CXCR2 Is Deregulated in ALS Spinal Cord and Its Activation Triggers Apoptosis in Motor Neuron-Like Cells Overexpressing hSOD1-G93A

Amyotrophic lateral sclerosis (ALS) is a multifactorial neurodegenerative disease characterized by progressive depletion of motor neurons (MNs). Recent evidence suggests a role in ALS pathology for the C-X-C motif chemokine receptor 2 (CXCR2), whose expression was found increased at both mRNA and protein level in cortical neurons of sporadic ALS patients. Previous findings also showed that the receptor inhibition is able to prevent iPSC-derived MNs degeneration in vitro and improve neuromuscular function in SOD1-G93A mice. Here, by performing transcriptional analysis and immunofluorescence studies, we detailed the increased expression and localization of CXCR2 and its main ligand CXCL8 in the human lumbar spinal cord of sporadic ALS patients. We further investigated the functional role of CXCR2/ligands axis in NSC-34 motor neuron-like cells expressing human wild-type (WT) or mutant (G93A) SOD1. A significant expression of CXCR2 was found in doxycycline-induced G93A-SOD1-expressing cells, but not in WT cells. In vitro assays showed CXCR2 activation by GROα and MIP2α, two murine endogenous ligands and functional homologs of CXCL8, reduces cellular viability and triggers apoptosis in a dose dependent manner, while treatment with reparixin, a non-competitive allosteric CXCR2 inhibitor, effectively counteracts GROα and MIP2α toxicity, significantly inhibiting the chemokine-induced cell death. Altogether, data further support a role of CXCR2 axis in ALS etiopathogenesis and confirm its pharmacological modulation as a candidate therapeutic strategy.


Introduction
Amyotrophic lateral sclerosis (ALS) is a debilitating condition characterized by the degeneration of motor neurons (MNs) in the primary motor cortex and spinal cord, resulting in progressive muscle weakness and death within 2-5 years [1,2]. Most of the cases (90%) are sporadic (SALS) without a family history, while the remaining 10% are familial (familial ALS, FALS) mainly inherited in a dominant manner [1,3]. Disease-causing mutations in the Cu/Zn superoxide dismutase type-1 (SOD1) gene are common in ALS and account for both FALS and SALS, explaining approximately 12-20% of the familial and 1-2% of the sporadic cases [4,5]. The clinical presentation of SALS and FALS are similar, and treatment options remain mainly supportive so far. Indeed, the two current FDA-approved drugs, i.e., the anti-excitotoxic Riluzole (Rilutek) and the antioxidant Edaravone are able to prolong the lifespan of patients by only a few months and counteract disease progression without a real resolutive outcome [6,7].
The pathogenic process underlying ALS neurodegeneration is still not fully determined, although described alterations primarily consist of aberrant RNA metabolism, neuroinflammation, impaired protein homeostasis, mitochondrial dysfunction, excitotoxicity, and oxidative stress [8]. We have recently provided novel interesting clues about the role of the G-protein-coupled C-X-C motif chemokine receptor 2 (CXCR2) in ALS pathophysiology [9]. By using a bulk transcriptomic-based patients stratification approach and a following inter-species meta-analysis, we identified CXCR2 mRNA as significant deregulated in human SALS motor cortex and SOD1-G93A mice at symptomatic stages as well, and therefore prioritized it as a candidate therapeutic target [8][9][10][11][12][13]. We further observed an increased immunoreactivity of the CXCR2 receptor in neuronal cell bodies and axons from ALS motor cortex [9], and functionally showed that receptor inhibition prevent inducible pluripotent stem cells (iPSC)-derived MNs degeneration in vitro and improve SOD1-G93A mice muscular functions in vivo, delaying the onset of neuromuscular decline by four weeks [9].
Deregulation of the CXCR2/ligands signaling has been previously described for further neuropathological diseases (traumatic brain injury, multiple sclerosis, ischemia, Alzheimer's disease, neuropathic pain) [14,27], but the biological significance of this alteration in ALS remains not fully explained.
In the present work, we investigated the expression and localization of both CXCR2 and CXCL8 in spinal cord specimens from control and ALS patients, and inspected the biological and functional role of the CXCR2/ligands axis in murine NSC-34 motor neuronlike cells expressing human wild-type (WT) or mutant G93A-SOD1.

Transcriptomic Profiling
For this study, we refer to a previously described transcriptome dataset [11,28] deposited in ArrayExpress (http://www.ebi.ac.uk/arrayexpress/, accessed on 30 January 2022) with the accession number E-MTAB-8635 (https://www.ebi.ac.uk/arrayexpress/ experiments/E-MTAB-8635/, accessed on 30 January 2022). The dataset consists of the expression profiles of spinal cord samples from SALS (n = 30) and control (n = 10) subjects obtained with 4 × 44 K Whole Human Genome Oligo expression microarrays containing 41,093 probes (Agilent Technologies, Santa Clara, CA, USA). A detailed description of the subject characteristics (origin, source code, age, gender, race, disease state, survival time from diagnosis date and post-mortem interval) and experimental procedures was previously reported [11,28]. The E-MTAB-8635 dataset has been previously queried to investigate the role of splicing players deregulation in sporadic ALS [28], but not to explore the CXCR2/CXCL8 axis. Raw intensity signals from samples hybridization were thresholded to 1, log2-transformed, normalized, and baselined to the median of all samples using GeneSpring GX (Agilent Technologies, Santa Clara, CA, USA). Values from multiple probe signals targeting the same gene were collapsed to create a gene-level analysis and filtered to focus on CXCR2/CXCL8. The statistical analysis between CTRL and SALS was performed using t-test followed by Tukey post hoc test to identify significant variation between groups.

Cell Viability
Cell viability was assessed using the colorimetric reagent-based MTT cell proliferation kit I, based on 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (Roche Diagnostics, Mannheim, Germany) salt, as previously described [32]. Cells were cultured into 96-well plates at a density of 1 × 10 4 cells/well in 100 µL of medium for 24 h. The day after doxycycline induction, cells were treated with increasing concentration of CXCL1/GROα (SRP4210, Sigma-Aldrich, Munich, Germany) and CXCL2/MIP2α (SRP4251, Sigma-Aldrich, Munich, Germany) for 24 h. Subsequently, 0.5 mg/mL of MTT was added to each well and incubated for 4 h at 37 • C. The reaction was stopped with 100 µL of solubilization solution, then formazan was measured spectrophotometrically (550-600 nm) using a microplate reader (Bio-Rad Laboratories, Hercules, CA, US). Six replicate wells were used for each group. Controls included untreated cells, and medium alone was used as a blank. The minimum effective dose of agonists in vitro to produce significant cell death, i.e., GROα (1 ng/mL) and MIP2α (100 nM), was chosen and used for following experiments alone or in combination with 10 µM reparixin (MedChem Express, Monmouth Junction, NJ, USA) as previously described [9].
To perform cell counting, NSC-34 WT and NSC-34 G93A cells were seeded into 24 well plates (25,000 cells for each well) and cultured for 24 h at 37 • C in 5% CO 2 in the medium previously described. The day after, hSOD1 expression was induced by adding 2 µg/mL doxycycline for 24 h. Then, the culture medium was replaced with fresh medium containing GROα (1 ng/mL) or MIP2α (100 nM) in the presence or absence of reparixin (Rep) (10 µM). After 24 h, cells were trypsinized and centrifugated at 10,000× g for 7 min at 4 • C. The pellet was resuspended in 1 mL of fresh medium, then 100 µL of cell suspension were added to a trypan blue solution for cell counting in a Bürker chamber.

Western Blot Analysis
Proteins were extracted from total cells lysate with RIPA buffer (Thermo Fisher Scientific, Paisley, UK) supplemented with phosphatase and protease inhibitors (Roche Diagnostics, Monza, Italy), homogenized by a Teflon-glass homogenizer and then sonicated by an ultrasonic probe, followed by centrifugation at 10,000× g for 10 min at 4 • C. Quant-iT Protein Assay Kit (Invitrogen, NY, USA) was used to determine protein concentration for each sample as previously described [33]. About 25 µg of protein homogenate was diluted in 2X Laemmli buffer (Invitrogen, NY, USA), heated at 70 • C for 10 min, separated on a Biorad Criterion XT 4-15% Bis-tris gel (Invitrogen, NY, USA) by electrophoresis and then transferred to a nitrocellulose membrane (Invitrogen). The transfer was monitored by a pre-stained protein molecular weight marker (BioRad Laboratories, Hercules, CA, USA). The nitrocellulose membranes were firstly incubated with Odyssey Blocking Buffer (Li-Cor Biosciences, Lincoln, Nebraska) and subsequently with specific primary antibodies: anti-CXCR2 (rabbit ab217314, Abcam, Cambridge, UK), anti-BAX (mouse sc-20067, Santa Cruz Biotechnology, Inc., Dallas, TX, USA), anti-BCL2 (mouse sc-509, Santa Cruz Biotechnology, Inc., Dallas, Texas), and anti-β-actin (mouse sc-47778, 1:500, Santa Cruz Biotechnology, Inc., Dallas, Texas).

Immunocytofluorescence
NSC-34 cells expressing human WT or SOD1-G93A were cultured on glass cover slips, fixed in 4% PFA in PBS for 15 min at room temperature, permeabilized with Triton X-100 (0.2%), blocked with BSA (0.1%) in PBS for 1 h at room temperature and probed with the anti-CXCR2 (rabbit ab14935, Abcam, Cambridge, UK) or anti-cleaved-caspase-3 Asp 175 (rabbit #9661, Cell Signaling) primary antibodies overnight. Subsequently, samples were incubated with the Alexa Fluor 488 goat anti-rabbit antibody for 1 h at room temperature shielded from light. DAPI was used to stain the nuclei (#940110 Vector Laboratories). Images were captured with a Nikon A1 confocal inverted microscope equipped with a Plan Apochromat lambda 60×/1.4 oil immersion lens (Nikon, Tokyo, Japan). Fluorescence was quantified by extrapolating the mean intensity of FITC channel from multiple regions of interest (ROI), normalized to the background by using the NIS-Elements AR (Advanced Research) software (version 4.60).

Statistical Analysis
Data are reported as the mean ± standard error of the mean (SEM). T-test and oneway analysis of variance (ANOVA) were used to compare differences among groups. Tukey-Kramer post hoc test was applied to assess the statistical significance (p ≤ 0.05). All statistics were run using the Prism 5.0a (GraphPad Software Inc., La Jolla, CA, USA) software packages.

CXCR2/CXCL8 Expression in Control and Sporadic ALS Spinal Cord Samples
Previous transcriptome profiling, qRT-PCR and immunohistochemistry experiments revealed a significant upregulation of CXCR2 in human sporadic ALS motor cortex compared to control, and its main localization in both somas and axons of cortical neurons [9,11]. Here, we focused on lumbar (L1) spinal cord specimens from the same cohort of patients, and analyzed CXCR2 and its main ligand CXCL8 at both mRNA and protein levels.
Gene-level transcriptional analysis (E-MTAB-8635 dataset) showed a significant increase of both CXCR2 and CXCL8 mRNA in ALS spinal cord tissue compared to controls (* p < 0.05 ALS vs. CTRL) ( Figure 1). In addition, we found an association trend between CXCR2 higher mRNA levels and short survival (Kaplan-Meier curves in Supplementary Figure S1).
Tukey-Kramer post hoc test was applied to assess the statistical significance (p ≤ 0.05). All statistics were run using the Prism 5.0a (GraphPad Software Inc., La Jolla, CA, USA) software packages.

CXCR2/CXCL8 Expression in Control and Sporadic ALS Spinal Cord Samples
Previous transcriptome profiling, qRT-PCR and immunohistochemistry experiments revealed a significant upregulation of CXCR2 in human sporadic ALS motor cortex compared to control, and its main localization in both somas and axons of cortical neurons [9,11]. Here, we focused on lumbar (L1) spinal cord specimens from the same cohort of patients, and analyzed CXCR2 and its main ligand CXCL8 at both mRNA and protein levels.
Gene-level transcriptional analysis (E-MTAB-8635 dataset) showed a significant increase of both CXCR2 and CXCL8 mRNA in ALS spinal cord tissue compared to controls (* p < 0.05 ALS vs. CTRL) (Figure 1). In addition, we found an association trend between CXCR2 higher mRNA levels and short survival (Kaplan-Meier curves in Supplementary Figure S1). Double staining with anti-CXCR2/CHAT antibodies revealed a substantial CXCR2 immunoreactivity in spinal anterior horns in correspondence to CHAT + neurons, which was visibly increased in ALS samples ( Figure 2). In the same region, CXCL8 immunoreactivity was mainly found in ALS ( Figure 3). Double staining with anti-CXCR2/CHAT antibodies revealed a substantial CXCR2 immunoreactivity in spinal anterior horns in correspondence to CHAT + neurons, which was visibly increased in ALS samples ( Figure 2). In the same region, CXCL8 immunoreactivity was mainly found in ALS ( Figure 3).

CXCR2 Activation by GROα and MIP2α Pro-Inflammatory Chemokines Induces Dose-Dependent Cell Death in NSC-34 Cells Overexpressing hSOD1-G93A
To explore the biological role of CXCR2 axis in ALS, we used an in vitro motor neuron-like model by using the mouse hybrid cell line NSC-34 overexpressing wild type (WT) and mutant human SOD1 (hSOD1-G93A) [30,31]. Western blot and immunofluorescence experiments were carried out to examine the expression of the receptor in both WT and mutant hSOD1-G93A cell line. Interestingly, we observed a consistent genotype-specific expression of CXCR2 in the activated NSC-34 cells carrying the mutated hSOD1 construct (*** p < 0.001 vs. WT + ), while the same was faintly detected in WT cells (Figure 4).

CXCR2 Activation by GROα and MIP2α Pro-Inflammatory Chemokines Induces Dose-Dependent Cell Death in NSC-34 Cells Overexpressing hSOD1-G93A
To explore the biological role of CXCR2 axis in ALS, we used an in vitro motor neuronlike model by using the mouse hybrid cell line NSC-34 overexpressing wild type (WT) and mutant human SOD1 (hSOD1-G93A) [30,31]. Western blot and immunofluorescence experiments were carried out to examine the expression of the receptor in both WT and mutant hSOD1-G93A cell line. Interestingly, we observed a consistent genotype-specific expression of CXCR2 in the activated NSC-34 cells carrying the mutated hSOD1 construct (*** p < 0.001 vs. WT + ), while the same was faintly detected in WT cells (Figure 4).
To further investigate the contribution of CXCR2 activation by endogenous ligands in ALS neuronal depletion, we exposed both WT and mutant hSOD1-G93A NSC-34 cells to increasing concentrations of the two murine functional homologs of CXCL8, i.e., MIP2α (100 pM, 1 nM, 10 nM, 100 nM, 1 µM) and GROα (1 pg/mL, 10 pg/mL, 100 pg/mL, 1 ng/mL, 10 ng/mL, 100 ng/mL), and assessed cellular viability after 24 h. Both GROα and MIP2α treatments determined a significant reduction of cellular viability in a dosedependent manner in hSOD1-G93A NSC-34 cells but not in WT, compared to untreated To further investigate the contribution of CXCR2 activation by endogenous ligands in ALS neuronal depletion, we exposed both WT and mutant hSOD1-G93A NSC-34 cells to increasing concentrations of the two murine functional homologs of CXCL8, i.e., MIP2α (100 pM, 1 nM, 10 nM, 100 nM, 1 µM) and GROα (1 pg/mL, 10 pg/mL, 100 pg/mL, 1 ng/mL, 10 ng/mL, 100 ng/mL), and assessed cellular viability after 24 h. Both GROα and MIP2α treatments determined a significant reduction of cellular viability in a dose-dependent manner in hSOD1-G93A NSC-34 cells but not in WT, compared to untreated cells (** p < 0.01 or *** p < 0.001 vs. Control) ( Figure 5). The minimum effective dose of both ligands to induce significant cell death in hSOD1-G93A NSC-34 cells (1 ng/mL GROα and 100 nM MIP2α) was chosen for the following experiments.
We investigated the CXCR2 expression following ligand treatment in the G93A + background, and observed an increase of CXCR2 immunoreactivity 24 h after MIP2α or GROα incubation (Supplementary Figure S2).  To investigate whether the in vitro neurotoxicity of GROα and MIP2α was specifically mediated by CXCR2, we tested reparixin, a non-competitive allosteric inhibitor of this receptor. Both MTT analysis ( Figure 6) and cell counting (Supplementary Figure S3) showed that treatment with reparixin (10 µM) counteracted the toxicity of both 1 ng/mL GROα and 100 nM MIP2α, significantly inhibiting chemokine-induced cell death and playing a significant neuroprotective role in hSOD1-G93A NSC-34 cells. We investigated the CXCR2 expression following ligand treatment in the G93A + background, and observed an increase of CXCR2 immunoreactivity 24 h after MIP2α or GROα incubation (Supplementary Figure S2).
To investigate whether the in vitro neurotoxicity of GROα and MIP2α was specifically mediated by CXCR2, we tested reparixin, a non-competitive allosteric inhibitor of this receptor. Both MTT analysis ( Figure 6) and cell counting (Supplementary Figure S3) showed that treatment with reparixin (10 µM) counteracted the toxicity of both 1 ng/mL GROα and 100 nM MIP2α, significantly inhibiting chemokine-induced cell death and playing a significant neuroprotective role in hSOD1-G93A NSC-34 cells. To investigate whether the in vitro neurotoxicity of GROα and MIP2α was specifically mediated by CXCR2, we tested reparixin, a non-competitive allosteric inhibitor of this receptor. Both MTT analysis ( Figure 6) and cell counting (Supplementary Figure S3) showed that treatment with reparixin (10 µM) counteracted the toxicity of both 1 ng/mL GROα and 100 nM MIP2α, significantly inhibiting chemokine-induced cell death and playing a significant neuroprotective role in hSOD1-G93A NSC-34 cells.

Activation of CXCR2 Axis by MIP2α Triggers Apoptosis in hSOD1-G93A NSC-34 Cells
To further explore the molecular mechanisms underlying cell death induced by CXCR2 activation, we examined by Western blot analysis the expression of two players involved in apoptosis, the pro-apoptotic BAX and the anti-apoptotic BCL2, following MIP2α and reparixin treatment. Densitometric analysis revealed that CXCR2 activation by MIP2α triggers apoptosis in hSOD1-G93A NSC-34 cells, prompting a significant upregulation of the ratio BAX/BCL2, while the simultaneous antagonism by reparixin significantly reduced the ratio compared to the agonist alone ( Figure 7). These data are consistent with the effects produced on cellular viability. No significant effect was observed for WT cells, with the exception of co-treatment with MIP2α and reparixin, which elicited a downregulation of the BAX/BCL2 ratio. consistent with the effects produced on cellular viability. No significant effect was observed for WT cells, with the exception of co-treatment with MIP2α and reparixin, which elicited a downregulation of the BAX/BCL2 ratio.
To further consolidate these results, we analyzed by fluorescent immunocytochemistry the levels of cleaved caspase-3. Results obtained showed an increased cleaved caspase-3 immunoreactivity in MIP2ɑ-incubated cells, while levels were decreased in combination with reparixin ( Figure 8).  The bar graph shows results from three independent experiments. Relative signal density was quantified using the ImageJ software (Version 1.53t). The protein levels were expressed as arbitrary units obtained after normalization to β-actin, which was used as loading control. Data are expressed as mean ± SEM (*** p < 0.001 vs. Control WT; § § § p < 0.001 vs. MIP2α WT; ### p < 0.001 vs. REP WT; ++ p < 0.01 or +++ p < 0.001 vs. Control G93A; $$ p < 0.01 vs. REP G93A; • • • p < 0.001 vs. MIP2α G93A, as determined by one-Way ANOVA followed by Tukey post hoc test).

Discussion
A plethora of molecular mechanisms have been proposed to account for neuronal damage in ALS, including aberrant RNA metabolism, impaired protein homeostasis, mitochondrial dysfunction, excitotoxicity, and oxidative stress, likely arising from a combination of environmental and genetic risk factors [28,[36][37][38]. Increasing evidence are suggesting that neuroinflammation plays a key role in neuronal degeneration in both SALS and FALS [25,39]. Chronically activated astrocytes, microglia, and infiltrating T cells represent prominent pathological features founded at sites of motor neuron injury on endstage pathology in both patients and animal models [39,40]. In addition, dysregulation in circulating lymphocyte and monocyte populations, as well as altered levels in inflammatory cytokines, chemokines, growth factors (such as VEGF, IFN-γ, TNF-α, IL-1β, IL-6, and IL-10) and their receptors have been reported at different disease stages [22,41].
In the present work, we provide further evidence supporting a role of the G-proteincoupled receptor CXCR2 axis in ALS pathophysiology. The expression of this receptor, mainly involved in mediating inflammatory response, has been previously reported in projections of cerebral cortex neurons, hippocampus, cerebellum [15], human cortical motor neurons (pyramidal cells in layer V) [9], neutrophils [42], monocytes [43], Tlymphocytes, mast cells [44,45], fibroblasts [46], and endothelial cells [47]. CXCR2 expression was also detected in mice spinal neurons (NeuN + cells) [27], and here we show for the first time that the receptor is expressed in human spinal cord anterior horns, and that its level significantly increases in ALS terminal stages (Figures 1 and 2). Interestingly, by correlating the CXCR2 mRNA level to disease progression, we found a trend of association between CXCR2 higher level and short survival (Kaplan-Meier curves in Supplementary Figure S1).

Discussion
A plethora of molecular mechanisms have been proposed to account for neuronal damage in ALS, including aberrant RNA metabolism, impaired protein homeostasis, mitochondrial dysfunction, excitotoxicity, and oxidative stress, likely arising from a combination of environmental and genetic risk factors [28,[36][37][38]. Increasing evidence are suggesting that neuroinflammation plays a key role in neuronal degeneration in both SALS and FALS [25,39]. Chronically activated astrocytes, microglia, and infiltrating T cells represent prominent pathological features founded at sites of motor neuron injury on end-stage pathology in both patients and animal models [39,40]. In addition, dysregulation in circulating lymphocyte and monocyte populations, as well as altered levels in inflammatory cytokines, chemokines, growth factors (such as VEGF, IFN-γ, TNF-α, IL-1β, IL-6, and IL-10) and their receptors have been reported at different disease stages [22,41].
In the present work, we provide further evidence supporting a role of the G-proteincoupled receptor CXCR2 axis in ALS pathophysiology. The expression of this receptor, mainly involved in mediating inflammatory response, has been previously reported in projections of cerebral cortex neurons, hippocampus, cerebellum [15], human cortical motor neurons (pyramidal cells in layer V) [9], neutrophils [42], monocytes [43], T-lymphocytes, mast cells [44,45], fibroblasts [46], and endothelial cells [47]. CXCR2 expression was also detected in mice spinal neurons (NeuN + cells) [27], and here we show for the first time that the receptor is expressed in human spinal cord anterior horns, and that its level significantly increases in ALS terminal stages (Figures 1 and 2). Interestingly, by correlating the CXCR2 mRNA level to disease progression, we found a trend of association between CXCR2 higher level and short survival (Kaplan-Meier curves in Supplementary Figure S1).
Accordingly, we confirmed a significant increase of CXCL8 level in the spinal cord of sporadic ALS patients and reported its expression in spinal ventral horns (Figures 1 and 3).
To further explore the involvement of CXCR2 axis in ALS, we investigated its functional biological role in murine hybrid neuroblastoma-spinal cord NSC-34 cells overexpressing human WT and mutant G93A-SOD1. These cells represent a widely used in vitro model to study MNs in an immortalized system, and constitutively express many phenotypic features of primary motor neurons, such as neurofilament triplet proteins expression, the generation of action potentials, synthesis/storage of acetylcholine, and sensitivity to glutamate insult [31]. Moreover, preliminary observations suggested that in several biological scenarios characterized by mutated SOD1, there is an overall deregulation of the CXCR2/CXCL8 pathway. Indeed, a significant upregulation of CXCR2 ligands emerged in cervical MNs isolated from SOD1 familiar ALS cases compared to control, as well as in fully differentiated iPSC-derived MNs carrying mutated SOD1 (Supplementary Table  S2). Moreover, a previous time-course metanalysis of different transcriptomic profiles revealed a significant increase of Cxcr2 in the spinal cords of SOD1-G93A mice at symptomatic/terminal stages (100-120 days of age) (Supplementary Figure S4) [10,48].
Consistent with these findings, we observed a genotype-specific expression of CXCR2 in hG93A-SOD1 NSC-34 cells, but not in WT hSOD1 (Figure 4). In this regard, a pathological increase of this receptor may provide a paracrine or autocrine milieu, which may enable inflammatory/immune responses inconsistent with healthy neural function [49,50]. Indeed, receptor activation by GROα and MIP2α, two murine endogenous ligands of Cxcr2 and functional homologs of CXCL8, triggered apoptosis and reduced cellular viability in a dose dependent manner (Figures 5-8). Moreover, treatment with reparixin, a non-competitive allosteric CXCR2 inhibitor, counteracted the effects of GROα and MIP2α, inhibiting their toxic effects (Figures 6-8). It is noteworthy that all these findings are consistent with previous observations establishing a neuroprotective effect of reparixin against MIP2αinduced toxicity in rodent-based motor neuronal primary cultures [51], against growthfactor deprivation-induced apoptosis in human degenerating iPSC-derived motor neurons, and in SOD1-G93A transgenic mice [9].

Conclusions
Although further studies are still necessary to completely elucidate the contribution of CXCR2/ligands in ALS pathogenesis and to define the impact of this signaling pathway in motor neuronal selective degeneration, our data further support a role of this axis in ALS pathogenesis and confirm its pharmacological modulation as a candidate therapeutic strategy against ALS.   Funding: This research was funded by the IRIB-CNR project "A multi-omics approach for the study of neurodegeneration" (grant number: DSB.AD007.304 to S.C). E.A. was supported by ALS Stichting (grant "ALS Tissue Bank-NL").

Institutional Review Board Statement:
The study was approved by an ethical committee (Ethics Committee of the Amsterdam Academic Medical Center, approved protocol: W11_073) for medical research and have been performed in accordance with ethical standards, as previously reported [11,12].
Informed Consent Statement: Informed consent was obtained from all individual participants included in the study for the use of tissue and for access to medical records for research purposes.