NOX4 Regulates NLRP3 by Inhibiting the Ubiquitination of LRRC8A to Promote Ferroptosis in Nucleus Pulposus Cells

Abstract

Intervertebral disc degeneration (IDD) is a significant contributor to low back pain, imposing a considerable socioeconomic burden. Ferroptosis, a novel form of cell death driven by iron and characterized by the accumulation of reactive oxygen species (ROS), has been associated with the progression of IDD. Nicotinamide adenine dinucleotide phosphate oxidase 4 (NOX4) has been widely recognized as a pivotal factor promoting ferroptosis across various diseases; however, its precise role in the pathogenesis of IDD remains incompletely understood. Our experimental findings demonstrated a marked upregulation of NOX4 in degenerated cells, accompanied by elevated ROS levels and a diminished mitochondrial membrane potential, indicating the participation of ferroptosis. Furthermore, the expression of the critical regulatory factor GPX4 was reduced, while ACSL4 levels were significantly increased, further corroborating the involvement of ferroptosis. Functional loss and gain experiments revealed that NOX4 overexpression augmented ferroptosis and ROS production while promoting the secretion of inflammatory cytokines. Subsequent studies indicated that the knockdown of NOX4 could reverse tert-butyl hydroperoxide (TBHP)-induced ferroptosis. Mass spectrometry analysis identified leucine-rich repeat-containing 8A (LRRC8A) as an interacting protein of NOX4, and further validation confirmed that they co-regulate Nod-like receptor pyrin domain-3 (NLRP3) activation through their interaction. Utilizing a rat model of intervertebral disc degeneration, we further corroborated the role of NOX4 in IDD. This study provides theoretical support for the potential application of NOX4-targeting drugs in the treatment of IDD.

Introduction

Intervertebral disc degeneration (IDD) is the most prevalent degenerative disorder of the musculoskeletal system, resulting in significant morbidity and mortality for patients, as well as substantial financial burdens on families and society [1]. As is well known, the intervertebral disc (IVD) comprises three anatomical sub-structures that are essential for its normal function: the nucleus pulposus (NP), the annulus fibrosus (AF), and the cartilaginous endplates (CEP) [2]. The primary component of the NP is nucleus pulposus cells (NPCs), which play a crucial role in maintaining the integrity of the IVD through the production of the extracellular matrix (ECM) [3]. Due to the lack of a direct blood supply, NPCs exist in a low-nutrient and hypoxic microenvironment, rendering them susceptible to inflammatory factors and the accumulation of reactive oxygen species (ROS) [4]. Moreover, the disordered synthesis of the ECM and the hypocellularity of the NPCs further contribute to the structural degradation and biomechanical instability of the IVD [5], and thereby aggravate IDD.

Regulated cell death (RCD), encompassing necrosis, apoptosis, pyroptosis, and ferroptosis, plays a crucial role in degenerative diseases. Ferroptosis, a novel form of RCD characterized by iron-dependent lipid peroxidation, has been linked to various pathological conditions [6, 7], including IDD [8–10]. This process involves the accumulation of iron and the subsequent production of ROS, leading to oxidative stress and cellular damage. Notably, ROS-mediated lipid peroxidation is a significant step in the progression of ferroptosis [11]. Nicotinamide adenine dinucleotide phosphate(NADPH)Oxidase4(NOX4), one of the most commonly expressed members of the NOX family, has been identified as a critical regulator of ferroptosis in various neurodegenerative diseases, such as intracerebral hemorrhage [12], Parkinson’s disease [13], and Alzheimer’s disease [14]. Current, research on the role of NOX4 in IDD primarily focuses on animal models. Feng et al. demonstrated that NOX4 expression is significantly upregulated in degenerated NPCs of rats, identifying it as a major source of ROS in these cells [15]. Furthermore, their studies on aging rat degenerative discs revealed that NOX4 regulates NPCs senescence, providing novel insights into the mechanisms of IDD and potential therapeutic targets [16]. In human NPCs, only Ma et al. reported that hydrogen peroxide (H₂O₂) upregulates NOX4 expression, which reverses the antioxidative, anti-inflammatory, and matrix-protective effects of anemonin [17]. However, there are currently no studies investigating the role of NOX4 in regulating ferroptosis in NPCs during IDD. Our previous bioinformatics analysis and verification suggest that NOX4 may act as a promoting factor for ferroptosis in IDD [18].

NOX4 can regulate intracellular ROS levels, thereby participating in the occurrence of ferroptosis [19]. However, ROS has also been identified as a key factor in activating the nod-like receptor pyrindomain-3 (NLRP3) inflammasome [20]. NLRP3 serves as a crucial element in the human immune system, significantly contributing to inflammation and cellular death in response to disease and stress [21]. The NLRP3 inflammasome consists of NLRP3, apoptosis-associated speck-like protein containing a CARD (ASC), and caspase-1. Notably, NLRP3 functions as a receptor that signals to the downstream effector protein caspase-1, thereby regulating chronic inflammation [22]. Recent research indicates that the NLRP3 inflammasome is predominantly triggered in IDD, with the subsequent inflammatory mediators playing a critical role in the progression of IDD. The activated NLRP3 inflammasome primarily participates in the inflammatory response, induces pyroptosis in cells, contributes to the degradation of the ECM, and facilitates the apoptosis of IVD cells [23, 24]. Therefore, controlling ROS/NLRP3-induced ferroptosis may represent a promising strategy to prevent IDD. Nonetheless, the precise mechanism by which NOX4 promotes ferroptosis in NPCs remains to be fully elucidated.

MicroRNAs (miRNAs) are single-stranded, non-coding RNA molecules approximately 22 nucleotides in length. When the miRNAs sequence is fully complementary to the target gene, it cleaves the target mRNA, leading to its degradation(22). Conversely, if the sequence is only partially complementary, the translation of the target gene is inhibited without affecting mRNA stability(23). The expression levels of miRNAs change significantly in degenerative tissues, with notable differences between normal and degenerated NP tissues(24). Studies have demonstrated that miRNAs regulate key biological processes in IDD, including cell proliferation, apoptosis, cytokine release, and ECM synthesis and degradation(25–27). We hypothesize that miRNAs may serve as upstream regulators of NOX4, and this study aims to identify novel miRNAs that directly bind to NOX4 in the context of IDD.

In this study, we performed a bioinformatics analysis to identify key molecules involved in IDD. Our findings imply that NOX4 is a crucial enzyme responsible for ROS generation and is significantly upregulated in IDD. Furthermore, NOX4 contains sequences that correspond to the seed region of miR-6843-3p. We subsequently investigated the role and underlying mechanisms of NOX4 in IDD through both in vitro and in vivo experiments.

Materials and Methods

Human NPCs Culture

The research was supervised and approved by the Ethics Committee of Fuyang City People's Hospital (ethical approval no. [2024]174). Informed consent was obtained from all donors prior to their surgical procedures. We procured degenerative NP tissues from a total of 15 patients diagnosed with lumbar disc herniation, lumbar spinal stenosis, or lumbar spondylolisthesis, all of whom underwent percutaneous endoscopic lumbar discectomy. Normal NP tissues ware obtained from 10 idiopathic scoliosis patients requiring surgical intervention. Patients with spinal magnetic resonance imaging Pfirrmann grades I and II were classified as the normal group, while those with grades III and IV were categorized as having mild degeneration. Patients with grade V were classified as having severe degeneration.

Human NPCs were obtained following a previously described protocol [25]. Briefly, small pieces of NP tissues were digested with 0.25 mg/ml type II collagenase (Invitrogen) for 5 h. After washing with PBS and centrifugation, the isolated NPCs were cultured in Dulbecco’s modified Eagle medium (DMEM)/F12 (Gibco), enriched with 15% fetal bovine serum (Gibco) and 1% penicillin–streptomycin (Invitrogen). The identification of NPCs was performed using specific markers (CD24, KRT18, Abcam), and cells at passage two were used for further experiments. In vitro, NPCs were exposed to tert-butyl hydroperoxide (TBHP) (Sigma) at concentrations of 0, 12.5, 25, 50, and 100 μM for either 12 or 24 h.

Evaluation of Cellular Viability and Proliferation

Cell viability was assessed using the Cell Counting Kit-8 (CCK-8) (C0037, Beyotime, Shanghai, China) in accordance with the manufacturer’s protocols. Absorbance measurements were recorded at a wavelength of 450 nm using a spectrophotometer. The proliferation of NPCs was evaluated through the incorporation of EdU (5-ethynyl-2′-deoxyuridine) (C0078S; Beyotime, Shanghai, China), and fluorescence images were captured with a fluorescence microscope (Zeiss, LSM 800, Germany).

Mitochondrial Membrane Potential and Apoptosis Detection

The cells were initially placed in a 24-well plate. Following the induction of apoptosis, they were centrifuged at 1000 rpm for 5 min. The medium was subsequently discarded, and the cells were washed with PBS. Next, a sequence of 188 µl of Annexin V-FITC binding solution, 5 µl of Annexin V-FITC, 2 µl of Mito-Tracker Red CMXRos, and 5 µl of Hoechst 33,342 was added, with gentle mixing after the introduction of each compound. The cells were then incubated at room temperature in the dark for 20 to 30 min, followed by treatment in an ice bath. Fluorescence microscopy was employed to observe the fluorescence emitted by Mito-Tracker Red CMXRos (red), Hoechst 33,342 (blue), and Annexin V-FITC (green).

In the experiment, DCFH-DA was initially diluted in serum-free medium at a ratio of 1:1000, resulting in a final concentration of 10 μmol/L. The cell culture medium was then discarded, and an adequate quantity of the diluted probe solution was added to ensure complete coverage of the cell surface. After incubating the cells at 37 °C for 20 min, they were washed three times with serum-free medium to remove any unincorporated DCFH-DA. Following these washes, the cells were stimulated for 20–30 min, and the levels of ROS were assessed using confocal microscopy. At this stage, a significant increase in ROS levels should be evident in the ROS-positive control group.

Western Blotting Analysis

Cells were subjected to protein extraction using RIPA buffer containing both protease and phosphatase inhibitors, while the concentration was assessed via a BCA assay. Protein samples, ranging from 20 to 40 µg, were denatured and then separated on SDS-PAGE gels composed of 8–12% acrylamide, followed by transfer to PVDF membranes. These membranes were blocked for 1 h at room temperature in a 5% non-fat milk solution in TBS-T, after which they were incubated overnight at 4 °C with primary antibodies targeting NOX4, GPX4, ACSL4, NLRP3, transferrin (TF), transferrin receptor (TFR), leucine-rich repeat containing 8A (LRRC8A), and β-Actin (Proteintech, Wuhan, China). Following washing in TBS-T, membranes were incubated with HRP-conjugated secondary antibodies for 1 h at room temperature. Protein bands were detected using ECL detection and captured with a Bio-Rad ChemiDoc system. Band intensity quantification was performed using ImageJ software, with protein levels normalized relative to β-Actin.

Quantitative Reverse Transcription Polymerase Chain Reaction (QRT-PCR)

Total RNA was obtained from cultured cells utilizing TRIzol reagent (Invitrogen) in accordance with the manufacturer's guidelines. The sequence of primers for each gene can be found in Table 1, with β-Actin utilized for normalization purposes.

Table 1.

Sequences of the QRT-PCR primers used for each gene

Gene	Forward primer sequence	Reverse primer sequence
GPX4	GAGGCAAGACCGAAGTAAACTAC	CCGAACTGGTTACACGGGAA
ACSL4	CATCCCTGGAGCAGATACTCT	TCACTTAGGATTTCCCTGGTCC
NOX4	CAGATGTTGGGGCTAGGATTG	GAGTGTTCGGCACATGGGTA
β-Actin	CTCGCCTTTGCCGATCC	GAATCCTTCTGACCCATGCC

Immunohistochemical Examination

The sections underwent defatting, rehydration, and were microwaved for 15 min to eliminate antigens. Subsequently, the activity of endogenous peroxidase was inhibited using 3% hydrogen peroxide for 10 min, and nonspecific binding sites were blocked with 5% bovine serum albumin for 30 min at room temperature. The sections were then incubated overnight at 4 °C with primary antibodies (NOX4, GPX4, ACSL4; Proteintech, Wuhan, China). Following this, appropriate HRP-conjugated secondary antibodies were applied, and the sections were counterstained with hematoxylin.

Detection of Reduced Glutathione (GSH)

To measure reduced GSH, prepare a 10 mM oxidized glutathione (GSSG) stock solution by dissolving 5 mg of GSSG in 816 µl Milli-Q water. Create a DTNB stock solution by dissolving 4.5 mg DTNB in 1.5 ml DMSO. For protein removal, mix 0.2 g of Protein Removal Reagent M with 4 ml assay buffer to prepare a 5% solution. Prepare a 40 mg/ml NADPH stock solution by dissolving 4 mg NADPH in 100 µl Milli-Q water. Prepare a 5 × diluted glutathione reductase solution by mixing 50 µl enzyme with 200 µl assay buffer. Adjust the total glutathione assay working solution based on sample number. Prepare 0.5 mg/ml NADPH by mixing 10 µl of the stock with 790 µl assay buffer. Dilute GSH removal assist solution and GSH removal reagent working solution as described. For standards, create GSSG solutions of varying concentrations and process them with diluted GSH removal assist solution and GSH removal reagent. Prepare cell samples by lysing in Protein Removal Reagent M solution, followed by freeze–thaw and centrifugation. Process samples similarly to standards. Measure absorbance at A412 using a 96-well plate, adding NADPH and incubating. Calculate GSH content from ΔA412/min using a standard curve.

Plasmids and Transfection

NOX4 was either overexpressed or silenced in human NPCs by the transfection of lentiviral plasmids (Genomeditech, Shanghai, China), in accordance with the manufacturer's guidelines. These lentiviral plasmids were converted into viral particles and introduced into NPCs with the assistance of 8 μg/mL polybrene. The transfection process was conducted at 37 °C in an atmosphere containing 5% CO₂ for a duration of 24 to 48 h, after which selection was performed using 2 μg/mL puromycin for a period of 7 to 10 days to confirm stable integration. The effectiveness of NOX4 overexpression and knockdown was verified through QRT-PCR and Western blotting, which measured both mRNA and protein levels, respectively. The cells that were successfully transfected were subsequently utilized to investigate the effects of NOX4 modulation on NPC functionality and related signaling pathways.

Co‐immunoprecipitation Assay

Collect NPCs during the logarithmic growth phase, then add 1 ml of RIPA lysis buffer and incubate on ice for 30 min. Following this, centrifuge the mixture to obtain the supernatant. Next, add agarose beads and NOX4 antibody, and shake the mixture thoroughly overnight at 4 °C. Afterward, collect the agarose beads, add SDS loading buffer, and boil for 10 min. The gel was subsequently stained with Coomassie brilliant blue, and NOX4-interacting protein molecules were detected using mass spectrometry analysis.

Ubiquitination Analysis

First, treat the cells with the proteasome inhibitor MG132 for 24 h in the cell culture medium to inhibit protein degradation. After lysing the cells, determine the protein concentration to ensure sample consistency. Next, incubate the lysate with either anti-ubiquitin antibody or the target protein antibody for immunoprecipitation to enrich the ubiquitinated proteins. Subsequently, use protein A/G magnetic beads to capture the antibody-protein complexes. Finally, separate the samples by SDS-PAGE, transfer to a membrane, and detect using ubiquitin-specific antibody(Anti-Ubiquitin Mouse mAb, PTM-1107). Visualization of the ubiquitination patterns can be achieved using chemiluminescence or fluorescence methods.

Mass Spectrometry Analysis

NPCs samples were thoroughly washed with pre-cooled PBS to remove impurities, followed by lysis using a buffer containing protease inhibitors. The samples were sonicated on ice to release proteins and centrifuged at 12,000 × g for 15 min at 4 °C to remove debris. The supernatants were collected, and protein concentrations were measured using a BCA assay kit, with all samples adjusted to the same concentration. Protein reduction was performed with DTT at 37 °C for 1 h, followed by alkylation with IAA at room temperature in the dark for 30 min. The proteins were then digested with trypsin (enzyme-to-protein ratio of 1:50) at 37 °C overnight. The resulting peptides were enriched and desalted using C18 solid-phase extraction columns, dissolved in 0.1% formic acid, and stored at −80 °C.

Peptide samples were separated using ultra-high-performance liquid chromatography (UPLC) and analyzed with a tandem mass spectrometer. Full scan (MS1) and fragmentation scan (MS2) data were recorded, with a mass range of 350–1,500 m/z and resolution parameters optimized for high-precision detection. Data analysis was performed using Mascot or MaxQuant software to match mass spectrometry data with protein databases, identifying peptides and corresponding proteins. Functional annotation and pathway enrichment analyses were conducted using gene ontology (GO) and kyotoencyclopedia of genes and genomes(KEGG) databases. Finally, statistically significant candidate proteins were screened for subsequent validation.

ELISA Assay

The NPCs culture supernatant was stored in a −80 °C refrigerator until measurement. R&D Systems offers ELISA kits (Beyotime,P1305; Beyotime,PT518) for detecting IL-1β and TNF-α levels. Measurements should be conducted following the manufacturer's guidelines.

Immunofluorescence

NPCs were treated with 4% paraformaldehyde for fixation, followed by permeabilization using 0.1% Triton X-100 and blocking with 5% BSA. Specific primary antibodies were then applied and incubated overnight at 4 °C. Upon washing, cells received fluorophore-conjugated secondary antibodies and were counterstained with DAPI. Coverslips were mounted with Fluoromount-G, and fluorescence images were acquired utilizing a confocal microscope. For the analysis of protein localization, ImageJ software was employed.

Construction and Treatment of Rat IDD Model

Forty male Sprague–Dawley rats (3 months old, 250 ± 30 g) were anesthetized with 2% pentobarbital and randomly divided into four groups: control, empty vector, NOX4 knockdown (NOX4 Sh), and NOX4 overexpression. The Co7/8 intervertebral disc was located by palpation and verified via X-ray. With the exception of the control group, the disc was punctured with an 18G needle, which was inserted to a depth of 5 mm, rotated 360 degrees, and maintained in position for 30 s to promote degeneration.The control group received no puncture. Treatments were administered at weeks 4 and 8 post-puncture, with 2 µL of PBS, empty vector adenovirus, NOX4 knockdown adenovirus, or NOX4 overexpression adenovirus injected into the disc using a 29G needle. The approach and volume for intradiscal injection were established based on earlier research [26]. The Ethics Committee for Animal Care at Anhui Medical University reviewed and approved all animal experiments conducted.

Magnetic Resonance Imaging (MRI)

MRI assessment took place 12 weeks post-surgery. The evaluation of signal and structural alterations on sagittal T2-weighted images (T2WI) utilized a 3.0 T clinical magnet (Philips Intera Achieva 3.0 MR). Blinded orthopedic investigators conducted the MRI evaluation employing the Pfirrmann MRI scoring system [27].

Histological Staining and Scoring Methods

In this study, tissue specimens underwent staining with hematoxylin–eosin (HE), Masson’s trichrome, and Alcian blue, followed by histological scoring to assess the degree of degeneration. Sections were first deparaffinized through three rounds of xylene and ethanol treatments, followed by rinsing with running water. HE staining involved hematoxylin staining, hydrochloric acid alcohol differentiation, lithium carbonate bluing, and eosin staining, followed by dehydration and clearing with ethanol and xylene, and finally mounting with neutral resin. Masson’s staining used Weigert’s iron hematoxylin, Ponceau-Fuchsin, and aniline blue, with intermediate washing in phosphomolybdic acid solution. After Alcian blue staining, sections were dehydrated, cleared, and mounted. Histological scoring was based on the morphology and cellular structure of the NP and AF, as well as the integrity of the NP-AF boundary. The scoring system varied from 0 to 15, where elevated scores signified greater degeneration [28].

Bioinformatics Analysis

To identify miRNAs that regulate NOX4, we employed bioinformatics tools including TargetScan, RNA22, MiRDB, and mirDIP to predict potential upstream miRNAs for NOX4. The predicted miRNAs from these four databases were analyzed using Venn2.1 to find the intersecting miRNAs, thus identifying key regulatory miRNAs upstream of NOX4.

Dual-Luciferase Reporter Assay

The wild-type or mutant 3’-UTR of NOX4 was constructed into dual luciferase reporters using the ZBFV-025 vector (Genomeditech, China), and 100 nM miR-6843-3p mimics or negative control (Genomeditech, China) were co-transfected with 2 μg wild-type or mutant luciferase reporter plasmids in NPCs. After 48 h, 200 µL of RIPA lysis buffer was utilized to lyse the cells in each well, and incubation was carried out for 10 min to achieve full lysis. The resulting lysates were then gathered and subjected to centrifugation, with the supernatants reserved for further analysis. In the luciferase assay, each sample received 100 µL of Firefly luciferase substrate, followed by mixing, and luminescence was recorded with a detection duration of 0.1 s. Afterward, 100 µL of Renilla luciferase substrate was introduced to the same sample, mixed once more, and luminescence was assessed again using the same detection duration.

Image Processing and Statistical Methods

All measurements were performed at least three times. Western blot, immunofluorescence, and immunohistochemistry analyses were conducted using Image J software, and statistical analysis and plotting were performed using GraphPad Prism 8. The comparative differences between the two groups were analyzed using t-test. Differences between multiple groups were compared using ANOVA. The results are expressed as mean ± standard deviation, with statistical significance marked by p < 0.05. The symbol * represents p < 0.05, ** signifies p < 0.01, and *** indicates p < 0.001.

Results

Ferroptosis Occurred in Degenerated NPCs Compared to Normal NPCs

To investigate the potential involvement of ferroptosis in IDD, we obtained NP tissues from individuals diagnosed with idiopathic scoliosis (normal group) and from those suffering from lumbar disc herniation (degenerative group). Mitochondrial membrane potential serves as a critical indicator of cellular viability and ferroptosis. In comparison to the normal group, the mitochondrial membrane potential in degenerative NPCs was significantly reduced, which was accompanied by an increase in apoptosis (Fig. 1A). Additionally, elevated levels of ROS, a hallmark of ferroptosis, were observed in the degenerative group (Fig. 1B).

Fig. 1.

Analysis of ferroptosis markers and cell viability in IDD.  A Mitochondrial membrane potential in NPCs was significantly reduced in the degenerative group, indicating decreased cell viability and increased ferroptosis, accompanied by elevated apoptosis. B ROS levels were markedly higher in the degenerative group, a key indicator of ferroptosis. C mRNA expression levels of ferroptosis regulators GPX4 and ACSL4, with reduced GPX4 and increased ACSL4 in the degenerative group compared to the normal group. D-E Western blot analysis confirmed corresponding protein level changes, with GPX4 downregulated and ACSL4 upregulated in the degenerative group. F Immunohistochemical staining showed that GPX4 expression decreased, while ACSL4 expression increased with advancing degeneration grade. NC: normal nucleus pulposus cells; IDD: Degraded nucleus pulposus cells; PI, PIII and PV: the grading of disc degeneration according to the Pfirrmann method.  *p< 0.05, **p < 0.01, ***p < 0.001

We conducted a comprehensive analysis of the expression levels of key regulators of ferroptosis, specifically GPX4 and ACSL4. Compared to the normal group, the mRNA expression of GPX4 was significantly reduced in the degenerative group, while ACSL4 exhibited a notable increase in expression within the degenerative group (Fig. 1C). Western blot analysis corroborated these differences at the protein level (Fig. 1D and E). Additionally, immunohistochemical analysis demonstrated that as the grade of degeneration increased, GPX4 expression decreased, whereas ACSL4 expression increased (Fig. 1F). These findings indicate the occurrence of ferroptosis in IDD, suggesting that ferroptosis may play a critical role in the pathological processes associated with disc degeneration.

TBHP Induces Ferroptosis in Normal NPCs and Upregulates NOX4 Expression in Degenerative NPCs

To establish an in vitro environment that simulates oxidative stress, NPCs were exposed to varying concentrations of TBHP to identify the most effective dose for inducing ferroptosis. Different doses of TBHP were administered to NPCs over 12- and 24-h periods, with cell viability assessed using the CCK8 assay. The IC50 value for TBHP was determined to be 39.42 μmol, leading to the selection of a concentration of 40 μmol for 24-h exposure in subsequent experiments (Fig. 2A).

Fig. 2.

NOX4 involvement in ferroptosis and oxidative stress in degenerative NPCs induced by TBHP. A Cell viability of NPCs treated with TBHP (0, 12.5, 25, 50, and 100 μmol) for 12 and 24 h, with 40 μmol selected for further experiments based on an IC50 of 39.42 μmol. B Western blot analysis shows TBHP-induced upregulation of NOX4 and ACSL4 and downregulation of GPX4 in a dose-dependent manner. C Fluorescence staining reveals dose-dependent decreases in mitochondrial membrane potential and increases in apoptosis and ROS levels. D GSH levels decrease with increasing TBHP concentrations. E, F Western blot and QRT-PCR indicate elevated NOX4 expression at both protein and mRNA levels in degenerative (IDD) NPCs compared to normal controls (NC). G Immunofluorescence shows cytoplasmic NOX4 localization, with higher intensity in degenerative NPCs. H Immunohistochemistry demonstrates a progressive increase in NOX4-positive cells in NP tissues with higher degeneration grades. NC: normal group; IDD: degraded group; PI, PIII and PV: the grading of disc degeneration according to the Pfirrmann method.  *p< 0.05, **p< 0.01, ***p< 0.001

To confirm TBHP-induced ferroptosis, we analyzed key markers associated with this cell death pathway. Western blotting demonstrated that, with increasing concentrations of TBHP, the protein levels of NOX4 and ACSL4 were upregulated, while the levels of GSH and GPX4 decreased (Fig. 2B, D). Fluorescence staining revealed a dose-dependent reduction in mitochondrial membrane potential, accompanied by increased apoptosis and ROS production (Fig. 2C). These findings suggest that TBHP enhances ROS production, thereby modulating ferroptosis markers, with NOX4 showing significant upregulation following TBHP treatment.

To investigate the expression of NOX4 in degenerative NPCs, we conducted Western blot and QRT-PCR analyses on isolated cultured NPCs. The results demonstrated a significant upregulation of NOX4 at both protein and mRNA levels in degenerative NPCs (Fig. 2E and F). Additionally, immunofluorescence staining of the collected NPCs revealed that NOX4 is localized in the cytoplasm, with a marked increase in fluorescence intensity observed in degenerative NPCs (Fig. 2G). To explore the correlation between IDD grade and NOX4 expression, we performed immunohistochemical analysis on nucleus pulposus tissues categorized by different degeneration grades. Our findings indicated that NOX4 expression increases in tandem with the severity of degeneration (Fig. 2H). Overall, these observations suggest that as IDD progresses, NOX4 expression is likely upregulated, implying a significant role for NOX4 in the development of IDD.

NOX4 Modulates Ferroptosis, ROS Production, and GSH Levels in NPCs

To enhance our understanding of NOX4's involvement in ferroptosis, we conducted experiments aimed at either loss or gain of function. Initially, we transfected NPCs with either the NOX4 overexpression vector or NOX4-siRNA to generate NPCs that either overexpress NOX4 or lack NOX4, respectively. Western blot analysis of ferroptosis-related markers revealed that NOX4 overexpression is associated with increased expression of ACSL-4, TF, and TFR, while NOX4 knockdown produced the opposite effect (Fig. 3A). GSH assays indicated that NOX4 knockdown results in GSH accumulation, whereas overexpression leads to increased GSH consumption (Fig. 3B). Furthermore, assays measuring mitochondrial membrane potential and ROS levels demonstrated that NOX4 knockdown is linked to increased mitochondrial membrane potential, reduced apoptotic cell counts, and decreased ROS levels, whereas NOX4 overexpression exhibited the contrary effects (Fig. 3C).

Fig. 3.

NOX4 modulates ferroptosis, ROS production, and GSH levels in NPCs. A Western blot analysis of ferroptosis-related proteins (NOX4, ACSL4, TF, and TFR) in NPCs with NOX4 overexpression (oe) or knockdown (Sh) and control groups (NC, oeNC). NOX4 overexpression increases ACSL4, TF, and TFR, while knockdown reduces their levels. B GSH levels in each group, showing accumulation with NOX4 knockdown and depletion with overexpression. C Fluorescence images of mitochondrial membrane potential (Mito-Tracker, red), apoptosis (Annexin V, green), and ROS levels. NOX4 knockdown increases membrane potential and reduces apoptosis and ROS, while overexpression has opposite effects. D Western blot of ferroptosis markers in TBHP-treated NPCs, with or without NOX4 knockdown. TBHP induces NOX4, ACSL4, TF, and TFR, but NOX4 knockdown alleviates these changes. E GSH levels after TBHP treatment, showing partial reversal of GSH depletion by NOX4 knockdown. F Fluorescence images of mitochondrial membrane potential, apoptosis, and ROS in TBHP-treated NPCs, with NOX4 knockdown reducing TBHP-induced changes. NC: knock down control group; Sh; knock down group; oeNC: overexpression control group; oe: overexpression group; TBHP: TBHP induction group; Sh: NOX4 knock down group; TBHP + NOX4: TBHP-induced + NOX4 knockdown group. ∗ p< 0.05, **p< 0.01, ***p< 0.001

To further investigate the role of NOX4 in ferroptosis in NPCs, we induced ferroptosis using TBHP while simultaneously knocking down NOX4 expression to assess its potential to reverse TBHP-induced ferroptosis. Following TBHP treatment, we observed an upregulation of NOX4, ACSL4, TF, and TFR protein levels (Fig. 3D). Additionally, there was a reduction in mitochondrial membrane potential, an increase in apoptotic cells, elevated ROS levels (Fig. 3F), and decreased GSH levels (Fig. 3E). Notably, NOX4 knockdown partially alleviated these changes, suggesting that the knockdown of NOX4 can reverse the ferroptosis phenotype induced by TBHP in NPCs.

NOX4 and LRRC8A Complex Regulates NLRP3 Inflammasome Activation and Cytokine Release

To further elucidate the mechanisms by which NOX4 contributes to IDD, we isolated NOX4-interacting protein complexes using immunoprecipitation and identified the associated proteins through mass spectrometry. Figure 4A presents the chromatographic peaks of proteins from the IgG control and NOX4 groups. Subsequently, enrichment analyses for GO and KEGG were performed on the identified NOX4-interacting proteins, revealing associations with activities related to NADPH oxidase, signaling through protein kinase B, and functions involving focal adhesion (see Fig. 4B). Among these proteins, LRRC8A, a crucial component of volume-regulated anion channels located on the plasma membrane, was particularly emphasized [29]. Immunoprecipitation additionally validated the direct interaction of LRRC8A with NOX4 (refer to Fig. 4C and D). To assess the spatial relationship between LRRC8A and NOX4, immunofluorescence staining was performed, revealing their colocalization within NPCs (Fig. 4E). Given that LRRC8A is a key factor in NLRP3 inflammasome activation [30], we assessed the effect of NOX4 expression on the levels of LRRC8A, NLRP3, and Caspase1 in transfected NPCs. The results indicated that the knockdown of NOX4 reduced the expression levels of LRRC8A, NLRP3, and Caspase1, while the overexpression of NOX4 led to their upregulation (Fig. 4F). When cells were treated with MG132 for 24 h, ubiquitination analysis revealed that NOX4 knockdown significantly enhanced the ubiquitin-mediated degradation of LRRC8A, whereas NOX4 overexpression inhibited its degradation. These findings suggest that NOX4 plays a crucial role in regulating the stability of the LRRC8A protein (Fig. 4G). Furthermore, ELISA analysis demonstrated that NOX4 overexpression increased the levels of TNF-α and IL-1β, while NOX4 knockdown resulted in a reduction of these pro-inflammatory cytokines (Fig. 4H and I). Collectively, these findings suggest that the NOX4/LRRC8A axis may promote ferroptosis in NPCs by activating the NLRP3 inflammasome.

Fig. 4.

NOX4 and LRRC8A complex regulates NLRP3 inflammasome activation and cytokine release. A Chromatographic peaks of proteins identified by mass spectrometry from immunoprecipitation assays in the IgG control and NOX4 groups. B GO and KEGG enrichment analyses of NOX4-interacting proteins, showing associations with NADPH oxidase activity, protein kinase B signaling, and focal adhesion functions. C, D Immunoprecipitation analysis confirming the physical interaction between NOX4 and LRRC8A. E Immunofluorescence staining showing colocalization of NOX4 and LRRC8A in NPCs. F Western blot analysis showing that NOX4 knockdown reduces, while overexpression increases, LRRC8A, NLRP3 and Caspase1 levels in transfected NPCs. G Ubiquitination analysis after 24-h MG132 treatment, demonstrating that NOX4 knockdown promotes, while overexpression inhibits, LRRC8A ubiquitin-mediated degradation. H, I ELISA results indicating that NOX4 overexpression increases TNF-α and IL-1β levels, while knockdown decreases them. NC: knock down control group; TBHP: TBHP induction group; Sh: NOX4 knock down group; TBHP + NOX4: TBHP-induced + NOX4 knockdown group.  *p<0.05, **p< 0.01, ***p< 0.001

Silencing NOX4 alleviates IDD in rats in vivo

To investigate the in vivo effects of NOX4, we induced disc degeneration in rats using caudal needle puncture and administered adenoviruses to either overexpress or knock down NOX4 in the punctured discs (Fig. 5A). After a 12-week period, MRI scans were conducted on the rats. The NOX4 knockdown group displayed normal disc height, structure, and signal intensity, while the overexpression group exhibited significant disc degeneration (Fig. 5B). Pfirrmann scores indicated that the NOX4 knockdown group achieved significantly better scores than the empty vector group, which in turn outperformed the overexpression group, with statistically significant differences observed (P < 0.01, Fig. 5B). These findings suggest that, in an injury-induced model, silencing NOX4 mitigates IDD in rats, whereas its overexpression exacerbates the condition. Histological staining with HE, Masson’s trichrome, and Alcian blue revealed a distinct boundary between the AF and NP in the control group, characterized by large, rounded NPCs and abundant ECM within the NP. Conversely, both NP volume and cell density were significantly diminished in the empty vector and overexpression groups, while notable recovery was observed in the NOX4 knockdown group (Fig. 5C). Moreover, the boundary between AF and NP was lost in the overexpression group. Histological scoring further corroborated the protective effect of NOX4 knockdown on the discs (Fig. 5C). Immunohistochemistry revealed that NOX4 expression was significantly downregulated in the NP tissue of the NOX4 knockdown group, while it was markedly increased in the overexpression group (Fig. 5D).

Fig. 5.

Effects of NOX4 on IDD in a rat model. A Experimental timeline for IDD induction via caudal vertebral puncture, followed by adenoviral NOX4 overexpression or knockdown, with MRI and histological analysis at 12 weeks. B MRI images and Pfirrmann scores indicate preserved disc structure in the NOX4 knockdown group, while NOX4 overexpression leads to significant degeneration (P < 0.01). C Histological staining (Alcian blue, HE, and Masson) shows clear NP-AF boundaries and abundant ECM in the control group. NOX4 knockdown preserves NP structure and ECM, whereas NOX4 overexpression disrupts these features. D Immunohistochemistry confirms reduced NOX4 expression in the knockdown group and increased expression in the overexpression group, with significant differences in positive cell proportions across groups. NC: normal group; IDD + Vector: injection of empty vector adenovirus; IDD + Sh: injection of NOX4 knockdown adenovirus; IDD + NOX4: injection of NOX4 overexpressed adenovirus. **p< 0.01, ***p< 0.001

MiR-6843-3p Regulates NOX4 Expression and Influences Ferroptosis in NPCs

The regulatory mechanisms governing NOX4 expression remain unclear. Our objective was to identify the upstream regulatory factors influencing NOX4. Utilizing bioinformatics tools such as TargetScan, RNA22, MiRDB, and mirDIP, we predicted potential upstream miRNAs for NOX4 and identified miR-6843-3p as the intersecting miRNA (Fig. 6A). In alignment with the trends observed in the bioinformatics analysis, QRT-PCR demonstrated that the levels of miR-6843-3p were significantly elevated in normal NP tissue but reduced in severely degenerated NP tissue (Fig. 6C), indicating a negative correlation with NOX4 expression levels (Fig. 6B). As predicted by the TargetScan database, NOX4 possesses a sequence complementary to the seed region of miR-6843-3p. Luciferase reporter assays revealed that the miR-6843-3p mimic significantly inhibited the luciferase activity of the NOX4 wild-type (WT) construct, while the luciferase activity of the NOX4 mutant (MUT) construct exhibited no significant change. This suggests that NOX4 can directly bind to miR-6843-3p through complementary targeting (Fig. 6D). Based on these findings, we hypothesize that miR-6843-3p may serve as a key upstream regulatory factor of NOX4.

Fig. 6.

miR-6843-3p regulates NOX4 expression and inhibits ferroptosis in NPCs. A Venn diagram identifying miR-6843-3p as a common upstream miRNA of NOX4 through bioinformatics analysis. B Negative correlation between miR-6843-3p and NOX4 expression in NP tissues (r =  −0.7836 , p = 0.0073). C QRT-PCR showing lower miR-6843-3p levels in IDD tissues compared to NC. D Luciferase assay showing that miR-6843-3p mimic reduces NOX4 WT construct activity, but not the MUT construct, indicating direct binding. E QRT-PCR showing miR-6843-3p overexpression decreases NOX4 mRNA levels. F Western blot confirming reduced NOX4 protein with miR-6843-3p mimic. G Fluorescence staining showing increased mitochondrial membrane potential and decreased apoptosis and ROS with miR-6843-3p mimic. H Elevated GSH levels with miR-6843-3p mimic. I Lower TNF-α and IL-1β in miR-6843-3p mimic group compared to IDD. J Western blot showing miR-6843-3p mimic downregulates ACSL4, TF, TFR, NLRP3, and LRRC8A. NC: normal group; IDD: degeneration group; mimic: miR-6843-3p mimic. *p< 0.05, **p< 0.01,***p< 0.001

To investigate whether miR-6843-3p serves as a key upstream regulator of NOX4, we conducted a series of experiments to explore their relationship. Transfection of the miR-6843-3p mimic into NPCs significantly increased the levels of miR-6843-3p (Fig. 6E). The overexpression of miR-6843-3p was found to negatively regulate the expression levels of NOX4 mRNA and protein (Fig. 6E and F). Our findings also demonstrated that the miR-6843-3p mimic enhanced mitochondrial membrane potential in degenerated NPCs, reduced cell apoptosis and ROS levels (Fig. 6G), and upregulated GSH levels (Fig. 6H). ELISA results indicated that the miR-6843-3p mimic led to decreased levels of TNF-α and IL-1β (Fig. 6I). Subsequent Western blot analysis revealed that the miR-6843-3p mimic downregulated the protein expression levels of ACSL4, TF, TFR, the NLRP3 inflammasome, and LRRC8A in NPCs (Fig. 6 J). These findings suggest that miR-6843-3p can regulate the expression of NOX4 and LRRC8A and inhibit ferroptosis in NPCs by modulating inflammatory responses.

Discussion

In IDD, ferroptosis in NPCs plays a pivotal role [8–10]. Ferroptosis is a novel form of cell death characterized by lipid peroxidation and the accumulation of ROS. During degeneration, elevated ROS levels and decreased activity of antioxidant enzymes such as GPX4, lead to lipid peroxidation and cell death [6, 7]. Our research is the first to report the critical role of NOX4 in mediating ferroptosis in NPCs and its significant contribution to IDD. Furthermore, it indicates that, in comparison to normal NPCs, degenerated NPCs exhibit increased levels of ROS and ACSL4, decreased mitochondrial membrane potential, and reduced expression of GSH and GPX4. Immunohistochemical analysis confirmed that GPX4 expression decreases while ACSL4 expression increases as degeneration progresses. In an in vitro ferroptosis model, stimulation with TBHP further validated these findings by reducing cell viability, downregulating GSH and GPX4, and upregulating ACSL4..

ROS-mediated lipid peroxidation serves as a central trigger of ferroptosis, leading to cellular damage and death [11]. In the context of IDD, oxidative stress in NPCs is intensified by factors such as cartilage endplate degeneration, annulus fibrosus rupture, and inflammation, all of which contribute to increased ROS production [31]. Extensive research has underscored the involvement of ROS in ECM metabolism, programmed cell death, and cellular senescence in IVD cells [16, 32]. NOX4, one of the most widely expressed members of the NOX family, plays a critical role in regulating intracellular ROS production and oxidative stress [33, 34]. While most NOX proteins primarily produce superoxide (O²⁻), NOX4 uniquely catalyzes the reduction of molecular oxygen to H₂O₂ [35]. This distinctive function endows NOX4 with specific roles in various cellular processes, including proliferation, differentiation, apoptosis, and senescence [36–39]. Its upregulation results in mitochondrial dysfunction and promotes ferroptosis in neurons and astrocytes. Although these mechanisms are well-characterized in those conditions, the specific role of NOX4 in regulating ferroptosis in NPCs during IDD remains insufficiently explored. In this study, our findings demonstrated a significant upregulation of NOX4 at both the protein and mRNA levels in degenerated NPCs, with NOX4 primarily localized in the cytoplasm. Immunohistochemical analysis of NP tissues further revealed a positive correlation between NOX4 expression and the severity of IDD. These results suggest that NOX4 may play a crucial role as a ferroptosis-related gene in IDD, emphasizing the necessity for further investigation into its underlying mechanisms.

Inflammatory cytokines play a crucial role in the pathogenesis of IDD by facilitating ECM degradation and attracting immune cells to the IVD tissue [40]. Key cytokines, such as IL-1β and TNF-α, are significant drivers of IVD inflammation and contribute markedly to IDD [41]. This inflammatory response is mediated through inflammasome complexes, with the NLRP3 inflammasome being particularly prominent in recent research on IDD [42]. Furthermore, ROS have been identified as critical factors in the activation of the NLRP3 inflammasome [20]. Recent findings have linked NLRP3 activation to inflammation, pyroptosis, ECM degradation, and apoptosis in IVD cells [42]. Our Western blot analysis demonstrated that NOX4 knockdown significantly decreases NLRP3 expression, while NOX4 overexpression increases it. ELISA assays revealed similar trends in IL-1β and TNF-α levels, suggesting that NOX4 may promote ferroptosis in NPCs by generating excessive ROS, which activates NLRP3.

To further elucidate the role of NOX4 in ferroptosis in NPCs, we conducted both loss-of-function and gain-of-function experiments. Our analysis of key indicators revealed that NOX4 overexpression significantly promoted ferroptosis, increased ROS production, and enhanced the secretion of inflammatory cytokines. Conversely, in a TBHP-induced NPC degeneration model, NOX4 knockdown reversed the ferroptosis induced by TBHP, suggesting a crucial regulatory role of NOX4 in this process. In addition, adenoviral injection experiments in an injury-induced rat model confirmed that NOX4 knockdown could partially alleviate the progression of IDD, whereas its overexpression exacerbated the condition. This study provides novel mechanistic insights into the role of NOX4 in ferroptosis and its involvement in IDD. Subsequently, we utilized mass spectrometry to identify proteins that interact with NOX4, leading to the discovery of LRRC8A, a crucial component of the volume-regulated anion channel. LRRC8A is essential for regulating various cellular processes, including cell survival, growth, energy sensing, and immune responses [29]. Notably, LRRC8A is essential for the activation of the NLRP3 inflammasome. Previous researches have shown that LRRC8A deficiency specifically inhibits the activation of the canonical NLRP3 inflammasome and reduces the accumulation of damaged mitochondria [30]. Furthermore, in vascular smooth muscle cells, LRRC8A has been reported to colocalize with Nox1 and its p22phox subunit [43]. Based on these findings, we hypothesized that LRRC8A may physically interact with NOX4 in NPCs. Through immunofluorescence staining, we provided the first evidence of colocalization between LRRC8A and NOX4 in NPCs, suggesting a functional interdependence between these two proteins. This colocalization indicates the potential formation of a molecular complex that may regulate critical cellular processes. Immunoprecipitation assays further validated the physical interaction between NOX4 and LRRC8A, thereby strengthening the hypothesis of their direct association. Interestingly, both knockdown and overexpression of NOX4 resulted in significant alterations in LRRC8A expression levels, highlighting a regulatory relationship. These findings imply that NOX4 and LRRC8A form a functional complex that plays a pivotal role in modulating downstream signaling pathways. Specifically, these results indicate that NOX4 and LRRC8A jointly regulate NLRP3 inflammasome activation, thereby contributing to ferroptosis in NPCs.

The biological significance of this interaction lies in its influence on ferroptosis within NPCs. By modulating the activation of the NLRP3 inflammasome, the NOX4-LRRC8A complex orchestrates a cascade of events that culminate in oxidative stress, mitochondrial dysfunction, and lipid peroxidation, which are hallmark features of ferroptosis. This regulatory mechanism not only contributes to the pathological progression of IDD but also offers novel insights into the molecular crosstalk between ferroptosis and inflammation in degenerative diseases. Consequently, targeting the NOX4-LRRC8A axis could provide a promising therapeutic strategy for alleviating IDD and related conditions.

MiRNAs, a class of small non-coding RNAs, regulate gene expression by binding to target mRNAs [44]. Recent studies have demonstrated that miRNAs are intricately involved in the key biological processes associated with IDD [45]. Our study reveals that miR-6843-3p is highly expressed in normal NP tissue but significantly decreased in severely degenerated NP tissue, indicating a negative correlation with NOX4 expression. The upregulation of NOX4 in IDD may be partly attributed to the downregulation of miR-6843-3p, which diminishes the inhibitory effect of miR-6843-3p on NOX4. This study demonstrates that miR-6843-3p regulates the expression of NOX4 and LRRC8A, and plays a role in the ferroptosis of NPCs. By modulating the NOX4/LRRC8A/NLRP3 axis, miR-6843-3p appears to exert a protective effect against ferroptosis in NPCs. The downregulation of NOX4 by miR-6843-3p reduces ROS levels, stabilizes LRRC8A activity, and consequently decreases NLRP3 activation. This regulatory pathway highlights NOX4 as a central mediator of ferroptosis and suggests that targeting NOX4 could provide a novel therapeutic approach to mitigating IDD.

This study has certain limitations. First, the small sample size of clinical specimens may introduce statistical bias. Second, significant differences exist between in vitro experimental conditions and the in vivo environment, which may contribute to variability in the results. Additionally, the biomechanical properties of caudal intervertebral discs differ from those of lumbar discs, and the biological functions of nucleus pulposus cells may also vary, potentially limiting the interpretation of the findings.

Conclusion

This study underscores the critical role of NOX4 in promoting ferroptosis in NPCs through the generation of excessive ROS and the activation of the NLRP3 inflammasome. The interaction between NOX4 and LRRC8A further promotes this process, indicating that the NOX4/LRRC8A/NLRP3 axis plays a role in the progression of IDD. Additionally, miR-6843-3p has been identified as a key regulator of NOX4 and ferroptosis in NPCs, presenting a potential therapeutic target for mitigating the progression of IDD. The specific regulatory mechanisms are detailed in Fig. 7. Future studies should aim to further elucidate the regulatory mechanisms of the NOX4/LRRC8A complex and explore the therapeutic potential of targeting NOX4 in clinical settings.

Fig. 7.

Schematic diagram of miR-6843-3p alleviates IDD by regulating NLRP3 through NOX4/LRRC8A axis to inhibit ferroptosis of NPCs(By Figdraw)

Abbreviations

IDD

Intervertebral disc degeneration

IVD

Intervertebral disc

NP

Nucleus pulposus

AF

Annulus fifibrosus

CEP

Cartilage endplate

NPCs

Nucleus pulposus cells

ECM

Extracellular matrix

ROS

Reactive oxygen species

RCD

Regulated cell death

CCK-8

Cell Counting Kit-8

EdU

5-Ethynyl-2′-deoxyuridine

QRT-PCR

Quantitative reverse transcription polymerase chain reaction

UPLC

Ultra-high-performance liquid chromatography

MRI

Magnetic resonance imaging

HE

Hematoxylin-eosin

LRRC8A

Leucine rich repeat containing 8A

TF

Transferrin

TFR

Transferrin receptor

NADPH

Nicotinamide adenine dinucleotide phosphate

NOX4

Nicotinamide adenine dinucleotide phosphate Oxidase4

ASC

Apoptosis-associated speck-like protein containing a CARD

GO

Gene ontology

KEGG

Kyotoencyclopedia of genes and genomes

WT

Wild-type

MUT

Mutant

Author Contributions

H. Yu., W. J. and X. C. conceived and designed research; F. Z. and D. performed experiments; Z. W. and Y. L. analyzed data; K. W. and H. L. prepared figures; F. Z. drafted manuscript; F. Z. and D. C. edited and revised manuscript; All authors have read and approved the final manuscript.

Funding

This research was supported by University Natural Science Research Project of Anhui Province in 2024(Project No. 2024AH050760), the Open Project Fund of Ministerial Key Laboratory of Bengbu Medical College in 2022 (Project No. AHTT2022A004, AHTT2022B003 and AHIAI2022R03) and Anhui Spinal Deformity Clinical Medical Research Center Innovation Fund Project in 2023(Project No. AHJZJX-GG2023-005 and AHJZJX-GG2023-003).

Data Availability

No datasets were generated or analysed during the current study.

Declarations

Conflicts of Interest

The authors declare no competing interests.