Sigma-1 receptor counteracts non-cell-autonomous poly-PR-induced astrocytic oxidative stress in C9orf72 ALS

Abstract

C9orf72-associated amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD) are characterized by the accumulation of toxic dipeptide repeat proteins (DPRs) generated from G₄C₂ hexanucleotide repeat expansions. Among these, the arginine-rich poly-PR (proline-arginine) species is the most neurotoxic, eliciting glial activation and neuroinflammation via non-cell-autonomous mechanisms. Although growing evidence implicates glial cells, particularly astrocytes, in disease progression, the molecular pathways linking neuron-derived poly-PR to astrocyte-mediated oxidative stress remain poorly understood. We demonstrate that exogenous poly-PR induces robust NOX4 expression and hydrogen peroxide (H₂O₂) production in astrocytes through activation of the IKK/IκB/NF-κB p65 signaling pathway. Mechanistically, poly-PR promotes nuclear translocation of p65 and enhances its binding to the NOX4 promoter, thereby amplifying astrocytic oxidative stress. Overexpression of the Sigma-1 receptor (Sigma-1R), an endoplasmic reticulum-resident chaperone, significantly attenuates poly-PR-induced NOX4 transcription and reactive oxygen species (ROS) production by interacting with p65 and blocking its nuclear translocation, independently of upstream p65 phosphorylation. Notably, clemastine, a clinically approved Sigma-1R agonist, suppresses astrocytic NOX4 expression by disrupting p65 binding to the NOX4 promoter. In a mouse model of C9orf72 ALS, Sigma-1R deficiency exacerbates poly-PR-induced neurodegeneration, astrogliosis, and NOX4 upregulation, whereas Sigma-1R sufficiency confers neuroprotection and anti-inflammatory effects. This study identifies Sigma-1R as a critical modulator of non-cell-autonomous poly-PR toxicity and establishes its activation as a potent suppressor of astrocyte-derived oxidative stress. Our findings uncover a previously unrecognized glial mechanism driving C9orf72 ALS pathogenesis and support Sigma-1R activation, via clemastine, as a promising therapeutic strategy to mitigate neuroinflammation and disease progression.

Graphical abstract

Highlights

• •Poly-PR induces astrocytic oxidative stress through non-cell-autonomous mechanisms.
• •The IKK/IκB/NF-κB/NOX4 signaling axis mediates ROS production upon poly-PR stimulation.
• •Sigma-1R exerts an antioxidative role in C9orf72-linked ALS.
• •Clemastine, a repurposed drug, shows therapeutic potential for C9orf72 ALS.

1. Introduction

Among the dipeptide repeat proteins (DPRs) produced from GGGGCC (G₄C₂)-RNA repeats in C9orf72-associated amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD), the arginine-rich poly-PR (proline–arginine) species is considered the most neurotoxic and has been strongly associated with neuroinflammation and glial activation [[1], [2], [3], [4]]. In vivo studies utilizing either the LoxP-Cre system or adeno-associated virus (AAV)-mediated gene delivery have demonstrated that poly-PR overexpression elicits pronounced activation of astrocytes and microglia in mouse models [1,5]. In human C9orf72 FTD cases, (G₄C₂)-RNA foci are predominantly detected in neurons [6,7]; nevertheless, they are consistently accompanied by marked glial activation, implicating non-cell-autonomous mechanisms in disease progression. Notably, DPRs, particularly poly-PR, have been shown to be secreted from neurons and transferred to neighboring neurons, astrocytes, and microglia in cellular models [8,9], further supporting the involvement of intercellular propagation in C9orf72-linked pathology.

Neuroinflammation is closely associated with the production of reactive oxygen species (ROS), primarily generated by NADPH oxidase (NOX) enzymes [[10], [11], [12]]. The NOX family comprises seven isoforms, NOX1, NOX2, NOX3, NOX4, NOX5, Duox1, and Duox2 [13]. Among these, NOX2 is predominantly expressed in microglia and neurons, NOX3 is restricted to neurons, and NOX4 is mainly found in astrocytes [14]. Astrocytes, the most abundant glial cell type in the central nervous system, play crucial roles in supporting neuronal communication, regulating ion homeostasis, and maintaining blood-brain barrier integrity [15,16]. Although these cellular functions are well recognized, the molecular mechanisms by which neuronal poly-PR elicits glial activation, particularly in astrocytes, and promotes ROS production via NADPH oxidases through non-cell-autonomous pathways remain largely unclear.

The Sigma-1 receptor (Sigma-1R) is a ligand-activated chaperone protein localized at the mitochondria-associated endoplasmic reticulum membrane [17,18]. Upon activation, Sigma-1R translocates to various cellular compartments, including the plasma membrane, nuclear envelope, and nucleoplasm [[18], [19], [20], [21], [22], [23]]. Sigma-1R is also known to interact with ion channels, nucleoporins, and transcription factors [2,19,22,24]. Previous studies have demonstrated that overexpression of Sigma-1R or treatment with Sigma-1R agonists ameliorates C9orf72-associated ALS phenotypes by stabilizing nucleoporins and facilitating the nuclear translocation of transcription factors in neurons [18,21,22]. In a Drosophila model, expression of (G₄C₂)-RNA repeats induced climbing deficits and retinal degeneration, both of which were rescued by Sigma-1R overexpression [22]. Despite the demonstrated neuronal benefits of Sigma-1R in C9orf72 ALS, whether it can also alleviate astrocyte-mediated non-cell-autonomous oxidative stress remains unknown.

Clemastine ((2R)-2-[2-[(1R)-1-(4-chlorophenyl)-1-phenylethoxy]ethyl]-1-methylpyrrolidine) is a well-characterized histamine H1 receptor antagonist [25,26], but it also functions as a non-selective ligand for the Sigma-1R, exhibiting a binding affinity of Ki = 67 nM [25,27]. While previous studies have attributed its neuroprotective and anti-inflammatory effects to histaminergic pathways [26,28], emerging evidence highlights its potential actions via Sigma-1R-mediated mechanisms [25,27]. Notably, clemastine has been reported to enhance autophagy [26] and reduce neuroinflammation [28]. In the present study, we focus on its role as a Sigma-1R ligand to investigate whether clemastine mitigates astrocyte activation and oxidative stress in C9orf72-associated ALS, particularly in the context of non-cell-autonomous toxicity induced by poly-PR.

In this study, we demonstrate that exogenous poly-PR dipeptide treatment induces NOX4 expression and H₂O₂ production in astrocytes. Mechanistically, the upstream transcription factor p65 translocates into the nucleus and binds to the NOX4 promoter to initiate transcription via the IKK/IĸB signaling pathway. Overexpression of Sigma-1R in astrocytes attenuates poly-PR-induced NOX4 expression and oxidative stress. Notably, activation of Sigma-1R by its ligand clemastine suppresses poly-PR-induced NOX4 expression and H₂O₂ production. Further analysis revealed that Sigma-1R interacts with p65 and blocks its nuclear translocation, thereby preventing its binding to the NOX4 promoter. In an in vivo model, Sigma-1R^+/+ mice exhibited prolonged survival compared to Sigma-1R^−/− littermates following AAV9-hSyn-EGFP-poly-PR₄₂ transduction. Additionally, astrocyte activation was significantly enhanced in Sigma-1R^−/− mice relative to Sigma-1R^+/+ mice. Together, these findings provide novel mechanistic insights into how Sigma-1R mitigates poly-PR-induced astrocytic oxidative stress and neuroinflammation, and they highlight Sigma-1R as a promising therapeutic target in C9orf72-associated ALS. Importantly, our results not only uncover a glial-mediated pathological mechanism but also support the translational potential of pharmacological Sigma-1R activation as a strategy to counteract non-neuronal oxidative stress in C9orf72 ALS.

2. Materials and methods

2.1. Cell culture

Human primary astrocytes were obtained from ScienCell™ Research Laboratories (Catalog #1800, Carlsbad, CA, USA) and cultured in astrocyte medium (Catalog #1801) supplemented with 10 % fetal bovine serum (FBS; Catalog #0010) and 5 mL of astrocyte growth supplement (AGS; Catalog #1852). NSC34 cells were purchased from CELLutions Biosystems, Inc. (Catalog #CLU140; Toronto, ON, Canada) and maintained in Dulbecco's Modified Eagle Medium (DMEM; GIBCO, 11965-092; Thermo Fisher Scientific, Waltham, MA, USA) supplemented with 10 % FBS and 1 % penicillin/streptomycin (GIBCO, 15140-122; Thermo Fisher Scientific). All cells were incubated at 37 °C in a humidified atmosphere containing 5 % CO₂. For transfection, plasmid DNA was mixed with poly-Jet transfection reagent (SignaGen Laboratories, Rockville, MD, USA) at a ratio of 1 μg DNA to 2 μL reagent. The mixture was incubated in 500 μL of serum-free DMEM for 20 min at room temperature (23 °C), then added to a 10-cm culture dish and incubated with the cells for the indicated duration. For cell treatment, astrocytes were seeded in 10-cm culture dishes and incubated for 24 h. The medium was then replaced with fresh astrocyte medium with or without clemastine (MedChemExpress, Monmouth Junction, NJ, USA) and incubated for 3 h. Subsequently, cells were treated with poly-PR₂₀ for the indicated duration. The peptides (poly-PR₂₀ and FITC-poly-PR₂₀) were purchased from Genomics Bioscience & Technology Co., Ltd. (Taipei, Taiwan). Their purity (>98 %) and molecular weight were verified by high-performance liquid chromatography (HPLC) and mass spectrometry, respectively. Lyophilized peptides were dissolved in sterile PBS, aliquoted, and stored at −80 °C until use. Astrocytes were pretreated with NE100 (3 μM; MedChemExpress) for 1 h prior to clemastine (2.5 μM) and poly-PR₂₀ peptide exposure.

2.2. Animal model

Based on a previously established approach [2], the AAV9 construct containing the hSyn promoter-EGFP-PR₄₂ was produced by the AAV Core Facility at Academia Sinica, Taiwan, which specializes in recombinant AAV generation. Neonatal wild-type or Sigma-1R knockout (SigmaR1^−/−) C57BL/6J mice at postnatal day 0 (P0) received intracerebroventricular injections of 2 μL (8.2 × 10¹³ genomes/mL) AAV9-hSyn-EGFP-PR₄₂, as previously described. Mice were monitored and analyzed at the indicated ages (postnatal days 11–13) for assessments of acute neurodegeneration, survival, and histological analyses. All animal procedures were conducted in accordance with the guidelines of the Institutional Animal Care and Use Committee (IACUC) of China Medical University (Protocol No. CMUIACUC-2025-007). C57BL/6JNarl mice were obtained from the National Center for Biomodels (NCB), National Institute of Animal Resources (NIAR), Taiwan. Mice were housed in a controlled environment maintained at 22 ± 0.5 °C with 60 ± 15 % relative humidity and a 12-h light/dark cycle, with ad libitum access to food and water.

2.3. Conditional media collection

Astrocytes or NSC34 cells were seeded in 10-cm culture dishes and incubated overnight. Cells were then transfected with the EGFP-PR₄₂ plasmid for 24 h, followed by washing and replacement with fresh serum-free DMEM. The cells were further incubated for an additional 48 or 72 h. Conditioned media were subsequently collected and concentrated tenfold using 10K centrifugal filter devices. Naïve astrocytes were then treated with the concentrated media for 48 h.

2.4. RT-qPCR

Total RNA was extracted from treated cells using the RNeasy Mini Kit (Qiagen, Hilden, Germany) according to the manufacturer's instructions. Briefly, cells were lysed in lysis buffer and mixed with 70 % ethanol, followed by purification using the kit's wash buffers. The eluted RNA was reverse transcribed into complementary DNA (cDNA) using the iScript™ cDNA Synthesis Kit (#1708891; Bio-Rad Laboratories, Hercules, CA, USA). Quantitative real-time PCR was performed using the CFX96 Touch Real-Time PCR Detection System (Bio-Rad Laboratories) with iTaq™ Universal SYBR Green Supermix (#1725122; Bio-Rad Laboratories) and gene-specific primers (Table 1). Cycle threshold (Ct) values were obtained, and relative gene expression levels were calculated using the 2^−ΔΔCt method.

Table 1.

Antibodies and primers.

REAGENT OR RESOURCE	SOURCE	IDENTIFIER
Antibodies
NOX4	Proteintech	14347-1-AP; RRID:AB_10638146
NOX4	Novus Biologicals	NB110-58849; RRID:AB_877739
GFP	Rockland Immunochemicals	600-901-215; RRID:AB_1537403
GFP	Proteintech	50430-2-AP; RRID:AB_11042881
GFP	Proteintech	66002-1-Ig; RRID:AB_11182611
NeuN	ABclonal	A19086; RRID:AB_2862578
NFκB (p65)	Merck	06-418; RRID:AB_11214166
NFκB (p-p65)	Merck	MAB3026; RRID:AB_2178887
p-IKKα/β	Cell Signaling	2697s; RRID:AB_2079382
GAPDH	Merck	MAB374; RRID:AB_2107445
IKKα	Novus	NB100-56704; RRID:AB_838409
IKKβ	Cell Signaling	8943s; RRID:AB_11024092
IgG	Cell Signaling	2729S; RRID:AB_1031062
IgG	Santa Cruz	sc-2025; RRID:AB_737182
p-IĸBα	Cell Signaling	9246s; RRID:AB_2267145
IĸBα	Cell Signaling	4812s; RRID:AB_10694416
GRP78 (BiP)	Invitrogen	PA5-19503; RRID:AB_1097803
GFAP	Merck	AB4674; RRID:AB_304558
β-actin	Proteintech	66009-1-Ig; RRID:AB_2687938
HA-tag	Proteintech	66006-2-Ig; RRID:AB_2881490
HA-tag	Proteintech	51064-2-AP; RRID:AB_11042321
Sigma-1R	Santa Cruz Biotechnology	sc-137075; RRID:AB_2285870
Donkey anti-Mouse IgG (H + L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor™ Plus 555	Invitrogen	A32773; RRID: AB_2762848
Donkey anti-Mouse IgG (H + L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor™ Plus 488	Invitrogen	A32766; RRID: AB_2762823
Donkey anti-Rabbit IgG (H + L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor™ Plus 555	Invitrogen	A32794; RRID: AB_2762834
Donkey anti-Rabbit IgG (H + L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor™ Plus 488	Invitrogen	A32790; RRID: AB_2762833
Peroxidase IgG Fraction Monoclonal Mouse Anti-Rabbit IgG, light chain specific	Jackson ImmunoResearch	211-032-171; RRID: AB_2339149
Peroxidase AffiniPure® Goat Anti-Mouse IgG, light chain specific	Jackson ImmunoResearch	115-035-174; RRID:AB_2338512
Virus strains
AAV-hsyn-EGFP	AAV Core Facility of Academia Sinica, Taiwan	N/A
AAV-hsyn-EGFP-PR42	This study	N/A
Oligonucleotides
Tubulin	GENOMICS	F: CGG GCA GTG TTT GTA GAC TTG C R: CTC CTT GCC AAT GGT GTA GTG C
NOX1	GENOMICS	F: GGT TTT ACC GCT CCC AGC AGA A R: CTT CCA TGC TGA AGC CAC GCT T
NOX2	GENOMICS	F: AGC TAT GAG GTG GTG ATG TTA GTG G R: CAC AAT ATT TGT ACC AGA CAG ACT TGA G
NOX3	GENOMICS	F: CCT GGA AAC ACG GAT GAG TGA G R: CCT CCC ATA GAA GGT CTT CTG C
NOX4	GENOMICS	F: GCC AGA GTA TCA CTA CCT CCA C R: CTC GGA GGT AAG CCA AGA GTG T
NOX5	GENOMICS	F: CCA CCA TTG CTC GCT ATG AGT G R: GCC TTG AAG GAC TCA TAC AGC C
SigmaR1 Wild-type	GENOMICS	F: TCT GAG TAC GTC CTG CTC TTC G R: CAG AAA TCT CAG CCC AGT ATC G
SigmaR1 Knockout	GENOMICS	F: AGG ATC TCC TGT CAT CTC ACC TTG CTC CTG R: AAG AAC TCG TCA AGA AGG CGA TAG AAG GGG

2.5. Western blot

Cells were harvested and lysed in IP lysis buffer composed of 50 mM NaCl, 0.5 % NP-40, and 10 mM Tris-HCl (pH 8.0), supplemented with a complete protease inhibitor cocktail (1 × ), and incubated on ice for 30 min. After quantification, equal amounts of protein were denatured by heating at 95 °C for 10 min. Samples were resolved using SDS-PAGE with gels of appropriate acrylamide concentrations and transferred onto PVDF membranes. Membranes were then blocked with 5 % Blotting-Grade Blocker (#1706404; Bio-Rad Laboratories) at room temperature for 1 h. Following blocking, membranes were incubated overnight at 4 °C with specific primary antibodies (Table 1). After three washes in TBST, membranes were incubated with HRP-conjugated secondary antibodies for 1 h at room temperature. Signals were detected using the Azure 400 imaging system (Azure Biosystems, Dublin, CA, USA), and band intensities were quantified using Image Studio Lite software (version 5.2; LI-COR, Lincoln, NE, USA) in accordance with the manufacturer's instructions.

2.6. Immunoprecipitation

Cell lysates from transfected and treated cells were prepared using immunoprecipitation buffer (50 mM NaCl, 0.5 % NP-40, 10 mM Tris-HCl, pH 8.0) supplemented with a protease inhibitor cocktail. Protein extracts were incubated with specific primary antibodies or control IgG at 4 °C for 2 h. The immune complexes were subsequently captured using protein A/G agarose beads (sc-2003; Santa Cruz Biotechnology, Dallas, TX, USA) and rotated overnight at 4 °C. After incubation, the beads were washed three times with IP buffer, eluted with 2 × Laemmli sample buffer containing β-mercaptoethanol, and denatured by heating at 95 °C for 10 min. Both immunoprecipitated and input protein samples were subjected to Western blot analysis.

2.7. H₂O₂ assay

Following the manufacturer's instructions (G8821, Promega, WI, USA), cells and conditioned media were incubated with the H₂O₂ Substrate Solution for 6 h. After incubation, 100 μL of freshly prepared ROS-Glo™ Detection Solution was added to each sample and incubated for 20 min at room temperature (23 °C). Luminescence was then measured using a SpectraMax® iD3 microplate reader.

2.8. DHE (dihydroethidium) assay

According to the manufacturer's instructions (ab236206, Abcam, MA, USA), cultured cells were first washed with 150 μL Cell-Based Assay Buffer. After aspirating the buffer, 130 μL of ROS Staining Buffer was added to each well and incubated for 30 min at 37 °C in the dark. Following incubation, the ROS Staining Buffer was removed and replaced with 100 μL of fresh Cell-Based Assay Buffer. Fluorescence was measured using a SpectraMax® iD3 microplate reader at an excitation wavelength of 480–520 nm and an emission wavelength of 570–600 nm.

2.9. CCK8 assay

Cells were exposed to varying concentrations of clemastine at 37 °C in a humidified incubator with 5 % CO₂ for 24 h. Following treatment, Cell Counting Kit-8 (CCK-8; ab228554, Abcam) reagent was added in diluted form, and the cells were incubated for an additional 1 h at 37 °C. After incubation, the resulting formazan product was solubilized in the medium. Absorbance was measured at 450 nm using a SpectraMax® iD3 microplate reader to assess cell viability.

2.10. Luciferase reporter assay

The 5′ flanking regions of the NOX4 gene (−710 to +1 and −210 to +1) were synthesized by Genomics (Taipei, Taiwan) and individually cloned into the pGL3-Basic vector (Promega). For the luciferase reporter assay, cells were transfected with the reporter constructs and indicated expression vectors using polyJet transfection reagent (SignaGen), with or without poly-PR₂₀ and clemastine treatment. Cell lysates were collected according to the manufacturer's instructions using the Luciferase Assay System (E1500; Promega). Luminescence was quantified using a SpectraMax® iD3 microplate reader.

2.11. Immunofluorescence

For cell staining, cells were seeded onto poly-l-lysine-coated coverslips and cultured overnight. Following the treatment protocol described in the Cell Culture section, cells were incubated with FITC-poly-PR₂₀. After treatment, cells were fixed with 4 % paraformaldehyde and permeabilized using 0.1 % Triton X-100 in PBS. Subsequently, cells were blocked with 10 % serum in PBS for 1 h at room temperature. For tissue staining, 15-μm-thick brain sections were mounted onto glass slides and washed in PBS for 10 min. Antigen retrieval was performed by incubating the slides at 80 °C for 10 min, followed by blocking with 10 % serum in PBS for 1 h at room temperature. Primary antibodies (Table 1) were applied to either coverslips or brain sections and incubated overnight at 4 °C. After washing with 0.1 % Triton X-100 in TBST, samples were incubated with appropriate secondary antibodies at room temperature for 1 h. Nuclei were counterstained with 4’,6-diamidino-2-phenylindole (DAPI) in PBS. Images were captured using an LSM900 confocal microscope (Carl Zeiss, Oberkochen, Germany) equipped with either a Plan-Apochromat 20 × or 63 × /1.3 NA objective and processed using confocal or Airyscan acquisition modes. Fluorescence intensities were quantified using ImageJ/Fiji software (version 1.53t; NIH, Bethesda, MD, USA).

2.12. Statistical analysis

All statistical analyses were performed using Prism software (version 10.4.2; GraphPad Software, La Jolla, CA, USA). Data were obtained from at least three biologically independent replicates for cell-based experiments and from a minimum of 10 mice in animal studies. Comparisons between two groups were conducted using two-tailed Student's t-tests. For comparisons involving more than two groups, one-way or two-way analysis of variance (ANOVA) was performed, followed by Tukey's or Šídák's multiple comparisons test when appropriate. Data are presented as mean ± standard error of the mean (SEM). Statistical significance was defined as ∗P < 0.05, ∗∗P < 0.01, ∗∗∗P < 0.001, and ∗∗∗∗P < 0.0001.

3. Results

3.1. Poly-PR induces NOX4 expression and H₂O₂ production in astrocytes through non-cell-autonomous mechanisms, but not via overexpression within astrocytes

Oxidative stress, such as H₂O₂ production, is primarily mediated by NADPH oxidase activation [[10], [11], [12]]. Among the NADPH oxidase isoforms, NOX4 is predominantly expressed in astrocytes (Supplementary Fig. 1A) [14]. Dipeptide repeat proteins, particularly poly-PR, can be secreted from neuronal cells [8]. In this study, we demonstrated that astrocytes or NSC34 cells expressing EGFP-PR₄₂ were capable of secreting poly-PR, which was subsequently internalized by naïve astrocytes (Supplementary Fig. 2). However, it remains unclear whether poly-PR, introduced either through plasmid-driven overexpression or exogenous peptide treatment, is capable of inducing NOX4 expression and stimulating H₂O₂ production in astrocytes. To address this, we first examined the effect of poly-PR₄₂ overexpression on NOX4 mRNA and protein levels in astrocytes. Our results showed no significant changes in NOX4 transcript or protein expression in poly-PR₄₂-expressing astrocytes (Fig. 1A–C). We then evaluated whether exogenous poly-PR peptide treatment could modulate NOX4 expression. Notably, treatment with synthetic poly-PR₂₀ peptides led to a significant increase in NOX4 mRNA and protein levels in astrocytes (Fig. 1D–F). Furthermore, exogenous poly-PR₂₀ peptide exposure elevated both extracellular and intracellular H₂O₂ production (Fig. 1G–I) and ROS formation (Supplementary Fig. 1B). These findings suggest that the arginine-rich dipeptide repeat protein poly-PR induces NOX4 expression and H₂O₂ generation in astrocytes via non-cell-autonomous mechanisms, but not through overexpression within astrocytes (Fig. 1J).

Fig. 1.

Exogenous, but not overexpressed, poly-PR induces NOX4 expression and increases H₂O₂ production in astrocytes. (A) RT-qPCR analysis revealed no significant difference in NOX4 mRNA expression between astrocytes expressing EGFP and those expressing EGFP-PR₄₂ at 24 h. (B) Western blot analysis showed no significant change in NOX4 protein levels in astrocytes overexpressing either EGFP or EGFP-PR₄₂. (C) Quantification of NOX4 protein levels from three independent experiments performed on biologically independent replicates, all yielding similar results. Data are presented as mean ± SEM and analyzed using an unpaired two-tailed t-test. (D) RT-qPCR analysis demonstrated a time-dependent increase in NOX4 mRNA levels in astrocytes following exogenous poly-PR₂₀ peptide treatment. Data are presented as mean ± SEM and analyzed using one-way ANOVA followed by Tukey's multiple comparisons test (∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001). N = 3 biologically independent experiments with consistent results. (E) Western blot analysis revealed increased NOX4 protein expression in astrocytes treated with exogenous poly-PR₂₀ peptides. (F) Quantification of NOX4 protein levels from three biologically independent experiments, yielding similar results. Data are presented as mean ± SEM and analyzed using one-way ANOVA followed by Tukey's multiple comparisons test (∗p < 0.05, ∗∗p < 0.01). (G) Schematic diagram illustrating the experimental design for H₂O₂ detection following poly-PR₂₀ treatment. (H, I) Extracellular (H) and intracellular (I) H₂O₂ levels were measured using a luminescence-based H₂O₂ assay (see Methods). Astrocytes were treated with varying concentrations of poly-PR₂₀ for 24 or 48 h prior to measurement. Data are presented as mean ± SEM and analyzed using two-way ANOVA followed by Tukey's multiple comparisons test (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001). N = 3 biologically independent experiments with consistent results. (J) Schematic representation showing that exogenous poly-PR, secreted from poly-PR-expressing motor neurons or astrocytes, can be internalized by naïve astrocytes, leading to NOX4 upregulation and increased H₂O₂ production.

3.2. Exogenous Poly-PR enhances NF-κB p65-mediated transcription of NOX4 via promoter binding

To explore upstream transcriptional regulators of NOX4, we employed the PROMO prediction tool to identify potential transcription factors binding to the NOX4 promoter. Five candidate transcription factors were identified: SOX2, FOS, OCT1, SP3, and NF-κB p65. To validate their regulatory effects, each transcription factor was individually overexpressed in human astrocytes. Among them, only NF-κB p65 significantly increased NOX4 mRNA expression, while SOX2, FOS, OCT1, and SP3 showed no appreciable effect (Fig. 2A–E). We next examined whether exogenous poly-PR dipeptide treatment could potentiate p65-induced NOX4 expression. Co-treatment with poly-PR and p65 overexpression further enhanced NOX4 mRNA and protein levels compared to poly-PR treatment alone in astrocytes (Fig. 2F–I), suggesting a synergistic effect of poly-PR and NF-κB p65 on NOX4 transcriptional activation. To determine whether p65 directly binds to the NOX4 promoter, we synthesized two promoter fragments: a full-length region spanning −710 to +1 (NOX4/FL-pGL3) and a proximal region spanning −210 to +1 (NOX4/PI-pGL3) (Fig. 2J). Luciferase reporter assays demonstrated that p65 significantly enhanced promoter activity in both constructs, with particularly strong activation observed in the −210 to +1 region (Fig. 2K). These results suggest that NF-κB p65 directly regulates NOX4 transcription by binding to its proximal promoter region, and that poly-PR enhances this transcriptional activation in astrocytes.

Fig. 2.

NF-κB p65 regulates NOX4 transcription via promoter binding and facilitates poly-PR₂₀-induced NOX4 expression. (A–E) RT-qPCR analysis of NOX4 mRNA levels in astrocytes following overexpression of the transcription factors SOX2 (A), FOS (B), OCT1 (C), SP3 (D), and NF-κB p65 (E). Only p65 significantly enhanced NOX4 mRNA expression. Data are presented as mean ± SEM and analyzed using an unpaired two-tailed t-test (∗∗∗p < 0.001). N = 3 biologically independent experiments with consistent results. (F) Schematic diagram illustrating the experimental design for assessing NOX4 mRNA expression, protein levels, and luciferase reporter activity following poly-PR₂₀ treatment and p65 overexpression. (G–I) Co-treatment with poly-PR₂₀ and p65 overexpression further increased NOX4 mRNA expression (G) and protein levels (H, I) compared to poly-PR₂₀ treatment alone. Data are presented as mean ± SEM and analyzed using two-way ANOVA followed by Tukey's multiple comparisons test (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). N = 3 biologically independent experiments with consistent results. (J) Schematic representation of predicted NF-κB p65-responsive elements within the NOX4 promoter and design of luciferase reporter constructs (NOX4/FL-pGL3: −710 to +1; NOX4/PI-pGL3: −210 to +1). (K) Luciferase reporter assay showing that p65 enhances transcriptional activity through the NOX4 promoter. Cells were co-transfected with HA-p65 and the indicated NOX4 reporter constructs. Data are presented as mean ± SEM and analyzed using one-way ANOVA followed by Tukey's multiple comparisons test (∗∗∗∗p < 0.0001). N = 3 biologically independent experiments with consistent results.

3.3. Exogenous poly-PR facilitates NF-κB p65 nucleocytoplasmic translocation via the IKK/IκB signaling pathway

To determine whether exogenous poly-PR dipeptide treatment activates p65 phosphorylation, astrocytes were treated with poly-PR₂₀ for 1, 3, and 6 h. Phosphorylated p65 levels were significantly increased at 3 h post-treatment (Fig. 3A and B). We next examined whether poly-PR₂₀ promotes the nuclear translocation of p65. Astrocytes were treated with FITC-poly-PR₂₀ for 3 h, and the results showed that p65 translocated into the nucleus following treatment (Fig. 3C). Line-scan analysis revealed that the p65 signal (red line) overlapped with the nuclear region (blue line), confirming nuclear localization (Fig. 3D). Moreover, FITC-poly-PR₂₀ treatment significantly increased the nuclear-to-cytosolic ratio of p65 (Fig. 3E). Given that the IKK/IκB pathway is a primary upstream regulator of p65 activation, we next investigated whether poly-PR₂₀ modulates this signaling cascade. Poly-PR₂₀ treatment induced phosphorylation of IKKα/β as early as 30 min (Fig. 3F–H), followed by IκB phosphorylation at 2 h (Fig. 3I and J). These findings indicate that exogenous poly-PR activates the IKK/IκB pathway, leading to p65 phosphorylation and nuclear translocation.

Fig. 3.

Exogenous poly-PR₂₀ activates p65 phosphorylation and nuclear translocation in astrocytes via the IKK/IκB pathway. (A) Western blot analysis of phosphorylated p65 (p-p65) and total p65 in astrocytes treated with poly-PR₂₀. p65 phosphorylation was notably increased at 3 h post-treatment. (B) Quantification of the p-p65/p65 ratio from three biologically independent experiments, all yielding consistent results. Data are presented as mean ± SEM and analyzed using one-way ANOVA followed by Tukey's multiple comparisons test (∗p < 0.05). (C) Immunofluorescence staining showing nuclear translocation of p65 in astrocytes treated with FITC-poly-PR₂₀ for 3 h. (D) Confocal microscopy with Z-stack and line-scan analysis confirmed p65 nuclear localization. Red: p65; Green: FITC-poly-PR₂₀; Blue: DAPI. (E) Semi-quantitative analysis of nuclear versus cytosolic p65 using ImageJ (version 1.53t) from confocal images. Data are presented as mean ± SEM and analyzed using an unpaired two-tailed t-test (∗∗∗p < 0.001). Total number of cells analyzed: control group, n = 24; FITC-poly-PR₂₀ group, n = 24. N = 3 biologically independent experiments with consistent results. (F) Western blot analysis showing activation of IKKα/β phosphorylation following poly-PR₂₀ treatment. (G, H) Quantification of phosphorylated IKKα/β relative to total IKKα (G) and IKKβ (H) revealed a significant increase at 30 min post-treatment. Data are presented as mean ± SEM and analyzed using one-way ANOVA followed by Tukey's multiple comparisons test (∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). N = 3 biologically independent experiments with consistent results. (I) Western blot analysis showing increased phosphorylation of IκBα in astrocytes following poly-PR₂₀ treatment. (J) Quantification of the p-IκBα/IκBα ratio showed a significant increase at 2 h. Data are presented as mean ± SEM and analyzed using one-way ANOVA followed by Tukey's multiple comparisons test (∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). N = 3 biologically independent experiments with consistent results.

3.4. Sigma-1R suppresses poly-PR-induced NOX4 expression and H₂O₂ production in astrocytes

Given the pivotal role of NOX4-mediated oxidative stress in poly-PR-induced astrocytic responses, we next investigated potential protective mechanisms that might counteract this effect. Sigma-1R, a multifunctional chaperone protein expressed in both neurons and astrocytes, is known to modulate oxidative stress and exert neuroprotective and anti-inflammatory effects [18,29]. However, whether astrocytic Sigma-1R can mitigate poly-PR-induced oxidative stress remains unclear. At the beginning, exogenous poly-PR peptide treatment did not alter Sigma-1R expression (Supplementary Fig. 3). We further found that Sigma-1R overexpression significantly reduced poly-PR-induced NOX4 transcription and protein expression in astrocytes (Fig. 4A–C). Moreover, Sigma-1R overexpression attenuated both intracellular and extracellular H₂O₂ production following poly-PR exposure (Fig. 4D–F). These findings suggest that astrocytic Sigma-1R alleviates poly-PR-induced oxidative stress by downregulating NOX4 expression (Fig. 4G).

Fig. 4.

Sigma-1R attenuates astrocytic NOX4 expression and H₂O₂ production. (A) Western blot analysis of NOX4 protein levels in astrocytes treated with exogenous poly-PR₂₀ following overexpression of Sigma-1R-GFP or GFP control. (B) Quantification of NOX4 protein levels revealed a significant reduction in NOX4 expression in Sigma-1R-overexpressing astrocytes compared to GFP controls under exogenous poly-PR₂₀ peptide treatment. Data are presented as mean ± SEM and analyzed using one-way ANOVA followed by Tukey's multiple comparisons test (∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001). N = 3 biologically independent experiments with consistent results. (C) RT-qPCR analysis demonstrated decreased NOX4 mRNA expression in Sigma-1R-expressing astrocytes compared to GFP controls. Data are presented as mean ± SEM and analyzed using one-way ANOVA followed by Tukey's multiple comparisons test (∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). N = 3 biologically independent experiments with consistent results. (D) Schematic diagram outlining the experimental design for assessing H₂O₂ production in astrocytes treated with exogenous poly-PR₂₀ and overexpressing Sigma-1R. (E, F) Intracellular (E) and extracellular (F) H₂O₂ levels were quantified using a luminescence-based H₂O₂ detection assay. Astrocytes overexpressing GFP or Sigma-1R-GFP were treated with poly-PR₂₀ for 24 or 48 h prior to measurement. Data are presented as mean ± SEM and analyzed using two-way ANOVA followed by Tukey's multiple comparisons test (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001). N = 3 biologically independent experiments with consistent results. (G) Schematic summary illustrating that Sigma-1R overexpression suppresses poly-PR-induced NOX4 expression and H₂O₂ production in astrocytes.

3.5. Clemastine, a clinically used Sigma-1R agonist, activates Sigma-1R to suppress NOX4 expression and attenuate H₂O₂ production in astrocytes

Clemastine is a histamine H1 receptor antagonist [25,26] and also functions as a non-selective ligand for the Sigma-1R [25,27]. Activation of Sigma-1R has been shown to reduce neuroinflammation and exert neuroprotective effects [28]. However, whether clemastine attenuates NOX4-mediated oxidative stress in astrocytes in the context of poly-PR-induced C9orf72-associated ALS remains unknown. To address this, we first determined the optimal concentration of clemastine required to activate Sigma-1R in astrocytes. Based on previous studies [17,18,21], Sigma-1R activation was assessed by evaluating the dissociation of BiP from Sigma-1R. We found that clemastine at 2.5 μM effectively promoted BiP dissociation, indicating optimal Sigma-1R activation (Fig. 5A and B). Importantly, clemastine did not exhibit cytotoxic effects on astrocytes, even at concentrations up to 10 μM (Fig. 5C). Given our earlier findings that Sigma-1R overexpression suppressed NOX4 expression and H₂O₂ production in poly-PR-treated astrocytes (Fig. 4), we next investigated whether clemastine could exert similar protective effects through pharmacological activation of Sigma-1R. Pre-treatment of astrocytes with clemastine significantly reduced poly-PR₂₀-induced NOX4 mRNA and protein expression (Fig. 5D–F). In addition, clemastine treatment alone did not alter basal NOX4 expression in astrocytes (Supplementary Fig. 4). Furthermore, clemastine markedly attenuated both intracellular and extracellular H₂O₂ production in astrocytes exposed to poly-PR₂₀ (Fig. 5G–I). Together, these results demonstrate that clemastine functions as a Sigma-1R agonist to suppress poly-PR-induced NOX4 expression and oxidative stress in astrocytes, highlighting its potential therapeutic value in C9orf72-associated ALS.

Fig. 5.

Clemastine, a clinically used Sigma-1R agonist, reduces NOX4 expression and H₂O₂ production in astrocytes. (A) Immunoprecipitation analysis of BiP-Sigma-1R interaction following clemastine treatment. Sigma-1R agonist activity leads to dissociation from BiP, indicative of functional activation. (B) Quantification of BiP associated with Sigma-1R showed a significant reduction at 2.5 μM clemastine. Data are presented as mean ± SEM and analyzed using one-way ANOVA followed by Tukey's multiple comparisons test (∗p < 0.05, ∗∗∗∗p < 0.0001). N = 3 biologically independent experiments with consistent results. (C) CCK-8 assay assessing astrocyte viability following clemastine treatment. (D) Western blot analysis of NOX4 protein levels in astrocytes treated with exogenous poly-PR₂₀, with or without 2.5 μM clemastine. (E) Quantification of NOX4 protein levels revealed a significant reduction in clemastine-treated astrocytes under poly-PR₂₀ exposure. Data are presented as mean ± SEM and analyzed using one-way ANOVA followed by Tukey's multiple comparisons test (∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). N = 3 biologically independent experiments with consistent results. (F) RT-qPCR analysis showed decreased NOX4 mRNA expression in clemastine-treated astrocytes under poly-PR₂₀ conditions. Data are presented as mean ± SEM and analyzed using one-way ANOVA followed by Tukey's multiple comparisons test (∗∗p < 0.01, ∗∗∗∗p < 0.0001). N = 3 biologically independent experiments with consistent results. (G) Schematic diagram illustrating the experimental timeline for H₂O₂ detection following clemastine and poly-PR₂₀ treatment. (H, I) Intracellular (H) and extracellular (I) H₂O₂ levels were measured using a luminescence-based detection assay. Astrocytes were pretreated with clemastine, followed by poly-PR₂₀ for 24 or 48 h. Data are presented as mean ± SEM and analyzed using two-way ANOVA followed by Tukey's multiple comparisons test (∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001). N = 3 biologically independent experiments with consistent results.

3.6. Sigma-1R and p65 interaction suppresses NOX4 expression by blocking p65 nuclear translocation without affecting phosphorylation

To elucidate the mechanism by which Sigma-1R suppresses NOX4 expression and H₂O₂ production in astrocytes under poly-PR₂₀ exposure, we first examined whether Sigma-1R modulates p65 phosphorylation. Our results showed that Sigma-1R overexpression did not alter poly-PR₂₀-induced p65 phosphorylation levels (Fig. 6A and B). We then investigated whether Sigma-1R affects the transcriptional activity of p65 on the NOX4 promoter. Luciferase reporter assays revealed that Sigma-1R overexpression significantly reduced p65-driven NOX4 promoter activity in poly-PR₂₀-treated astrocytes compared to the GFP control group (Fig. 6C and D). Similarly, pre-treatment with clemastine, a Sigma-1R agonist, also diminished p65-mediated NOX4 promoter activation under poly-PR₂₀ stimulation (Fig. 6E and F). Co-immunoprecipitation experiments demonstrated that Sigma-1R physically interacts with p65 in astrocytes following poly-PR₂₀ treatment (Fig. 6G). Moreover, Sigma-1R overexpression reduced p65 nuclear translocation in astrocytes exposed to exogenous poly-PR dipeptide treatment (Fig. 6H). These findings suggest that Sigma-1R suppresses NOX4 transcription not by altering upstream signaling but by directly interacting with p65 and blocking its binding to the NOX4 promoter.

Fig. 6.

Sigma-1R interacts with p65 and reduces its binding to the NOX4 promoter in astrocytes under exogenous poly-PR₂₀ peptide treatment. (A) Western blot analysis of phosphorylated p65 (p-p65) levels in astrocytes overexpressing Sigma-1R-GFP or GFP control, followed by exogenous poly-PR₂₀ peptide treatment. Astrocytes were transfected for 24 h and treated with exogenous poly-PR₂₀ peptide for 3 h. (B) Quantification of the p-p65/p65 protein ratio from three biologically independent experiments, each yielding consistent results. Data are presented as mean ± SEM and analyzed using an unpaired two-tailed t-test. (C) Schematic diagram illustrating the experimental timeline for the luciferase reporter assay following co-transfection of GFP or Sigma-1R-GFP with HA-p65, followed by exogenous poly-PR₂₀ peptide treatment. (D) Overexpression of Sigma-1R-GFP reduced p65-mediated activation of the NOX4 promoter. Data are presented as mean ± SEM and analyzed using an unpaired two-tailed t-test (∗p < 0.05). N = 3 biologically independent experiments with consistent results. (E) Schematic diagram illustrating the experimental timeline for the luciferase reporter assay after clemastine and poly-PR₂₀ treatment. (F) Clemastine treatment also reduced p65-driven NOX4 promoter activation. Data are presented as mean ± SEM and analyzed using an unpaired two-tailed t-test (∗p < 0.05). N = 3 biologically independent experiments with consistent results. (G) Immunoprecipitation analysis of p65 and Sigma-1R interaction in astrocytes treated with poly-PR₂₀ confirmed a physical interaction between Sigma-1R and p65. (H) Immunofluorescence analysis demonstrated that Sigma-1R overexpression in astrocytes attenuated p65 nuclear translocation in response to exogenous poly-PR peptide treatment. Data are presented as mean ± SEM and analyzed using an unpaired two-tailed t-test (∗∗∗∗p < 0.0001). Total number of cells analyzed: EGFP + poly-PR₂₀ group, n = 18; Sigma-1R-EGFP + poly-PR₂₀ group, n = 13. N = 3 biologically independent experiments with consistent results.

3.7. Sigma-1R deficiency exacerbates poly-PR-induced neurodegeneration in vivo

To investigate the in vivo role of Sigma-1R in poly-PR-induced neurodegeneration, we employed an adeno-associated virus (AAV)-mediated system to selectively express poly-PR in neurons under the control of the hSyn promoter, thereby establishing a C9orf72 ALS mouse model (Fig. 7A). Sigma-1R wild-type and knockout mice were genotyped and validated by Western blot to confirm Sigma-1R deficiency (Fig. 7B). Survival analysis revealed that neuronal poly-PR expression significantly shortened the lifespan of Sigma-1R knockout mice compared to wild-type controls, with an average reduction of 15.38 % (Fig. 7C). Additionally, poly-PR expression led to a reduction in body weight and brain length in both genotypes (Fig. 7D–G). Morphological examination further revealed greater brain atrophy in Sigma-1R-deficient mice relative to wild-type littermates. These findings indicate that loss of Sigma-1R exacerbates poly-PR-induced neurodegeneration, supporting its neuroprotective role in C9orf72-associated ALS pathogenesis.

Fig. 7.

SigmaR1^−/− mice exhibit exacerbated neurodegeneration and reduced survival compared to SigmaR1^+/+ mice. (A) Schematic diagram illustrating the experimental design for generating the poly-PR-induced C9orf72 ALS mouse model. (B) Genotyping (top) and Western blot analysis (bottom) confirming SigmaR1^+/+ and SigmaR1^−/− backgrounds. (C) Kaplan-Meier survival analysis showing that EGFP-Poly-PR₄₂ expression significantly reduced survival in SigmaR1^−/− mice (n = 10) compared to SigmaR1^+/+ mice (n = 10). (D, F) Representative images of whole-body (D) and brain size (F) in EGFP and EGFP-Poly-PR₄₂ mice. EGFP-Poly-PR₄₂-expressing mice exhibited visibly smaller body and brain sizes in both genotypes. (E) Quantification of body weight in SigmaR1^+/+ mice (EGFP: n = 11; EGFP-PR₄₂: n = 12) and SigmaR1^−/− mice (EGFP: n = 4; EGFP-PR₄₂: n = 8). (G) Quantification of brain length in EGFP or EGFP-Poly-PR₄₂-injected SigmaR1^+/+ mice (EGFP: n = 8; EGFP-PR₄₂: n = 5) and SigmaR1^−/− mice (EGFP: n = 4; EGFP-PR₄₂: n = 4). Data are presented as mean ± SEM and analyzed using two-way ANOVA followed by Šídák's multiple comparisons test (∗∗∗∗p < 0.0001).

3.8. Sigma-1R deficiency enhances astrogliosis and NOX4 upregulation in poly-PR-induced C9orf72 ALS

Building on our in vitro findings that exogenous poly-PR promotes NOX4 expression and oxidative stress in astrocytes via non-cell-autonomous mechanisms, we next investigated whether these effects are recapitulated in vivo and whether Sigma-1R modulates this astrocytic response. To this end, we employed an AAV9-hSyn-poly-PR system to selectively express poly-PR in neurons of Sigma-1R wild-type (Supplementary Fig. 5) and knockout mice, thereby modeling C9orf72-associated ALS pathology. Using this approach, we successfully induced astrocyte activation in the poly-PR-driven C9orf72 ALS mouse model (Supplementary Fig. 6). Consistent with our cell-based observations, immunofluorescence analysis revealed pronounced astrogliosis in poly-PR-injected brains, as evidenced by elevated GFAP expression in the motor cortex (Supplementary Fig. 6). Notably, this astrocytic activation was significantly intensified in Sigma-1R knockout mice compared to wild-type controls (Fig. 8A and B). In parallel, NOX4 expression, previously identified as a key mediator of oxidative stress in astrocytes, was markedly upregulated in Sigma-1R-deficient mice following poly-PR expression (Fig. 8C and D). Importantly, co-localization analyses confirmed that NOX4 upregulation occurred predominantly in GFAP-positive astrocytes. These results provide in vivo validation of our mechanistic model and suggest that Sigma-1R acts as a crucial regulator that suppresses poly-PR-induced astrocytic reactivity and NOX4-driven oxidative stress. Collectively, our findings highlight the protective role of Sigma-1R in modulating non-cell-autonomous neuroinflammatory responses in C9orf72 ALS.

Fig. 8.

Neuronal poly-PR expression induces greater astrocyte activation and NOX4 upregulation in SigmaR1^−/− mice compared to SigmaR1^+/+ mice. (A) Immunofluorescence staining showing increased astrocyte activation, as indicated by GFAP-positive cells, in SigmaR1^−/− mice compared to SigmaR1^+/+ mice following EGFP-PR₄₂ expression. All confocal images were acquired under identical settings, and uniform brightness/contrast adjustments were applied. (B) Quantification of GFAP-positive astrocytes in EGFP-PR₄₂-injected SigmaR1^−/− mice (n = 4 per group) revealed a significant increase compared to SigmaR1^+/+ mice. Cell numbers were quantified using the Cell Detection function in QuPath software. Data are presented as mean ± SEM and analyzed using an unpaired two-tailed t-test (∗∗p < 0.01). (C) Immunofluorescence analysis showing intensity of NOX4 with GFAP-positive astrocytes. The number of colocalized signals was significantly higher in SigmaR1^−/− mice compared to SigmaR1^+/+ mice. (D) Quantification of NOX4 intensity in EGFP-PR₄₂-injected SigmaR1^−/− mice (n = 4 per group) confirmed a significant increase relative to SigmaR1^+/+ controls. NOX4 intensities were obtained using the Cell Detection function in QuPath software. Data are presented as mean ± SEM and analyzed using an unpaired two-tailed t-test (∗p < 0.05).

4. Discussion

This study demonstrated that exogenous poly-PR dipeptide treatment induces NOX4 expression and H₂O₂ production in astrocytes via the IKK/IκB/NF-κB p65 signaling pathway. Notably, overexpression of Sigma-1R or treatment with clemastine, a clinically approved Sigma-1R agonist, suppressed NOX4 expression and H₂O₂ production by interacting with p65 and preventing its nuclear translocation and binding to the NOX4 promoter. In vivo, Sigma-1R deficiency exacerbated poly-PR-induced astrocyte activation, upregulated NOX4 expression, and further reduced survival, highlighting its neuroprotective role. Collectively, these findings suggest that exogenous poly-PR promotes astrocyte-mediated neuroinflammation and oxidative stress through non-cell-autonomous mechanisms, and that Sigma-1R exerts protective effects by modulating this pathological response (Fig. 9).

Fig. 9.

Proposed model of Sigma-1R-mediated protection against exogenous poly-PR-induced oxidative stress in astrocytes. Neuron-derived poly-PR is released and internalized by astrocytes, where it activates the IKK/IκB/NF-κB pathway, driving p65 nuclear translocation, NOX4 transcription, and H₂O₂ production. Sigma-1R overexpression interacts with p65, limiting its nuclear translocation and binding to the NOX4 promoter, thereby suppressing astrocytic oxidative stress and neuroinflammation. This model underscores the non-cell-autonomous neuron-to-astrocyte mechanism contributing to C9orf72 ALS pathogenesis.

Poly-PR propagation and its engagement with glial cells represent critical aspects of C9orf72-associated pathogenesis. Previous studies suggest that the pathogenic effects of poly-PR extend beyond the neurons in which it is produced, exerting deleterious effects on neighboring glial cells and neurons via non-cell-autonomous mechanisms [8,9]. In our study, we demonstrated that poly-PR, when expressed in either neuronal cells or astrocytes, is secreted into the extracellular space and subsequently internalized by naïve astrocytes (Supplementary Fig. 2), consistent with previous observations in cellular models [8,9]. Notably, poly-PR has been shown to propagate through exosome-independent mechanisms from poly-PR-expressing neurons to recipient neurons [8]. However, the detailed transmission and uptake mechanisms of poly-PR in astrocytes remain unclear. We speculate that poly-PR entry into astrocytes likely occurs via endocytic pathways, a mechanism previously implicated in the intercellular spread of toxic proteins such as α-synuclein, SOD1, and TDP-43 in various neurodegenerative diseases [30,31]. Although the precise mode of internalization has yet to be elucidated, our data using FITC-conjugated poly-PR₂₀ confirmed robust uptake and nuclear accumulation in astrocytes (Fig. 3C and D).

Consistent with previous reports [32], EGFP-poly-PR with 20∼60 repeats predominantly accumulates in the nucleus and nucleolus (Supplementary Fig. 7), which may explain why overexpression does not strongly activate cytoplasmic signaling cascades such as IKK/IκB/NF-κB. In contrast, longer poly-PR repeats or other arginine-rich DPRs, such as poly-GR (Supplementary Fig. 7), show greater cytosolic localization and may more directly engage NF-κB signaling. Whether overexpression of long-repeat poly-PR or poly-GR in astrocytes can activate the IKK/IκB/NF-κB/NOX4 axis remains to be determined. In line with this, our results demonstrate that poly-PR promotes p65 phosphorylation, yet Sigma-1R overexpression attenuates p65 nuclear accumulation without affecting its phosphorylation (Fig. 6). These findings suggest that Sigma-1R regulates p65 at a post-activation step, most likely within the cytoplasmic or perinuclear compartment, thereby limiting nuclear import and NOX4 promoter occupancy.

Notably, immunofluorescence analysis further showed that Sigma-1R overexpression reduced p65 nuclear translocation in astrocytes exposed to exogenous poly-PR peptides (Fig. 6H). This observation highlights an important question regarding the relationship between Sigma-1R and nucleocytoplasmic transport. Previous studies have reported that Sigma-1R stabilizes nucleoporin expression, particularly Pom121, thereby promoting nucleocytoplasmic transport in NSC34 motoneuron-like cells [2,18,21,22]. In contrast, in astrocytes, Sigma-1R overexpression restricted p65 nuclear entry, suggesting a cell-type-specific effect. Supporting this idea, a prior study showed that Pom121 can limit NF-κB p65 nuclear translocation under inflammatory stimulation in macrophages [33]. We therefore speculate that either cell-type differences or distinct stress contexts determine how Sigma-1R regulates nucleocytoplasmic transport and p65 translocation. This hypothesis is consistent with Sigma-1R's established roles at ER-associated membranes and with evidence that nucleoporin pathway components can restrict NF-κB nuclear import under inflammatory stress. Future studies using compartment-resolved assays will be required to determine whether Sigma-1R also modulates p65 function once inside the nucleus.

Importantly, exogenous poly-PR triggered a pronounced proinflammatory response in astrocytes through activation of the IKK/IκB/NF-κB pathway (Fig. 9). Phosphorylation of the NF-κB subunit p65 represents a pivotal regulatory step that facilitates its nuclear translocation and transcriptional activity [34,35], thereby linking upstream IKK activation to the expression of oxidative and inflammatory mediators. This modification provides the mechanistic basis by which exogenous poly-PR exposure enhances NOX4 transcription and ROS production in astrocytes. The precise mechanism by which poly-PR activates IKK remains unresolved, but two non-mutually exclusive hypotheses may be considered. First, exogenous poly-PR, upon uptake through endocytosis, could engage membrane-associated or endosomal stress sensors that converge on IKK signaling. Second, extracellular peptide exposure may impose acute stress or redox imbalance that activates upstream kinases, whereas plasmid-driven overexpression predominantly results in nuclear or nucleolar accumulation with limited engagement of cytoplasmic signaling cascades [32]. Although the exact molecular link awaits clarification, our findings clearly establish that exogenous poly-PR enhances p65 phosphorylation and nuclear translocation, leading to NOX4 induction and H₂O₂ production (Fig. 3). This mode of delivery thus critically shapes NF-κB activation and reinforces the role of astrocytic oxidative stress in C9orf72 ALS pathogenesis.

Additionally, the pathogenic actions of poly-PR are not confined to astrocytes. Previous studies have shown that poly-PR disrupts nucleocytoplasmic transport and promotes stress granule assembly in neurons [2,4,7], and it may similarly contribute to microglial activation. The widespread cellular vulnerability to poly-PR underscores the importance of targeting its downstream effectors to mitigate neuroinflammation and oxidative stress in C9orf72-associated ALS.

Neuroinflammation is tightly linked to increased production of ROS [10,11,36]. Oxidative stress is a common feature of neurodegenerative diseases, particularly ALS [[37], [38], [39], [40], [41]]. ROS can be generated through various mechanisms, including activation of NADPH oxidases and endoplasmic reticulum (ER) stress [42,43]. In addition to these sources, ROS production arises from multiple converging pathways. Mitochondrial dysfunction, a hallmark of neurodegenerative conditions, contributes substantially to ROS accumulation through impaired oxidative phosphorylation and electron leakage from the respiratory chain [39,44,45]. Abnormal iron metabolism is another critical contributor, as ferrous iron catalyzes the Fenton reaction, converting H₂O₂ into highly reactive hydroxyl radicals [46,47]. These overlapping sources of ROS not only sustain oxidative stress but also intensify neuroinflammatory responses, thereby accelerating disease progression in ALS and related neurodegenerative disorders. Taken together, our findings highlight that among the multiple sources of ROS, NADPH oxidase-derived ROS, especially via astrocytic NOX4, plays a pivotal role in mediating neuronal damage in response to non-cell-autonomous poly-PR. Mechanistically, exogenous poly-PR activates the IKK/IκB/NF-κB p65 signaling cascade, leading to enhanced NOX4 transcription and subsequent H₂O₂ production in astrocytes.

An important unresolved question raised by our findings is why NOX4 activation is observed following exogenous poly-PR peptide treatment but not with plasmid-driven poly-PR overexpression in astrocytes. Several possibilities may account for this discrepancy. First, intracellular localization and processing may differ between the two delivery methods. Plasmid-driven expression often results in nuclear or nucleolar accumulation [32], whereas exogenous peptides are efficiently internalized into the cytoplasm and nucleus, thereby engaging cytoplasmic signaling pathways such as IKK/NF-κB/NOX4 more robustly. Second, concentration and exposure kinetics may also play a role. Exogenous peptide application provides a rapid and high local concentration of poly-PR, potentially surpassing the threshold required for NOX4 induction, while gradual accumulation from plasmid expression may remain below this threshold. Third, the non-cell-autonomous nature of exogenous peptide uptake may better reflect neuron-to-astrocyte transmission in C9orf72 ALS, whereas direct overexpression in astrocytes does not mimic this disease-relevant communication. Collectively, these considerations suggest that the mode of poly-PR delivery critically shapes downstream NF-κB activation and subsequent NOX4 induction.

In addition to clemastine, other Sigma-1R agonists, including pridopidine and fluvoxamine, have been reported to exert beneficial effects in ALS models, further supporting the therapeutic relevance of targeting Sigma-1R [18,21,[48], [49], [50]]. These agents primarily act on neuronal cells; however, whether Sigma-1R agonists can attenuate glial activation, particularly by suppressing oxidative stress, in the context of C9orf72-associated ALS remains unclear. In our study, we demonstrated that clemastine activates Sigma-1R, as evidenced by BiP-Sigma-1R dissociation, and suppresses NOX4 expression and H₂O₂ production by inhibiting p65 binding to the NOX4 promoter regions (Fig. 5, Fig. 6F). These findings not only expand the understanding of clemastine's mechanism of action beyond neurons but also provide the first direct evidence that pharmacological activation of Sigma-1R can mitigate astrocyte-mediated oxidative stress in C9orf72 ALS. Future studies are warranted to determine whether other Sigma-1R agonists exert similar glial-modulatory effects and to evaluate their therapeutic potential in broader neuroinflammatory contexts. According to previous studies, the antihistaminergic agent clemastine protects motor neurons and suppresses inflammation in SOD1-G93A mice, primarily by inhibiting microglial activation [28,51]. Our findings reveal an additional mechanism whereby clemastine exerts its protective effects: activation of Sigma-1R to reduce astrocytic activation and oxidative stress. Supporting this, the selective Sigma-1R antagonist NE100 abolished the ability of clemastine to suppress poly-PR-induced NOX4 expression (Supplementary Fig. 8). These results indicate that clemastine acts primarily through Sigma-1R activation rather than histamine receptor blockade, although genetic approaches such as siRNA-mediated knockdown will be valuable for further confirmation. Taken together, these dual actions suggest that clemastine may serve as a promising multi-target therapeutic agent for ALS, capable of modulating both neuronal and glial components of the disease.

Despite the compelling evidence presented in this study, several limitations should be considered. First, although our findings indicate that clemastine activates Sigma-1R and suppresses astrocyte-mediated oxidative stress, the extent to which these effects are specifically mediated by Sigma-1R, as opposed to its antihistaminergic properties, remains to be fully clarified. While we employed functional assays to support Sigma-1R involvement, the use of genetic overexpression would further strengthen the specificity of this pathway. Second, the mouse model used in this study selectively expresses poly-PR in neurons, which may not fully recapitulate the complex spatial and temporal dynamics of poly-PR propagation and glial involvement observed in human C9orf72 ALS. Although our AAV9-hSyn construct was validated as neuron-specific and no direct expression of poly-PR₄₂ was detected in GFAP-positive astrocytes, we cannot fully exclude the possibility that poly-PR₄₂ is transferred to astrocytes through intercellular propagation. Third, our investigation focused primarily on astrocyte responses, whereas contributions from other glial populations, such as microglia and oligodendrocytes, remain to be elucidated. Fourth, our cellular experiments utilized commercially available human primary astrocytes rather than patient-derived induced pluripotent stem cell (iPSC)-derived astrocytes. While this approach offers experimental reproducibility and accessibility, it lacks the genetic background and pathological relevance of C9orf72 mutation carriers. The use of iPSC-derived astrocytes would provide valuable insights, but is constrained by limited availability, high cost, and technical challenges. Fifth, we cannot exclude the potential involvement of other DPRs, such as poly-GR and poly-GA, in mediating astrocyte activation and oxidative stress. Additionally, the mechanism by which poly-PR penetrates astrocytes, whether through passive diffusion, endocytosis, or other pathways, requires further investigation. Finally, although clemastine showed therapeutic promise in our experimental models, additional studies are necessary to evaluate its long-term efficacy, blood-brain barrier permeability, and translational potential across different ALS subtypes. Addressing these limitations in future research will be essential to deepen our mechanistic understanding and to inform the development of effective Sigma-1R-based therapies for C9orf72 ALS.

5. Conclusions

This study uncovers a previously unrecognized glial-mediated mechanism by which exogenous poly-PR exacerbates C9orf72 ALS pathology through non-cell-autonomous activation of the IKK/IκB/NF-κB signaling cascade in astrocytes, culminating in NOX4-driven H₂O₂ production and oxidative stress. We demonstrate that Sigma-1R, a multifunctional ER chaperone, exerts potent protective effects by disrupting p65-driven NOX4 transcription without altering upstream phosphorylation events. Importantly, pharmacological activation of Sigma-1R using clemastine, a clinically approved compound, recapitulates the antioxidant and anti-inflammatory effects of Sigma-1R overexpression, supporting its translational potential. In vivo, Sigma-1R deficiency markedly worsened poly-PR-induced neurodegeneration, astrogliosis, and NOX4 expression, reinforcing its critical role in modulating glial reactivity. Collectively, our findings establish Sigma-1R as a pivotal regulator of astrocyte-mediated oxidative stress in C9orf72 ALS and identify clemastine as a promising repurposable therapeutic agent. This work not only advances mechanistic understanding of non-neuronal contributions to C9orf72 pathology but also lays the groundwork for glia-targeted therapeutic strategies aimed at halting ALS progression (Fig. 9).

CRediT authorship contribution statement

Hsuan‐Cheng Wu: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Software, Validation. Teng-Wei Huang: Funding acquisition, Writing – original draft, Writing – review & editing. Eddie Feng‐Ju Weng: Conceptualization, Investigation, Methodology, Resources, Validation. Chun-Yu Lin: Writing – review & editing. Tsung‐Ping Su: Conceptualization, Resources, Visualization. Hsiang‐En Wu: Conceptualization, Methodology. Shao‐Ming Wang: Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing – original draft, Writing – review & editing.

Ethics approval and consent to participate

All animal procedures were performed in compliance with the ethical guidelines and regulations approved by the Institutional Animal Care and Use Committee (IACUC) of China Medical University (Protocol No. CMUIACUC-2025-007) and adhered to the Guide for the Care and Use of Laboratory Animals. Clinical trial registration: not applicable.

Consent for publication

Not applicable.

Availability of data and materials

Data supporting the findings of this study are available from the corresponding author upon reasonable request.

Declaration of generative AI and AI-assisted technologies

During the preparation of this work, the authors used ChatGPT for language editing. The authors subsequently reviewed and revised the content as necessary and take full responsibility for the final version of the publication.

Funding

Funding was obtained from the National Science and Technology Council of Taiwan (grant number MOST 111-2628-B-039- 006-MY3 and NSTC 114-2320-B-039 -064 -MY3) and China Medical University, Taiwan (grant number CMU113-MF-06).

Declaration of competing interest

The authors declare that they have no competing interests.

Abbreviations

C9orf72

Chromosome 9 open reading frame 72

ALS

amyotrophic lateral sclerosis

FTD

frontotemporal dementia

DPRs

dipeptide repeat proteins

PR

proline–arginine

GR

glycine–arginine

GA

glycine–alanine

AAV

adeno-associated virus

hSyn

human synapsin I promoter

NOX

NADPH oxidase

ROS

reactive oxygen species

Sigma-1R

Sigma-1 receptor

H₂O₂

hydrogen peroxide

CCK-8

Cell Counting Kit-8

DAPI

4′,6-diamidino-2-phenylindole

iPSC

induced pluripotent stem cell

Appendix A. Supplementary data

The following are the Supplementary data to this article.