^G93A mice to investigate the expression of several ALS-related genes in both motor cortex and lumbar spinal cord. Then, to specifically study the stress effect on MNs, we used murine hSOD1^G93A NSC-34 cell lines and human TARDBP^G298S iPSCs-derived MNs, which underwent oxygen-glucose deprivation (OGD), with the aim to further explore stress-affected genes and molecular pathways. Results In vivo study of stress effects in female and male hSOD1^G93A mice To verify whether stress could affect disease progression in ALS-predisposed conditions, its effect was evaluated in hSOD1^G93A mice. To this end, mice were subjected to a 28-day chronic unpredicted mild stress protocol^[50]11,[51]12, followed by behavioural observations and molecular analyses (RT-qPCR) (Fig. [52]1; Fig. [53]S1). Fig. 1. Fig. 1 [54]Open in a new tab Chronic unpredicted mild stress effects in female and male ALS-predisposed mice. A-B. Rotarod and weight results in female hSOD1^G93A are showed at starting and end point in STRESSED and CTRL (not stressed) groups. Results are represented as median and quartile range of three mice per group, considering the average of three trials. Mixed-effects analysis followed by Uncorrected Fisher’s LDS analysis was used for statistical results. Rotarod of female hSOD1^G93A STRESSED at end point vs. female hSOD1^G93A STRESSED at starting point **p < 0.01; weight of female hSOD1^G93A CTRL at end point vs. female hSOD1^G93A CTRL at starting point *p < 0.05. C-D. Rotarod and weight results in male hSOD1^G93A are described at starting and end point in STRESSED and CTRL (not stressed) groups. Results are represented as median and quartile range of three mice per groups, considering the average of three trials. Mixed-effects analysis followed by Uncorrected Fisher’s LDS analysis was used for statistical results. Weight of male hSOD1^G93A CTRL at end point vs. male hSOD1^G93A CTRL at starting point **p < 0.01. E. Heatmap graph (created using GraphPad Prism version 10.1.2 for Windows) shows the relative gene expression in motor cortex and lumbar spinal cord samples of STRESSED groups normalized to CTRL (distinguishing female and male groups). Results are showed as red-to-green scale, using data of three mice per groups and indicating 1 (black) as baseline reference point of female or male hSOD1^G93A CTRL (not shown). Two-way ANOVA followed by Sidák’s multiple comparison test was used as statistical analysis and the significantly results were listed in Table [55]1. Concerning the females, the rotarod performance was significantly reduced at the end of the stress period (end point) compared to the beginning (starting point), whereas no differences were observed in males (Fig. [56]1A and C). Body weight remained unchanged after the 28 day-stress protocol, in both female and male ALS stressed mice (Fig. [57]1B and D). In contrast, it was significantly increased in absence of stress, as expected for young growing mice. Regarding the RT-qPCR analysis, the relative gene expression of 39 ALS-correlated genes (see gene list in Table [58]S1; Prime PCR Disease State Panels, Bio-Rad Laboratories) was investigated in both motor cortex and lumbar spinal cord samples, by normalizing the values of stressed hSOD1^G93A mice (STRESSED) to not-stressed hSOD1^G93A mice (CTRL) (Fig. [59]1E). As summarized in Table [60]1, Col1a1 was significantly up-regulated in stressed females as compared to CTRL ones, both in the motor cortex and lumbar spinal cord. Col1a2 was also up-regulated in the female group, but only in the motor cortex. Instead, Il6 was significantly up-regulated in stressed ALS male mice compared to CTRL ones in the motor cortex, whereas no statistically significant differences in the expression of other genes were observed in lumbar spinal cord samples. Table 1. Genes significantly deregulated in Fig. [61]1. Gene Comparison Tissue Sex p-value Col1a1 STRESSED vs. CTRL Motor cortex Female 0.0021 Col1a2 STRESSED vs. CTRL Motor cortex Female 0.0055 Col1a1 STRESSED vs. CTRL Spinal cord Female < 0.0001 Il6 STRESSED vs. CTRL Motor cortex Male < 0.0001 [62]Open in a new tab Significant statistically differences referred to relative gene expression (2^–∆∆Ct method) of hSOD1^G93A stressed (STRESSED) mice normalized to hSOD1^G93A non-stressed (CTRL) mice per each tissue. Two-way ANOVA was applied as statistical analysis. Characterization of NSC-34 differentiation in MN-like cells To investigate the direct effect of stress on ALS MNs, we first used NSC-34 MN-like cells, a hybrid cell line derived by spinal cords of mouse embryos and mouse neuroblastoma cells. First, we defined the cell differentiation conditions. Differentiation of NSC-34 cells (naïve, hSOD1^WT and hSOD1^G93A) into MN-like cells was obtained by using serum starvation and RA treatment: various RA concentrations were tested at different DIV (Days In Vitro) by the MTT assay to identify the highest non-toxic concentration that did not interfere with the doxycycline effect (needed for the hSOD1 gene activation). By testing different RA concentrations (namely 1, 5, 10, 15 and 20 µM) and normalizing data to naïve NSC-34 cells, the cell viability of hSOD1^G93A cells was significantly different compared to the hSOD1^WT ones, almost at each DIV (Fig. [63]S2). However, by analysing each cell type independently (Fig. S3), we identified a 4 day-RA treatment as the optimal condition to obtain MN-like cell differentiation in each condition, in agreement with the literature^[64]13: in fact cell viability was unchanged at DIV4 in all cell lines, w/wo doxycycline, using any RA concentration as compared to control condition (1% serum) (Fig. S3A-F). On the contrary, cell viability at DIV2, 6 and 8 was significantly modified with almost all the RA concentrations. Then, the following morphological analyses were performed for all NSC-34 cell lines combining serum starvation (1% serum) with 20 µM of RA for 4 DIV (i.e. the highest non-toxic concentration at 4 DIV), in order to assess the MN differentiation. As shown in Fig. [65]2, in presence of RA, cell-body clusters, neurite length and branching were significantly increased in NSC-34 naïve (Fig. [66]2A-D), hSOD1^WT (Fig. [67]2E-H) and hSOD1^G93A (Fig. [68]2I-L), w/wo doxycycline, as compared to control conditions (1% serum w/o RA). The MN maturation was demonstrated by analysing ChAT expression by immunofluorescence (IF): by culturing cells in 1% serum and RA 20 µM, w/wo doxycycline, the corrected total cell fluorescence (CTCF) was significantly increased in NSC-34 naïve (Fig. [69]2M-N), hSOD1^WT (Fig. [70]2O-P) and hSOD1^G93A (Fig. [71]2Q-R), compared to control cells (1% serum only). No significant differences in morphological parameters, nor in ChAT expression have been observed among the cell lines, upon doxycycline administration (Fig. S4). These results confirm that all NSC-34 lines underwent proper differentiation into MN-like cells, supporting the conclusion that any subsequent differences observed could not be due to variability in the differentiation process. Fig. 2. [72]Fig. 2 [73]Open in a new tab Morphological analysis related to RA treatment (20 µM for 4 days). A, E, I. Representative images acquired using the built-in Basler Ace 1920 –155 μm camera of the Incucyte system. Scale bar = 200 μm. B-D, F-H, J-L. Cell-body cluster, neurite length and neurite branch point are significantly increased after RA treatment, w/wo doxycycline in NSC-34 naïve, hSOD1^WT and hSOD1^G93A. Results are represented as median and quartile range of three independent experiments. Two-way ANOVA followed by Sidák’s multiple comparisons test was used as statistical analysis. 1% serum + RA 20 µM vs. 1% serum *p < 0.05, **p < 0.01 and ***<0.001. M, O, Q. NSC-34 naïve, hSOD1^WT hSOD1^G93A treated with RA and immunolabeled by anti-ChAT antibody (red). Representative images acquired by Eclipse E600 (Microfire Camera 2-Megapixel Color Imaging, 1600 × 1200). Scale bar = 50 μm. N, P, R. CTCF is significantly increased in NSC-34 naïve, hSOD1^WT hSOD1^G93A, w/wo doxycycline. Results are represented as median and quartile range of three independent experiments (n ≥ 10 neurons for each experiment). Two-way ANOVA followed by Sidák’s multiple comparisons test was used as statistical analysis. 1% serum + RA 20 µM vs. 1% serum *p < 0.05, **p < 0.01 and ***<0.001. Until this point, naïve cells were included to validate the cell differentiation protocol and exclude potential interferences between the RA administration and the doxycycline treatment, and vice versa. However, from this point onward forward, we have excluded naïve cells to avoid inappropriate comparisons between different experimental systems: indeed hSOD1 NSC-34 cells express a hemagglutinin’s epitope, making them a modified system. Thus, hSOD1^WT will be considered the most appropriate control for the following analyses and statistical comparisons. Oxygen-glucose deprivation (OGD) as a stress model in vitro OGD was used to model a chronic stress in vitro. We evaluated different concentrations of CoCl[2] (a hypoxia-mimetic agent) in low glucose medium for 24 h, using the MTT assay to determine the dose inducing approximatively 50% cell death. As shown in Fig. [74]3A, all the employed CoCl[2] concentrations (50, 100, 200, 300 and 400 µM) significantly reduced cell viability in both NSC-34 hSOD1^WT and hSOD1^G93A cells, as compared to low glucose condition. However, 100 µM CoCl[2] was the lowest concentration that induced 50% cell loss and revealed a significant difference between NSC-34 hSOD1^WT and hSOD1^G93A (Fig. [75]3A). Therefore, this concentration was used in subsequent experiments, starting with the OGD stress effect validation. Fig. 3. [76]Fig. 3 [77]Open in a new tab OGD stress validation in NSC-34 hSOD1 cells. (A) The in vitro stress model was established using OGD conditions, by culturing NSC-34 cells in low glucose (LG) medium and CoCl[2]. All the tested concentrations of CoCl[2] (50, 100, 200, 300 and 400 µM) are able to induce a reduction of cell viability, both in NSC-34 hSOD1^WT and hSOD1^G93A, compared to LG culture; ~50% of cell death is observed using 100 µM of CoCl[2]. Results are represented as mean ± SD of three independent experiments. Two-way ANOVA followed by Sidák’s multiple comparisons test was used for comparing different LG and CoCl[2] concentrations in each cell line and different cell lines in each experimental condition. LG + CoCl[2] 50 µM/ 100 µM/ 200 µM/ 300 µM/ 400 µM vs. LG *p < 0.05, **p < 0.01 and ****p < 0.0001. NSC-34 hSOD1^G93A vs. NSC-34 hSOD1^WT ##p < 0.01. (B) NSC-34 hSOD1^WT and hSOD1^G93A, CTRL or STRESSED, were treated with MitoTracker Red CMXRos. Representative images acquired in live imaging by built-in Basler Ace 1920 –155 μm camera of the Incucyte system. (C) Quantifications of orange signal (visible in red), related to mitochondria membrane potential, performed using automated fluorescent analysis of the Incucyte system. Total orange object integrated intensity results is normalized to phase object count. Results are represented as median and quartile range of six independent experiments. Two-way ANOVA followed by Tukey’s multiple comparison test was used as statistical analysis. NSC-34 hSOD1^G93A CTRL vs. NSC-34 hSOD1^WT STRESSED and NSC-34 hSOD1^G93A STRESSED vs. NSC-34 hSOD1^G93A CTRL *p < 0.05. (D) Representative images showing HIF1α and cleaved caspase 3 protein levels of NSC-34 hSOD1^WT and hSOD1^G93A, in both experimental conditions (CTRL and STRESSED). NG stands for Normal Glucose. E-F. Protein expression levels quantified by normalizing all experimental conditions vs. NSC-34 hSOD1^WT CTRL. Full-length blots are shown in Fig. S5. Results are represented as median and quartile range of three independent experiments. Two-way ANOVA followed by Tukey’s multiple comparison test was used as statistical analysis. NSC-34 hSOD1^WT STRESSED vs. NSC-34 hSOD1^WT CTRL *p < 0.05 and **p < 0.01; hSOD1^G93A STRESSED vs. NSC-34 hSOD1^WT STRESSED *p < 0.05; hSOD1^G93A CTRL vs. NSC-34 hSOD1^WT CTRL*p < 0.05. First, we evaluated the mitochondrial potential membrane status in NSC-34 hSOD1^WT and hSOD1^G93A cells, both in stressed and not-stressed (CTRL) conditions. We analysed the integrated intensity normalized to phase object count using MitoTracker Red CMXRos (Fig. [78]3B-C). The signal intensity showed only a decreasing trend in hSOD1^WT stressed cells respect to CTRL, while it was significantly reduced in hSOD1^G93A stressed cells as compared to CTRL conditions. Even if stress exposure elicited a comparable response trend in both cell lines, it induced significant changes only in hSOD1^G93A cells, highlighting a possible mitochondrial impairment and an increased vulnerability only in stressed NSC-34 hSOD1^G93A cells. These findings suggest a cumulative effect of the SOD1^G93A mutation and cellular stress. To further validate this effect, the expression of specific hypoxic (HIF1α) and pro-apoptotic (cleaved caspase 3) markers was investigated by western blot (Fig. [79]3D-F; Fig. S5). The HIF1α protein levels were significantly up-regulated under stress conditions in hSOD1^WT cells, while no remarkable differences were observed in hSOD1^G93A cells (possibly due to their already impaired state). Interestingly, a significant down-regulation was observed in stressed hSOD1^G93A compared to stressed hSOD1^WT cells (Fig. [80]3E). Similarly, a significant up-regulation of cleaved caspase 3 protein levels was noted in hSOD1^WT cells after stress exposure compared to CTRL condition (Fig. [81]3F). In hSOD1^G93A cells the stress apparently did not induce remarkable effects: cleaved caspase 3 protein levels were significantly increased in not-stressed hSOD1^G93A cells compared to hSOD1^WT ones (Fig. [82]3F). Overall, these results highlight a reduced ability of hSOD1^G93A to endure and respond to stress conditions, also denoting the activation of cell death pathways in CTRL conditions. Gene expression analysis in NSC-34 hSOD1 upon stress exposure Once validated the OGD stress model, the expression of the 39 ALS-correlated genes was analysed by RT-qPCR in NSC-34 hSOD1^WT and hSOD1^G93A in basal conditions (CTRL) or upon stress exposure (STRESSED). As shown in the heatmap of Fig. [83]4A, most of the genes were deregulated under stress conditions, both in WT and mutated cells, in comparison to NSC-34 hSOD1^WT CTRL. Seventeen out of 39 genes were found to be significantly deregulated (the statistical details are listed in Table [84]2) and included: Ang, Bdnf, Casp1, Cldn5, Col1a1, Col4a1, Col4a2, Fgfr1, Gsk3b, Hdac7, Hspb1, Igf1, Il6, Nanog, Nes, Pou5f1 and Tfgb1. Fig. 4. [85]Fig. 4 [86]Open in a new tab Gene expression and relative PPI in CTRL and STRESSED cells. (A) Heatmap graph (created using GraphPad Prism version 10.1.2 for Windows) shows relative gene expression of NSC-34 hSOD1^WT STRESSED, NSC-34 hSOD1^G93A CTRL and STRESSED, normalized to NSC-34 hSOD1^WT CTRL. Results are showed as red-to-green scale of nine independent experiments, indicating 1 (black) as baseline of NSC-34 hSOD1^WT CTRL (not shown). Two-way ANOVA followed by Tukey’s multiple comparisons test was used as statistical analysis and the details were listed in Table [87]2. (B) Heatmap graph (created using GraphPad Prism version 10.1.2 for Windows) shows the relative expression of genes of interest, in both CTRL and STRESSED experimental conditions. Results are showed as red-to-green scale of nine independent experiments, indicating 1 (black) as baseline of NSC-34 hSOD1^WT (not shown). Unpaired t-test statistical analysis for each gene was used: NSC-34 hSOD1^G93A CTRL vs. NSC-34 hSOD1^WT CTRL, NSC-34 hSOD1^G93A STRESSED vs. NSC-34 hSOD1^WT STRESSED. C-D. Bar plot graphs show the genes significantly up-regulated and/or down-regulated, based on the unpaired t-test statistical analysis previously described in CTRL and STRESSED conditions. E-F. PPI, performed by STRING analysis, shows interactions among predicted proteins in CTRL and STRESSED conditions. Table 2. Genes significantly deregulated in Fig. [88]4. Gene Comparison p-value Ang hSOD1^WT STRESSED vs. hSOD1^WT CTRL 0.0055 hSOD1^G93A CTRL vs. hSOD1^WT CTRL < 0.0001 hSOD1^G93A STRESSED vs. hSOD1^WT CTRL 0.0055 hSOD1^G93A CTRL vs. hSOD1^WT STRESSED < 0.0001 hSOD1^G93A STRESSED vs. hSOD1^G93A CTRL < 0.0001 Bdnf hSOD1^WT STRESSED vs. hSOD1^WT CTRL 0.0155 hSOD1^G93A STRESSED vs. hSOD1^WT CTRL 0.0079 hSOD1^G93A CTRL vs. hSOD1^WT STRESSED < 0.0001 hSOD1^G93A STRESSED vs. hSOD1^G93A CTRL < 0.0001 Casp1 hSOD1^G93A CTRL vs. hSOD1^WT CTRL < 0.0001 hSOD1^G93A STRESSED vs. hSOD1^WT CTRL < 0.0001 hSOD1^G93A CTRL vs. hSOD1^WT STRESSED < 0.0001 hSOD1^G93A STRESSED vs. hSOD1^WT STRESSED < 0.0001 hSOD1^G93A STRESSED vs. hSOD1^G93A CTRL < 0.0001 Cldn5 hSOD1^WT STRESSED vs. hSOD1^WT CTRL 0.0424 hSOD1^G93A STRESSED vs. hSOD1^WT CTRL 0.0108 Col1a1 hSOD1^G93A CTRL vs. hSOD1^WT STRESSED 0.0308 hSOD1^G93A STRESSED vs. hSOD1^G93A CTRL 0.0073 Col4a1 hSOD1^G93A CTRL vs. hSOD1^WT STRESSED < 0.0001 hSOD1^G93A STRESSED vs. hSOD1^G93A CTRL 0.0002 Col4a2 hSOD1^G93A STRESSED vs. hSOD1^WT STRESSED 0.0139 Fgfr1 hSOD1^G93A CTRL vs. hSOD1^WT STRESSED 0.0042 hSOD1^G93A STRESSED vs. hSOD1^G93A CTRL 0.0183 Gsk3b hSOD1^G93A CTRL vs. hSOD1^WT STRESSED 0.0396 hSOD1^G93A STRESSED vs. hSOD1^G93A CTRL 0.0032 Hdac7 hSOD1^G93A CTRL vs. hSOD1^WT STRESSED 0.0123 hSOD1^G93A STRESSED vs. hSOD1^G93A CTRL 0.0048 Hspb1 hSOD1^G93A STRESSED vs. hSOD1^G93A CTRL 0.0349 Igf1 hSOD1^G93A CTRL vs. hSOD1^WT CTRL < 0.0001 hSOD1^G93A CTRL vs. hSOD1^WT STRESSED < 0.0001 hSOD1^G93A STRESSED vs. hSOD1^G93A CTRL < 0.0001 Il6 hSOD1^WT STRESSED vs. hSOD1^WT CTRL 0.0906 hSOD1^G93A STRESSED vs. hSOD1^WT CTRL 0.0236 hSOD1^G93A CTRL vs. hSOD1^WT STRESSED 0.0001 hSOD1^G93A STRESSED vs. hSOD1^G93A CTRL < 0.0001 Nanog hSOD1^G93A CTRL vs. hSOD1^WT STRESSED 0.0001 hSOD1^G93A STRESSED vs. hSOD1^G93A CTRL < 0.0001 Nes hSOD1^G93A STRESSED vs. hSOD1^WT CTRL 0.0430 Pou5f hSOD1^WT STRESSED vs. hSOD1^WT CTRL 0.0458 hSOD1^G93A STRESSED vs. hSOD1^WT CTRL 0.0299 Tgfb1 hSOD1^G93A CTRL vs. hSOD1^WT STRESSED 0.0026 hSOD1^G93A STRESSED vs. hSOD1^G93A CTRL 0.0012 [89]Open in a new tab Significant statistically differences referred to relative gene expression (2^–∆∆Ct method) and normalized to NSC-34 hSOD1^WT CTRL. Two-way ANOVA was applied as statistical analysis. To understand whether these specific deregulations were mutation-dependent, i.e. disease-related, and/or triggered by stress in disease-predisposed conditions, the expression level of these 17 genes was analysed by comparing hSOD1^G93A CTRL vs. hSOD1^WT CTRL MNs and, independently, hSOD1^G93A STRESSED vs. hSOD1^WT STRESSED MNs (Fig. [90]4B). These analyses revealed the significant deregulation of 11 and 10 genes in hSOD1^G93A versus to hSOD1^WT MNs in CTRL (Fig. [91]4C) and STRESSED condition, respectively (Fig. [92]4D). As shown in Fig. [93]4C, Cldn5 was significantly down-regulated, while Bdnf, Casp1, Col4a1, Fgfr1, Gsk3b, Hdac7, Igf1, Il6, Nanog and Tgfb1 were up-regulated in hSOD1^G93A CTRL cells compared to hSOD1^WT CTRL. These findings indicate that the expression of these genes is significantly modulated by the presence of the mutation itself. Instead, upon stress exposure, Ang, Cldn5, Gsk3b, Hdac7, Hspb1, Il6 were significantly down-regulated (Table [94]2) in ALS mutated MNs compared to hSOD1^WT cells, whereas Casp1, Col4a2, Igf1 and Tgfb1 were significantly up-regulated (Fig. [95]4D). Therefore, the stress exposure can affect the expression of some genes in hSOD1^G93A STRESSED cells. To predict the interactions among the proteins synthetized by the deregulated genes of interest, we evaluated CTRL and STRESSED conditions through separated protein-protein interaction (PPI) networks by STRING analysis (Fig. [96]4E-F). All the proteins, excluding Hdac7, were predicted to interact among each other in CTRL condition (Fig. [97]4E); while, the PPIs were reduced upon stress exposure: in particular Ang, Col4a2 and Hdac7 did not show any interaction with the other predicted proteins (Fig. [98]4F). Gene ontology (GO) and pathway enrichment analysis were performed on genes previously described in the bar plots of Fig. [99]4C-D for both experimental conditions. As shown in Fig. [100]5A and C, biological processes (BP), cellular components (CC) and molecular function (MF) revealed some similarities between the two groups (i.e. CTRL and STRESSED). However, among the BPs, “positive regulation of protein transport”, “positive regulation of establishment of protein localization”, “positive regulation of protein secretion”, “positive regulation of peptide secretion”, and “regulation of interleukin-1 beta production” were detected only upon stress exposure. Similarly, a limited numbers of CCs and MFs were specific for STRESSED experimental conditions: “basement membrane”, “grow cone”, “site of polarized growth” (in the framework of CCs) and “protein kinase regulator activity”, “kinase regulator activity”, “protein kinase C binding” (referred to MFs). Finally, pathway enrichment analysis identified “PI3K-Akt signalling pathway” as a relevant pathway in both CTRL and STRESSED conditions, while “Focal adhesion” as implicated only in STRESSED one (Fig. [101]5B and D). Although PI3K-Akt signalling pathway was relevant into both analyses, the genes were involved in different manner: by comparing STRESSED and CTRL conditions, we observed that (i) Gsk3b and Il6 were differentially deregulated, (ii) two Col4 chains were implicated (Col4a1 in CTRL analysis and Col4a2 in STRESSED analysis), (iii) only Igf1 was up-regulated in both cases. In addition, Fgfr1 and Bdnf were involved only in CTRL condition. Interestingly, Gsk3b, Igf1 and Col4a2 were also implicated in the “Focal adhesion”, the specific mechanism related to STRESSED analysis. Fig. 5. [102]Fig. 5 [103]Open in a new tab GO and pathway enrichment analysis in CTRL and STRESSED NSC-34 cells. A-B. Log2FC of gene expression related to NSC-34 hSOD1^G93A and NSC-34 hSOD1^WT for CTRL condition, was used to performed GO (including BP, CC and MF in GO three ontology plots) and pathway enrichment analysis by SRplot (free online platform: [104]http://www.bioinformatics.com.cn/SRplot, accessed on 8 May 2024).C-D Log2fc of interesting gene expression related to NSC-34 hSOD1^G93A and NSC-34 hSOD1^WT for STRESSED condition, was used to performed GO (including BP, CC and MF in GO three ontology plots) and pathway enrichment analysis by SRplot (free online platform: [105]http://www.bioinformatics.com.cn/SRplot, accessed on 15 March 2024). Investigation of stress effects in ALS TARDBP^G298S human iPSC-derived MNs In order to extend these analyses to human cells, the same OGD stress was applied to hiPSC-derived MNs, either healthy (56c2) or bearing the TDP43 mutation - TARDBP^G298S - (Fig. [106]6A). Interestingly, TARDBP mutations are more prevalent than SOD1 mutations: therefore, the use of TARDBP mutated cells may support the generalizability of our findings across distinct ALS subtypes. Fig. 6. [107]Fig. 6 [108]Open in a new tab OGD stress effects in human MNs (healthy and TDP-34). A. Representative images acquired using the built-in Basler Ace 1920 –155 μm camera of the Incucyte system. Scale bar = 100 μm. B-C. Neurite length and branch points are significantly changed upon stress exposure in healthy MNs (MN_healthy). TDP-43 MNs (MN_TDP-43) show relevant morphological alterations both in not stressed (CTRL) and STRESSED experimental conditions, as compared to healthy cells. Results are represented as median and quartile range of three independent experiments. Two-way ANOVA followed by Uncorrected Fisher’s LSD analysis was used for statistical results. MN_healthy STRESSED, MN-TDP-43 CRTL and STRESSED vs. MN_healthy CTRL ****p < 0.0001. D-M. Violin plots show relative gene expression of MN_healthy STRESSED and MN_TDP-43 CTRL and STRESSED normalized to MN_healthy CTRL. Results are represented as median and quartile range of three independent experiments. Two-way ANOVA followed by Tukey’s multiple comparison test was used as statistical analysis for each gene was studied: *p < 0.05, **p < 0.01, ***<0.001 and ****p < 0.0001. We observed that the neurite length (Fig. [109]6B) and branch points (Fig. [110]6C) were significantly reduced in TDP-43 MNs as compared to the healthy ones in basal conditions. Upon stress exposure, all the morphological parameters were remarkably changed in healthy cells compared to not stressed condition (CTRL). No significant differences were observed in TDP-43 mutated iPSC-derived MNs upon stress exposure, as compared to CTRL ones. Notably, all the morphometric parameters were significantly affected in mutated cells (both CTRL and STRESSED), compared to healthy MNs CTRL (Fig. [111]6B-C), emphasizing that the morphological alterations affecting TDP-43 cells are already evident in basal conditions. Regarding the molecular aspects, we investigated the expression of the genes previously identified in the murine MN analysis, by normalizing healthy STRESSED, TDP-43 CTRL and TDP-43 STRESSED MNs vs. healthy CTRL MNs. Similarly to hSOD1 mouse cells, despite the different mutation, the majority of the selected genes (in particular Ang, Col1a1, Col1a2, Col4a1, Col4a2, Hdac7, Igf1) were significantly up-regulated in mutated hiPSC-derived MNs compared to healthy ones under CTRL conditions (Fig. [112]6D-M). Moreover, following stress exposure, nearly all analyzed genes (except Casp1 and Il6) were significantly down-regulated in TDP-43 MNs compared to mutated CTRL MNs. On the other hand, upon stress exposure, few selected genes, including Ang, Col1a1 and Col4a2 were significantly up-regulated in TDP-43 MNs compared to healthy ones, while only Gsk3b was significantly down-regulated (Fig. [113]6D-M). Overall, these findings highlight once again the key role of Gsk3b, confirming the previous observations in the NSC-34 cells. Discussion Our lifestyle can positively or negatively affect our health. Among various factors, the importance of regular physical activity, a balanced diet and adequate sleep has been widely studied, and it is well recognized that these elements can greatly influence our overall wellbeing^[114]14. In contrast, harmful life habits or specific events can induce stress, potentially contributing to the development of dementia, AD and PD^[115]10,[116]15–[117]17. Additionally, post-traumatic stress disorders have been identified as risk factors for the onset of PD^[118]10,[119]18,[120]19 and frontotemporal dementia^[121]20. However, the precise link between a stressful lifestyle and ALS pathogenesis remains unclear. An international online case-control study found no association between stressors and ALS^[122]21, while a Japanese case-control study suggested that a combination of different lifestyle factors, along with reduced antioxidant defences in MNs, may increase ALS risk^[123]22. In our study, we aimed to clarify the potential relationship between stress and ALS pathogenesis, by exploring the molecular mechanisms triggered by stress in ALS-predisposed conditions, using both in vivo (SOD1^G93A mice) and in vitro (hSOD1^G93A NSC-34 and TDP43-mutated hiPSC-derived MNs) models. First, we employed the most largely known ALS mouse model, the SOD1^G93A mice, to conduct a pilot study assessing whether stress exposure could affect the disease onset and progression (Fig. [124]1; Fig. [125]S1). Due to the well-recognized gender ALS-related differences, we considered females and males separately. We mimicked a stressful lifestyle by applying the chronic unpredictable mild stress protocol, a well-established rodent model used to induce depressive- and anxiety-like behaviours^[126]11,[127]12. Indeed a stressful lifestyle can increase the risk of developing anxiety and/or depression, affecting, among the others, mitochondrial membrane depolarization, increased ROS levels and microglial activation^[128]23–[129]28. Interestingly, we observed that the chronic unpredicted mild stress protocol influenced the ALS onset in mice, in particular by affecting the body weight gain (in both females and males) and determining early motor defects (in females). It is worth noting that a lower Body Mass Index at the time of diagnosis has been associated with a worse prognosis in ALS^[130]29. Therefore, detecting a stress-related effect on the body weight at the symptom onset may be relevant in the context of this study. Moreover, in the female hSOD1^G93A stressed group, among 39 ALS-related genes, Col1a1 was significantly up-regulated in both the motor cortex and spinal cord, compared to not stressed transgenic mice (CTRL). Furthermore, Col1a2 was also up-regulated in motor cortex of female stressed group. Otherwise, we observed a significant up-regulation of Il6 in the motor cortex of hSOD1^G93A stressed male mice. Il6 is a crucial cytokine in CNS, with a largely recognized role in several biological processes, including neuroinflammation^[131]30, while collagen is an important protein which can provide structural stability of the extracellular matrix^[132]31. Interestingly, increased levels of Il6^[133]30,[134]32 and collagen accumulation^[135]33 have been referred in ALS patients. Indeed, in our experiments, the altered expression of Il6, Col1a1 and Col1a2 could be related to oxidative stress, neuroinflammation, DNA damage, apoptosis, autophagy and fibrosis processed, possibly exacerbated by stress^[136]34–[137]37, with evident effects in ALS-predisposed conditions, especially in females. Sex-based divergence is likely attributable to differences in stress response, particularly in terms of reactivity, resilience and adaptation^[138]38–[139]40. In fact, our results are consistent with previous works showing higher vulnerability of female mice to stressful conditions compared to males^[140]41. To specifically investigate the MN involvement, we moved to an in vitro experimental model of MNs widely used in ALS research: the NSC-34 cell line. By introducing slight changes to the available protocols for differentiating NSC-34 in MN-like cells^[141]13,[142]42,[143]43, we standardized a simple method to promote MN maturation in both naïve and hSOD1-transfected cells (Fig. [144]S2, Fig. S3, Fig. S4 and Fig. [145]2), enhancing cell-body clusters, neurite length, neurite branch points and ChAT expression. In the in vitro stress model, we selected OGD to mimic the in vivo stress effects, specifically targeting key stress-activated molecular pathways and associated cellular outcomes. Indeed, OGD can induce neurotoxicity through ROS accumulation and is associated with mitochondrial impairment, neuroinflammation and neuronal death^[146]44–[147]47. These cellular dysfunctions are related both to ALS pathogenesis and chronic stress^[148]8. Therefore, we consider OGD (combining low-glucose medium with CoCl[2]) as an appropriate model for mimicking chronic stress^[149]48,[150]49,. Our results highlighted a reduced ability of hSOD1^G93A NSC34 to endure stress (Fig. [151]3). In hSOD1^WT cells, upon stress exposure, the mitochondrial membrane potential slightly decreased, while the expression of hypoxic (HIF1α) and apoptotic (cleaved caspase 3) markers increased, suggesting a physiological stress response^[152]49–[153]54. In contrast, OGD triggered a significant decrease of hSOD1^G93A mitochondrial membrane potential, without affecting HIF1α and cleaved caspase 3 protein levels. Elevated levels of cleaved caspase 3 have been previously reported in hSOD1^G93A NSC-34 cells under basal conditions^[154]55. These observations are in line with our findings, suggesting that the apoptotic pathways are already activated in mutated cells, even in absence of external stress. Thus, we hypothesize that additional stress exposure cannot further amplify this response that may have already reached a plateau due to the pre-existing disease-related activation of apoptosis. In light of these findings, we investigated the same set of genes in NSC-34 hSOD1 MNs, normalizing hSOD1^WT STRESSED, hSOD1^G93A CTRL and STRESSED, to hSOD1^WT CTRL MNs (Fig. [155]4). Then, to better discriminate the mutation- or stress-related effects on gene expression, the significantly deregulated genes were further analysed independently, for both CTRL (i.e. not stressed) and STRESSED groups. The gene expression of mutated MNs has been normalized to WT MNs showing several genes significantly deregulated in CTRL condition (Cldn5 down-regulated; Bdnf, Casp1, Col4a1, Fgfr1, Gsk3b, Hdac7, Igf1, Il6, Nanog and Tgfb1 up-regulated). Similarly, the expression of some genes was remarkably affected in hSOD1^G93A cells compared to hSOD1^WT upon stress exposure (Ang, Cldn5, Gsk3b, Hdac7, Hspb1, Il6 down-regulated; Casp1, Col4a2, Igf1 and Tgfb1 up-regulated). Therefore, stress interferes with some specific intracellular mechanisms, also affecting the predicted PPIs (Fig. [156]4). Interestingly, the involvement of certain genes (Gsk3b, Il6, Igf1, together with collagen) seems to be more crucial than others and, based on the GO and pathway enrichment analyses, converges to the key role of PI3K/Akt pathway in the response to stress (Fig. [157]5). Notably, although PI3K/Akt pathway appeared to be implicated in both experimental conditions, the expression of certain genes (as Gsk3b and Il6) was different between CTRL and STRESSED groups, highlighting stress-dependent differential regulatory mechanisms in ALS-predisposed conditions. The up-regulation of Gsk3b found in the CTRL mutated group aligns with well-known mechanisms of neurodegenerative diseases, including ALS^[158]56–[159]58, in which the Akt inhibition, leading to apoptotic processes and determining caspase 3 protein level increment, may support our results observed in Fig. [160]3. On the contrary, the down-regulation of Gsk3b observed in the STRESSED mutated group can reflect Akt activation, potentially counteracting the apoptotic processes. Concerning Il6, it is known to trigger neuroinflammation in ALS and, although it is a cytokine mainly produced by immune cells, also neurons can express it^[161]30,[162]32,[163]59: in our experiments, stress apparently modifies its expression, potentially representing an additional protective response. On the other hand, high levels of Igf1 can be correlated with a better prognosis of ALS^[164]60, by exerting neuroprotective effects. Based on our results, their levels are strongly reduced upon stress exposure, potentially indicating neuronal distress. Overall, these findings suggest that in the hSOD1^G93A MNs stress can trigger different pathways that on one hand could be detrimental, and on the other might represent protective endogenous mechanisms (likely insufficient to halt disease progression, as evidenced by persistently elevated caspase expression). Pathway enrichment analysis revealed another interesting result concerning the specific pathway activated only in STRESSED experimental condition, i.e. the focal adhesion pathway, with the involvement of Gsk3b, Igf1, and Col4a2 (Fig. [165]5). Notably, focal adhesion, often by the mediation of Gsk3b and Igf1, regulates oxidative stress induced by chronic stress, depressive symptoms, diet, and general lifestyle factors^[166]61–[167]66. Moreover, collagen gene family has been previously characterized in brain regions of male mice underwent chronic agonistic interactions^[168]67, supporting our in vivo results for which COL genes could be key players of stress regulatory mechanisms in ALS-predisposed conditions. Interestingly, the results obtained from hSOD1^G93A NSC-34 cells were consistent with those from hiPCS-derived MNs (Fig. [169]6), despite the different mutated gene (i.e., TARDBP). The morphological parameters of healthy cells (healthy line) indicated a physiological stress response affecting the neurite length and branch point number^[170]68–[171]71. Moreover, compared to healthy cells, the mutated ones showed significant morphological alterations already in CTRL conditions, that were not further worsened by OGD stress. However, OGD induced significant transcriptional changes in human mutated MNs. Interestingly, gene expression in hiPCS-derived MNs aligned with the results observed in NSC-34 hSOD1 cell experiments. For instance, Gsk3b was down-regulated, and Col4a2 was up-regulated in STRESSED condition. In conclusion, our in vivo results (although representing a pilot study, intended to lay the groundwork for our working hypothesis) highlight evident negative effects of stress on the ALS onset/progression. Future investigations at later symptomatic stages and with increased sample size could provide further insights into the impact of stress on disease progression in vivo. Moreover, the in vitro experiments on MNs suggest a crucial role of the PI3K/Akt pathway (triggered by Gsk3b, Il6, Igf1 and collagen) in mediating stress response in different ALS-predisposed conditions. The investigation of Gsk3b/GSK3B phosphorylation and the consequent signalling cascade in both PI3K/Akt and focal adhesion pathways could clarify its role in the modulation of Igf1/IGF1, Il6/IL6, and Col/COL proteins. Of course, the absence of protein-level validation and rescue experiments represents a limitation of this study. However, the transcriptomic convergence observed across the used experimental models provides a solid basis for the identification of stress-responsive molecular pathways in ALS-predisposed conditions. Future work will be necessary to validate key candidate genes at the protein level and to functionally assess their contribution to the disease-related phenotype. Therefore, although further investigations are needed to support our hypothesis, this study enhances our comprehension of the stress role in the ALS-related neurodegenerative mechanisms, paving the way for novel therapeutic targets and encouraging different lifestyle habits. Methods Animal care and use The experimental in vivo procedures were performed in strict accordance with institutional guidelines in compliance with national (D.L. N.26, 04/03/2014) and international law and policies (new directive 2010/63/EU). The study was approved by the Italian Ministry of Health (permit n. 17/2010-B, 30 June 2010, and protocol n. 391/2021-PR). Additionally, an ad hoc Ethical Committee of the University of Turin approved this study. The study is reported in accordance with ARRIVE guidelines. Mice were maintained in the cages, under standard conditions with 12/12-h light/dark cycle and free access to food and water. Male and female B6SJL-Tg (SOD1*G93A)1Gur/J mice (The Jackson Laboratory; Bar Harbor, Maine, USA; ref 002726) were used to evaluate in vivo the stress effects. These are transgenic animals expressing a human SOD1 (hSOD1) containing the Gly93Ala mutation, which determines a progressive upper and lower MN loss. The colony was maintained by breeding hemizygous carrier males and B6SJLF1/J females. The offspring genotype was identified by DNA extracted from the mouse tail in order to identify the presence of the hSOD1 transgene. Forward (oIMR0113; 5’ – CAT CAG CCC TAA TCC ATC TGA – 3’) and reverse primers (oIMR0114; 3’ – CGC GAC TAA CAA TCA AAG TGA – 5’) were used, following the Jackson Laboratory instructions and Taq DNA polymerase (GoTaq Flexi DNA Polymerase; Promega; Madison, Wisconsin, USA; ref 9PIM829) protocols for the polymerase chain reaction (PCR). Overall, 6 transgenic females and 6 transgenic males were used for the experiments: 3 per group were housed in standard conditions (12 h darkness/12 h light), while the others underwent a chronic unpredicted mild stress protocol according to Kumar et al.^[172]11,[173]12, starting at postnatal day 60 (P60). Briefly, the overall duration of the protocol was 28 days, during which cage shaking, cage tilt, cold and warm swim, food/water deprivation, tail pinch, moist bedding, overnight illumination, and a reverse day-night cycle were repeated at least three, non-consecutive times^[174]11,[175]12. At P90, mice were euthanized by cervical dislocation in order to collect lumbar spinal cord and motor cortex from all the experimental groups. Rotarod test and weight assessment Rotarod test was performed to detect motor dysfunctions of the transgenic mice, either housed in standard conditions (CTRL) or stressed. The test was performed from P45 to P89, twice a week, using a 7650-accelerating model of a rotarod apparatus (Ugo Basile, Italy). The first two weeks were considered training sessions. Each test was repeated for three trials, with an accelerated speed from 5 to 32 rpm, and an arbitrary cut-off time of 300 s. The average of three trials was considered. In addition, the mouse weight was assessed after each rotarod session. The values at P89 (end point) were normalized to that at P60 (starting point). NSC-34 cell cultures The murine MN-like cell lines, NSC-34 naïve, hSOD1^WT and hSOD1^G93A were kindly provided by Prof. Mariotti’s laboratory (University of Verona, Italy). hSOD1^WT and hSOD1^G93A cells were stably transfected with the human wild type SOD1 gene or with the mutated one, in previous works^[176]72,[177]73. The human gene activation is regulated by a doxycycline inducible promoter: 5 µg/ml of doxycycline (Merck, Sigma-Aldrich; St. Louis, MO, USA; ref D9891) for 24 h were used to obtained the human gene expression. Cells were grown in standard experimental condition using DMEM (Merck, Sigma-Aldrich; ref D5796) supplemented with 10% of heat-inactivated fetal bovine serum (FBS; Merck, Sigma-Aldrich; ref F7524), 100 U/ml penicillin and 100 µg/ml streptomycin (Merck, Sigma-Aldrich; ref P4333). Cell differentiation was induced reducing FBS concentration (from 10 to 1%) and adding 20 µM of RA (Merck, Sigma-Aldrich; ref R2625) for 4 days of culture. The culture medium was changed every two days for both standard and differentiation culture conditions. The OGD stress was induced at DIV (days in vitro) 4 for 24 h using DMEM low glucose (Merck, Sigma-Aldrich; ref D5921), supplemented with 100 µM of CoCl[2] (Merck, Sigma-Aldrich; ref 232696). Cells, both under standard and stress conditions, were incubated at 37 °C in 5% CO2. Human iPSC-derived MN cell cultures Human MN progenitors were obtained using two lines of human iPSCs, one derived from a healthy patient (56c2) and the other one derived from an ALS patient with the human TARDBP^G298S mutation, following a standardized protocol^[178]74,[179]75, that has been patented and licensed to Stem Cell Technology (Patent PCT/IB2020/000972). Briefly, the cells were differentiated in embryonic bodies that were dissociated at day 10. The resulting MN progenitor cells were stored into biobanks of I-Stem (Corbeil-Essonnes, France) and Institut Imagine (Paris, France). For the stress experiment, those MN progenitors were seeded in 6 well plates coated with poly-L-ornithine solution (Merck, Sigma-Aldrich; ref P4957) before laminin (Thermo Fisher, Gibco; Waltham, Massachusetts; ref USA23017015), at ~ 1 × 10^5 cells/cm^2 of density. The progenitors were seeded and further differentiated intro mature MNs for 14 days in medium consisting to DMEM-F12 Glutamax/Neurobasal (Thermo Fisher, Gibco; ref 10565018/21103049; 1:1 ratio), N2 supplement/B27 no vitamin A supplement (Thermo Fisher, Gibco; ref 17502048/A3353501; 1:2 ratio), 0.1% of β-mercaptoethanol (Thermo Fisher, Gibco; ref 31350010) and 0.1% of Pen Strep (Thermo Fisher, Gibco; ref 15140122), supplemented with 100 nM of retinoic acid (Merck, Sigma-Aldrich; ref R2625), 0.5 µM of Smoothened Agonis (SAG; STEMCELLS Technologies; Vancouver, British Columbia, Canada; ref 73414;), 1 µM of Brain-Derived Neurotrophic Factor (BDNF; PreproTech; Rocky Hill, New Jersey, USA; ref 17874463), 1 µM of Glial-Derived Neurotrophic Factor (GDNF; PreproTech; ref 17814073), 10 nM of γ-secretase inhibitor (DAPT, STEMCELLS Technologies; ref 72792) and 10 µM of Y-27,632 (ROCK Inhibitor; STEMCELLS Technologies; ref 72307). Y-27,632 was eliminated the day after seeding at by medium refreshing and medium was refreshed every three days. The OGD stress was induced at day 24 using DMEM low glucose, supplemented with 100 µM of CoCl[2] (Merck, Sigma-Aldrich; ref 232696). Cells were incubated at 37 °C in 5% CO[2] during the entire period of the experiments. Cell viability assay 3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyl tetrazolium bromide) assay (MTT assay) for RA and CoCl[2] treatments on NSC-34 cells was performed using cell proliferation kit I (MTT) (Merck, Roche; Basel, Switzerland; ref 11465007001) following the manufacture’s instruction. Cells were plated in 96-wells plates (3 × 10^3 cells/well) and maintained in standard growth culture conditions for 24 h before testing different concentrations of RA (1 µM, 5 µM, 10 µM, 15 µM and 20 µM) for 2, 4, 6 and 8 days, with and without doxycycline treatment. In order to perform CoCl[2] MTT, cells were also seeded in 96-wells plates (3 × 10^3 cells/well) and were kept in standard growth culture for 24 h. Cells were differentiated for 4 days: then, DMEM low glucose supplemented with doxycycline and different concentrations of CoCl[2] (50 µM, 100 µM, 200 µM, 300 µM and 400 µM) were tested for 24 h. The absorbance was measured at 550 nm using a microplate reader (Tecan Infinite M nano) in at least two wells per condition, replicating three independent experiments. Neurotrack analysis Incucyte Neurotrack software module was used to validate MN differentiation of NSC-34 naïve, hSOD1^WT and hSOD1^G93A cells, evaluating cell-body cluster, neurite length and neurite branch point. The same morphological parameters were measured for analysing the stress impact on human MNs (hMNs). NSC-34 cells were seeded in 96-well plates at the density of 1.5 × 10^4 cells/cm^2 for four experimental conditions: standard medium with 1% fetal bovine serum (FBS), with and without doxycycline, and differentiated condition (standard medium with 1% serum, supplemented with 20 µM of RA), with and without doxycycline. Instead, hMNs (both healthy and TDP-43 mutated cells) were plated in 6-well plates at 1 × 10^5 cells/cm^2 for two experimental conditions: control and stressed conditions. Five and two wells per conditions of NSC-34 cells and hMNs, respectively, were analysed using three independent experiments [each performed at different passages (p) of the same NSC-34 cell line, specifically between p18 and p24], after the acquisition by the built-in Basler Ace 1920 –155 μm camera of the Incucyte system (Sartorius; Göttingen, Germany). NSC-34 segmentation mode related to cell-body cluster was set up in brightness considering 1 as range of segmentation adjustment. Clean-up was regulated for min cell width, set to 7 μm, while cell-body cluster filters were not applied. Filtering related to neurite parameters was configured as better setting 0.3 for neurite sensitivity and 1 μm for neurite width. For hMNs, filtering of neurite parameters was configured ad best, while the neurite sensitivity and the neurite width were set-up to 0.35 and 2, respectively. Immunofluorescence To analyse cell differentiation in vitro, murine cells were seeded in 24 wells plates onto poly-D-lysine (Merck, Sigma-Aldrich; ref A-003-E) coated-coverslips at the density of 1.5 × 10^4 cells/cm^2 under four experimental conditions: standard medium 1% serum, with and without doxycycline, and differentiated condition, with and without doxycycline. Cells were fixed in 4% paraformaldehyde for 10 min at DIV5. Cell permeabilization was performed in PBS-0.2% Triton-X100, while blocking of unspecific sites was done in PBS-0.1% Triton-X100 and 5% Normal Donkey Serum (NDS; Merck, Sigma-Aldrich; ref. D9663). As primary antibody, we used rabbit Anti-Choline Acetyltransferase (ChAT) (Merck, Sigma-Aldrich; ref Ab143; 1:200), incubated in PBS-0.1% Triton-X100 and 5% NDS. As secondary antibody, we used Cy3 AffiniPure Donkey Anti-Rabbit IgG (H + L) (Jackson ImmunoResearch; West Grove, Pennsylvania, USA; ref 711-165-152; 1:200) incubated in PBS and 2% NDS. Coverslips were mounted on microscope slides using mounting medium Mowiol/Dabco and the image acquisitions were realized by Eclipse E600 (Microfire Camera 2-Megapixel Color Imaging, 1600 × 1200). Images were thresholded identically across all experimental conditions and CTCF was analysed for at least ten cells from three independent experiments using ImageJ software. Mitochondrial imaging Mitochondrial membrane potential was evaluated using MitoTracker Red CMXRos (Thermo Fisher; ref M7512), a mitochondrial dye based on X-rosamine derivate. The probe was added to the medium of NSC-34 hSOD1^WT and hSOD1^G93A cells under two experimental conditions: control and stressed conditions. The final concentration of the probe was 100 nM and was maintained for 25 min, keeping the cell plates in the incubator at 37 °C in 5% CO[2]. Then, medium was changed and the image acquisitions were performed by the built-in Basler Ace 1920 –155 μm camera of the Incucyte system. “Cell-by-cell” software module was used in order to obtain the Total Orange Object Integrated Intensity normalized to Phase Object Count. All the analysis parameters were configured following Sartorius recommendations. Western blot Western blot was performed in order to verify the stress impact on NSC-34 hSOD1^WT and hSOD1^G93A cells. Proteins were extracted using a lysis buffer consisting of RIPA buffer (Merck, Millipore; Darmstadt, Germany; ref 20–188) supplemented with cOmplete Protein Inhibitor Cocktail [1X] (Merck, Roche; ref 11697498001), sodium orthovanadate [2 mM] (Merck, Sigma-Aldrich; ref S6508), Phenylmethanesulfonyl fluoride [1 mM] (Thermo Fisher; ref 36978) and dithiothreitol [1 mM] (Thermo Fisher; ref R0861). Lysates were sonicated for 8 min using Ultrasonic baths Bandelin Sonorex at room temperature and centrifuged at 3 × 10^3 rpm for 10 min at 4 °C. Protein quantifications were performed following manufacturer’s instructions of Bio-Rad Protein Assay Dye Reagent Concentrate (Bio-Rad Laboratories; Hercules, California, USA; ref 5000006). 25 µg of proteins were mixed to NuPAGE LDS Sample Buffer [1X] (Thermo Fisher; ref NP0007) and NuPAGE Sample Reducing Agent [1X] (Thermo Fisher, Invistrogen; ref NP0004), denatured for 5 min at 95 °C, separated using 4–20% Mini-PROTEAN TGX Precast Protein Gels (Bio-Rad Laboratories; ref 4561094;) and transferred into nitrocellulose membrane using Trans-Blot Turbo RTA Mini 0.2 μm Nitrocellulose Transfer Kit (Bio-Rad Laboratories; ref 1704270). Blots were blocked with EveryBlot Blocking Buffer (Bio-Rad Laboratories; ref 12010020) and incubated overnight with selected primary antibodies: mouse anti HIF1α (Santa Cruz; Dallas, Texas, USA; ref 28b; sc-13515; 1:500), rabbit anti Cleaved Caspase-3 (Cell signalling; Danvers, Massachusetts, USA; ref Asp 175; 9661; 1:500) and mouse anti β-Tubulin (Merck, Sigma-Aldrich; ref T8328; 1:1,000). Subsequently, membranes were incubated for 1 h with appropriate secondary antibodies: Goat Anti-Mouse IgG (H + L)-HRP Conjugate (Bio-Rad Laboratories; ref 1706516; 1: 10,000) and Goat Anti-Rabbit IgG (H + L)-HRP Conjugate (Bio-Rad Laboratories; ref 1706515; 1:10,000). The bands were exposed to Clarity Max Western ECL Substrate (Bio-Rad Laboratories; ref 1705062) and their signals were detected using ChemiDoc imaging system (Bio-Rad Laboratories; USA). Densitometric analysis of the bands was performed using Bio-Rad Laboratories Image Lab software and target values were normalized to β-Tubulin signal. Quantitative reverse transcription polymerase chain reaction (RT-qPCR) Total RNA extraction from murine tissues and NSC-34 cell lines was performed using Maxwell RSC simply RNA Cells Kit (Promega; ref AS1390), while for hMNs the RNeasy plus mini kit was used (QIAGEN; Venlo, Netherlands; ref 74134). RNA concentrations were quantified using Nano Quant Plate available for microplate reader Tecan Infinite M nano (Tecan, Switzerland) and iScript Advanced cDNA Synthesis Kit (Bio-Rad Laboratories; ref 1725038) was used for RNA to cDNA retrotranscription. For qPCR, species-specific (mouse or human) Prime PCR Disease State Panels (Bio-Rad Laboratories; ref 10036315 or 10034810) were used, based on SYBR detection workflows (iTaq Universal SYBR Green Supermix; Bio-Rad Laboratories; ref 1725124) and predesigned for ALS-related genes (Table [180]S1). cDNA concentration was calibrated for each experimental model: charged-quantity of cDNA for mouse tissues and human MNs was 25 ng/µl, while for NSC-34 cell lines was 75 ng/µl. The results were analysed following 2^–∆∆Ct method considering Gapdh gene as housekeeping. Heatmaps, violin blots and bar plots were used to represent gene expression analyses. Protein-protein interaction (PPI) data Protein-protein interaction (PPI) related to deregulated genes in hSOD1^G93A NSC-34 cells compared to hSOD1^WT ones was analysed in separated analysis for control and stressed experimental condition using The Search Tool for the Retrieval of Interacting Genes (STRING) database (free online platform: [181]https://string-db.org, accessed on 8 May 2024 and 19 March 2024, respectively). GO and pathway enrichment analyses GO and pathway enrichment analyses were performed on genes significantly deregulated in hSOD1^G93A NSC-34 cells compared to hSOD1^WT ones. Separated analyses were applied for control and stressed experimental condition using SRplot (free online platform: [182]http://www.bioinformatics.com.cn/SRplot, accessed on 8 May 2024 and 15 March 2024, respectively). GO three ontology plots were used to show biological processes (BP), cellular component (CC) and molecular function (MF), highlighting enrichment score in the x-axis of the graph. Instead, pathway enrichment analysis was visualized through cnet plots to facilitate the interpretation of gene interactions and their involvement in a specific pathway (or category), by using log2 fold change about the gene expression and bubble size for the gene’s numbers implicated in a specific category. Statistical analysis Data are shown as media ± SD and/or as median and quartile range of at least three independent experiments. Statistical analyses were conducted using two-way ANOVA (followed by Dunnett or Sidák or Tukey multiple comparison or by Uncorrected Fisher’s LDS) or Mixed-effects (followed by Uncorrected Fisher’s LDS or Tukey’s multiple comparison tests) or Unpaired t-test. These analyses and corresponding graphs (including heatmaps) were performed using GraphPad Prism version 10.1.2 for Windows (GraphPad Software, Boston, Massachusetts USA, [183]www.graphpad.com). Results were considered significant at P < 0.05. Electronic supplementary material Below is the link to the electronic supplementary material. [184]Supplementary Material 1^ (3.4MB, docx) [185]Supplementary Material 2^ (18.1KB, docx) Acknowledgements