Integrating bioinformatics and experimental validation to Investigate IRF1 as a novel biomarker for nucleus pulposus cells necroptosis in intervertebral disc degeneration

Abstract

Intervertebral disc degeneration (IDD) is a prevalent spinal disorder and the principal cause of lower back pain (LBP). Diverse forms of programmed cell death (PCD) have been identified as the key phenotypes of the disease and have the potential to serve as new indicators for the diagnosis and prognosis of IDD. However, the mechanism underlying necroptosis in IDD remains unclear. This study aimed to identify novel biomarkers that promote nucleus pulposus cell necroptosis in IDD using bioinformatic analysis and experimental validation. We analyzed multiple datasets of IDD from the Gene Expression Omnibus (GEO) database to identify necroptosis-related IDD differential genes (NRDEGs). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed, followed by logistic least absolute shrinkage and selection operator (LASSO) and support vector machine-recursive (SVM) algorithms to identify key genes. Gene set enrichment analysis (GSEA) and logistic regression analysis were used to ascertain the potential functions of these genes and to identify key genes, respectively. We then constructed mRNA-miRNA, mRNA-TF, mRNA-drug, and functional similarity gene interaction networks for the seven key genes identified. We used IDD clinical samples and necroptotic cell model to validate our findings. Immunohistochemical staining, RT-qPCR, and western blotting results indicated that IRF1 may be a hub necroptosis-related gene. To further elucidate the function of IRF1, we constructed IRF1 knockdown and overexpression models, which revealed that IRF1 promotes necroptosis in rat nucleus pulposus cells, increases mitochondrial ROS levels, and decreases ATP levels. These findings provide new insights into the development of necroptosis in IDD and, for the first time, validate the role of IRF1 as a novel biomarker for the diagnosis and treatment of IDD.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-81681-8.

Introduction

Low back pain (LBP) is a prevalent global health issue, affecting 22–48% of the population, with intervertebral disc degeneration (IDD) being one of its primary causes¹. It adversely impacts individuals’ quality of life and results in socio-economic consequences^2,3, Studies have shown that low back pain accounts for more than 30% of absenteeism, leading to a significant decrease in productivity. In addition, the direct medical costs associated with low back pain are considerable. These costs not only have an impact on individuals and families, but also put pressure on social care systems and hinder economic development^4,5. IDD typically manifests as back pain, radicular symptoms, and impaired spinal mobility, posing significant challenges to patient well-being and health care management⁶. It is characterized by the progressive deterioration and structural changes of intervertebral discs, which are fibrocartilaginous structures located between adjacent vertebrae. These discs play a crucial role in providing spinal flexibility, distributing loads, and absorbing shocks⁷. Intervertebral discs are susceptible to degeneration, leading to spinal disorders such as lumbar disc herniation and IDD. The pathogenesis of IDD is complex and involves multiple factors, such as genetic predisposition, environmental influences^8–14, mechanical stress, nutritional status, and alterations in cellular biology^15–17. A comprehensive understanding of these intricate mechanisms and contributing factors is crucial for advancing therapeutic interventions and optimizing patient outcomes for spinal health.

Programmed cell death (PCD) is crucial for maintaining tissue homeostasis and eliminating harmful cells¹⁸. The regulation of programmed cell death is a complex process that involves a variety of molecular mechanisms¹⁹. These include the activation of intracellular signalling pathways, the regulation of transcription factors and the involvement of effector molecules that execute cell death²⁰. In the context of IDD, PCD, particularly in the nucleus pulposus (NP), the central part of the disc, is a key feature of the degenerative process²¹. Necroptosis, a programmed form of cell death, has garnered considerable attention in cell biology and immunology due to its unique characteristics, bridging the gap between apoptosis and necrosis^22,23. Unlike traditional apoptosis, necroptosis does not rely on caspase activation; instead, it involves the activation of receptor-interacting protein kinases (RIPKs) and mixed lineage kinase domain-like (MLKL) protein^24–26. This intricate signaling cascade leads to plasma membrane rupture and the release of intracellular contents, causing inflammation and immunogenic cell death. Necroptosis plays a vital role in various physiological and pathological contexts, including embryonic development, tissue homeostasis, and host defense against pathogens. However, dysregulation of this process has been linked to numerous diseases, such as inflammatory disorders, neurodegeneration, and specific types of cancer. The broad implications of necroptosis make it a compelling area of research in understanding cell death mechanisms and their impact on human health.

Currently, research on necroptosis in IDD has shown promising results. Intervertebral discs, which are cartilaginous structures situated between vertebral bones, are susceptible to degeneration, leading to spinal disorders like lumbar disc herniation and intervertebral disc degenerative diseases. Studies have identified necroptosis-related proteins, including RIP1 and RIP3, in intervertebral disc tissues^27,28. These findings suggest that necroptosis plays a vital role in cell death among intervertebral disc cells^29,30, potentially affecting tissue degeneration by influencing cell survival and death. Additionally, IDD is often associated with an inflammatory response^31,32, raising the possibility that necroptosis might be linked to this inflammatory process in intervertebral disc tissue, and participates in regulating cell death and survival in inflammation-related cells. Despite progress in elucidating the role of necroptosis in IDD, this field remains in its infancy and warrants further investigation.

The present study aimed to unravel the complex interplay between necroptosis and IDD by analyzing diverse GEO datasets using a comprehensive bioinformatics approach. Our analysis identified 18 differentially expressed genes (NRDEGs) that were dysregulated in IDD, with a focus on seven key genes. Notably, our bioinformatics analysis and experimental validation revealed that IRF1 is significantly upregulated in necroptotic processes within IDD. Further gain-of-function and loss-of-function experiments confirmed that IRF1 can promote necroptosis of nucleus pulposus cells. This discovery suggests that IRF1 is not only a novel biomarker for necroptosis in IDD, but also a promising target for therapeutic intervention. Our findings have the potential to improve diagnostic and treatment strategies and offer new hope for individuals with IDD and related spinal disorders.

Materials and methods

Data collection and preprocessing

IDD-related datasets, namely GSE56081 (Normal = 5, IDD = 5)³³, GSE70362 (Normal = 8, IDD = 16)³⁴, GSE147383 (Normal = 2, IDD = 2)³⁵, and GSE34095 (Normal = 3, IDD = 3)³⁶ were from the Gene Expression Omnibus (GEO) database using the GEOquery package³⁷. The datasets GSE56081, GSE70362, and GSE147383 were merged into a GEO merged dataset called “Combined Datasets,” while GSE34095 was kept separate as the validation set.

To identify Necroptosis-related genes (NRGs), firstly, 8 NRGs were retrieved from the necroptosis-related gene set available in the Molecular Signatures Database (MsigDB)³⁸, using the keyword “Mitophagy.” And 93 NRGs were collected associated with “Necroptosis” from the GeneCards database³⁹. Lastly, 212 NRGs were obtained from references^40–42. The combined list from these three sources resulted in a total of 277 NRGs.

Determination of NRDEGs

In order to obtain the differentially expressed genes (DEGs) associated with different groups (IDD, Control), we performed batch correction using the R package sva⁴³ on the datasets GSE56081, GSE70362, and GSE147383. The batch-corrected dataset, named “Combined Datasets,” comprised 23 samples from IDD and 15 samples from the control group.

The DEGs were procured using the R package “limma”. The criteria^44,45 employed for identifying DEGs encompassed |logFC| > 0.3 and p.value < 0.05. To pinpoint DEGs linked with necroptosis, the amalgamation of all DEGs extracted from the differential analysis of the Combined Datasets and the NRGs was undertaken. The overlapping genes were referred to as Necroptosis-Related Differentially Expressed Genes (NRDEGs). Using a Venn diagram and created volcano plots and differential expression ranking plots with the “ggplot2” package in R to visualize the overlapping genes. Additionally, a heatmap of the NRDEGs using the “pheatmap” package was generated for further visualization and analysis.

Functional enrichment analysis

GO and KEGG annotation analyses were conducted using the “clusterProfiler” package⁴⁶ in R. To identify enriched entries, we applied the significance criteria of p.adjust < 0.05 and a False Discovery Rate (FDR) value (q-value) < 0.25, utilizing the Benjamini-Hochberg (BH) method for multiple testing correction. The enriched pathways from the KEGG database were visualized using the R package Pathview⁴⁷ to display relevant pathway diagrams for a comprehensive analysis of our data.

GSEA enrichment analysis

Pathway enrichment analysis of DEGs in the Combined Datasets was performed by ranking the DEGs based on logFC (fold change). The “clusterProfiler” package and the gene set “c2.all.v2022.1.Hs.symbols.gmt” from the MSigDB database⁴⁸ was utilized to conducted enrichment analysis. Significance criteria for enrichment were set as p.adjust < 0.05 and FDR value (q-value) < 0.25.

Construction of the diagnostic model of IDD

LASSO regression analysis using the R package “glmnetp“⁴⁹ was performed based on the hub genes present in the Combined Datasets. Diagnostic model plots and variable trajectory plots were used to visualize the outcomes of the LASSO regression analysis. Furthermore, a Forest Plot was employed to illustrate the molecular expression profiles of the LASSO regression model’s hub genes.

In addition, a SVM⁵⁰ model was developed using the hub genes extracted from the Combined Datasets. The intersection of the LASSO regression model and the SVM model facilitated the identification of necroptosis-associated hub genes (p < 0.05). For mapping the chromosomal positions of these hub genes, the RCircos package was harnessed to create a chromosome localization plot⁵¹. To establish the diagnostic model for IDD from the Combined Datasets, logistic regression analysis was conducted on the hub genes. The results of this analysis were then used to construct a nomogram⁵² using the “rms” package in R.

Validation of the diagnostic model of IDD

Decision Curve Analysis (DCA)⁵³ was performed on the Combined Datasets to analyze necroptosis through the R package ggDCA. Additionally, ROC curves were separately plotted for the key genes and the logistic regression model in both the Combined Datasets and the independent validation dataset GSE34095, using the R package “pROC”. Furthermore, the area under the ROC curve (Area Under the Curve, AUC) was calculated to assess the diagnostic performance of the key genes’ expression levels and the logistic prediction values for IDD occurrence. The AUC of the ROC curve is generally between 0.5 and 1, where values closer to 1 indicate better diagnostic performance.

Construction of regulatory network

The miRDB database⁵⁴ was used to predict the miRNAs that interact with the key genes, and the CHIPBase database⁵⁵ and the hTFtarget database were utilized to search for TFs that bind to the key genes. The CTD database⁵⁶ was used to predict potential drugs or small molecular compounds that interact with the key genes. Subsequently, using the Cytoscape software to visualized these interactions. Finally, the GeneMANIA⁵⁷ website was employed to predict the interaction network of similar genes to the key genes.

Immune infiltration analysis

The gene expression matrix data from the Combined dataset were uploaded to CIBERSORTx (http://cibersortx.stanford.edu/). After filtering the output data, only the immune cells with enrichment scores greater than zero were retained, resulting in specific results for the immune cell infiltration matrix.

The ssGSEA algorithm³⁷ was employed to quantify the relative abundance of 28 types of immune cell infiltrates in the Combined dataset. At the same time, a correlation analysis on the degree of infiltration of various immune cells in the sample and the expression of key genes was performed.

Clinical samples collection

This study recruited 9 cases of nucleus pulposus(NP) tissue with low degeneration(obtained from patients who required surgery due to adolescent idiopathic scoliosis) and 9 cases of nucleus pulposus tissue with high degeneration diagnosed with intervertebral disc degeneration (including: lumbar stenosis, lumbar spondylolisthesis, lumbar disc herniation) for Magnetic resonance images (MRIs) Supplementary Table S1, among which, 6 pairs were used for quantitative polymerase chain reaction RT-qPCR analysis (details see “Cell model construction, RNA extraction, and RT-qPCR”), and the other 3 pairs were used for pathological staining and immunohistochemical staining. Pfirrmann grading standards of IDD was shown in Supplementary Table S2, grade I and II are defined as low degeneration group, grade IV and V are defined as high degeneration group. The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Ethics Committee of Lanzhou University Second Hospital (No.2024 A-651) for studies involving humans, and all patients signed informed consent.

Haematoxylin eosin (HE), safranin O/Fast green and immumohistochemical staining

Samples for histology were fixed in 4% methanol-free paraformaldehyde (G1101-500ML, Servicebio, Wuhan, China) overnight, and then sequentially dehydrated, clarified, paraffin embedded and cut into 5 μm-thick sections using a pathology slicer(RM2016,Shanghai Leica Instrument Co., Ltd).

Sections were rehydrated and stained with Haematoxylin-Eosin (HE), Safranin O/Fast green and Immumohistochemical staining, according to standard protocols. It is worth noting that the proteins we detected by immunohistochemical staining were p-MLKL(1:100, AF7420, Affinity Biosciences, China) and IRF1 (1:100, 11335-1-AP, Proteintech, China).

Preparation of rat primary nucleus pulposus cells

Primary nucleus pulposus cells were isolated from the caudal vertebrae of healthy male Sprague-Dawley rats (SD rats, 150–180 g). After SD rats were sacrificed with an overdose of sodium pentobarbital (100 mg/kg), their tails were severed from their roots with tissue shears and immersed in 75% ethanol for 5 min. The intervertebral disc was cut along the intervertebral space with a sterile scalpel, and the nucleus pulposus tissue was exposed in turn. Then the gelatinous nucleus pulposus tissue was removed with forceps and placed in a small dish containing 2 ml DMEM/F12 complete culture medium. An additional 2 ml of 0.2% collagenase II solution was added and digested in a 37 °C incubator for 20 min. Then, the cells can be cultured in 5% CO2 incubator at 37 °C with saturated humidity. After 3–5 days, the cells were basically adherent, and the cells of passages 2–4 were used for subsequent experiments. The results of phenotype identification of NP cells are shown in supplementary Fig. S1.

The study was conducted in accordance with the Declaration of ARRIVE guidelines and confirmed that all experimental protocols were approved by the Institutional Ethics Committee of Lanzhou University Second Hospital (No.D2024-609) for studies involving rats, and all methods were carried out in accordance with relevant guidelines and regulations.

Cell model construction, RNA extraction, and RT-qPCR

Following the method reported in the literature, we constructed a model of necroptosis of nucleus pulposus cells (NP cells) with treating the cells with Necroptosis Inducer Kit with TNF-α, SM-164 and Z-VAD-FMK (TSZ, 1:1000; C1058S, Beyotime, China) for 12 h^58–60. The result of cell time gradient viability identification for treatment at the TSZ reagent is shown in Supplementary Fig. S2. After 12 h, discarding the culture medium, washing the cells with 1 ml of PBS, and then using AG RNA ex Pro RNA Extraction Reagent (AG21101, Accurate Biology, China) to sequentially extract RNA following the instructions provided in the product manual.

Total RNA was extracted from tissues and cells using TRIzol (AG21101; ACCURATE BIOLOGY AG, Guangzhou, China). The RNA was reverse-transcribed into cDNA using the Evo M-MLV RT Mix Kit with gDNA Clean for qPCR Ver.2 (AG11728, Accurate Biology, China), following the procedures outlined in the manual. For q-PCR analysis, we utilized the SYBR Green Premix Pro Taq HS qPCR Kit (AG11701, Accurate Biology, China) and the Cobas z 480 analyzer (Roche, Switzerland). To determine the relative gene expression levels, we employed the 2^−ΔΔCT method, with GAPDH serving as the reference gene⁶¹. After the fresh nucleus pulposus tissue collected clinically was mechanically broken, RNA was extracted using the same method and RT-qPCR experiments were performed to detect the expression of 7 key genes. The primer sequences for human NP tissue and rat NP cells used are showed below (5’–3’) (Table 1 for human, Table 2 for rat).

Table 1.

The primer sequences of key genes for human.

Gene primers		Sequences(5′–3′)
IFNGR2	(F)	CTCTCCCTTTGACATCGCTGATAC
	(R)	GCCTTTGACCTGTTGGATTCCTC
SLC25A6	(F)	TGGACAAGCACACGCAGTTC
	(R)	ACACGAAGCAGAGGGAGGTC
IRF1	(F)	AGCCCTGATACCTTCTCTGATGG
	(R)	GCATGTAGCCTGGAACTGTGTAG
MAP3K7	(F)	TGCCGCCTCCTCCTCCTC
	(R)	GCTCCTCTTCCAACAACCTCTTC
PELI1	(F)	GATGGCTCGTTAATTGACCTCTGTG
	(R)	GCTTCTAAATGCTTCACGGTAGGAG
TNFRSF1A	(F)	TGACCTGCTGCTGCCACTG
	(R)	ACACACTATCTCTCTTCTCCCTGTC
TNF	(F)	CTCATCTACTCCCAGGTCCTCTTC
	(R)	CGATGCGGCTGTTGGTGTG
GAPDH	(F)	GATTCCACCCATGGCAAATTC
	(R)	CTGGAAGATGGTGATGGGATT

Table 2.

The primer sequences of key genes for rat.

Gene primers		Sequences(5′–3′)
IFNGR2	(F)	ACTGTACGGACATCGCAGAGAC
	(R)	GCAGGAAGACTGTGTATGAGTATGG
SLC25A6	(F)	GGTTCGCCGTCGTATGATGATG
	(R)	GATCTTCCGCCAGCAGTCAAG
IRF1	(F)	TGTCACCGTGCGTCGTCAG
	(R)	TCCCTTCCTCATCCTCATCTGTTG
MAP3K7	(F)	CAGTCTGTCTTGTGATGGAGTATGC
	(R)	AGCAGCAGTGTAATAAGGCAATGG
PELI1	(F)	TCTCTCCCAAACGGCGACAG
	(R)	TGCGGTCCTTGTTGCTTATTGC
TNFRSF1A	(F)	GCTGTTGCCTCTGGTTATCTTCC
	(R)	TCCTTCACCCTCCACCTCTTTG
TNF	(F)	CTGTGCCTCAGCCTCTTCTC
	(R)	ACTGATGAGAGGGAGCCCAT
GAPDH	(F)	GACATGCCGCCTGGAGAAAC
	(R)	AGCCCAGGATGCCCTTTAGT
RIPK3	(F)	CCAGCTCGTGCTCCTTGACT
	(R)	TTGCGGTCCTTGTAGGTTTG
MLKL	(F)	TCTCCCAACATCCTGCGTAT
	(R)	TCCCGAGTGGTGTAACCTGTA

Western blotting

Cellular proteins were extracted using RIPA (P0013B, Beyotime, China). The subsequent protein concentration and purity were assessed using a BCA kit (Beyotime, Shanghai, China). Western blotting analysis of protein expression was performed according to standard experimental procedures. The proteins were loaded onto an SDS-PAGE gel before being transferred to a PVDF membrane (IPVH00010, Millipore, USA), blocked with blocking solution (Beytiome, China) for 1 h, and incubated overnight at 4 degree centigrade with antibodies. The main antibodies were used: anti-IRF1 (1:1000, #8478S, CST), anti-RIPK1(1:1000, #3493T, CST), anti-p-RIPK1(1:1000,YP1467,Immunoway), anti-RIPK3(1:1000, #15828S, CST), anti-p-RIPK3 (1:1000, AF7443, Affinity Biosciences), anti-MLKL (1:5000, 66675-1-Ig, Proteintech), anti-p-MLKL (1:1000, AF7420, Affinity Biosciences), anti-GAPDH (1:5000, ab219620, Abcam), which were normalized by blotting the same membrane with an antibody against GAPDH. Secondary antibodies were incubated for 1 h. ECL Detection Reagent (Solarbio, Beijing, China) was used to determine the protein signals.

Mitochondrial ROS, cellular ATP detection and PI staining

Mitochondrial ROS levels were measured using Mito-SOX Red dye (Invitrogen). NP cells were incubated with Mito-Sox Red reagent for 30 min at 37 °C and then washed three times with PBS. After the above treatment, three randomly selected images were captured using a light microscope (Olympus Inc.). The NP cellular ATP levels were measured using a firefly luciferase-based ATP assay kit (Beyotime) based on a fluorescence technique (Genmed Scientifics Inc.).

Cell death analysis was performed by propyridine iodide (PI) staining, numerical counting was performed by fluorescent nuclear staining with ImageJ, and TUNEL-positive nuclei were manually counted on a set of high-power immunofluorescence images. Statistical evaluations were performed using the unpaired Student’s t-test (two-tailed).

Cell transfection

IRF1 shRNA was purchased from Genechem (China). Cells were seeded at a density of 1 × 10⁴ cells/well in 6-well plates and cultured until reaching 30–50% confluency. Then, they were infected with lentiviruspackaged sh-NC, sh-IRF1 (GCACCACTGATCTGTACAACT) or oe-IRF1(LV-Irf1) at a multiplicity of infection (MOI) of 20. Following a 12-hour transfection period, the culture medium was replaced, and the cells were further cultured for 3–5 days in a complete growth medium. Upon reaching approximately 80% confluency, the cells were passaged at a ratio of 1:3. The transfection efficiency was confirmed by means of polymerase chain reaction (PCR) and Western blot analysis.

Statistical analysis

R 4.1.1 software was employed for data processing and analysis. The details of R packages as follows: GEOquery (Version 2.70.0, https://bioconductor.org/packages//GEOquery), ggplot2 (Version 3.4.4, https://github.com/tidyverse/ggplot2), limma (Versio n 3.58.1, https://bioconductor.org/packages/limma), clusterProfiler (Version 4.10.0, https://bioconductor.org/packages/clusterProfiler), pheatmap (Version 1.0.12, https://cran.r-project.org/web/packages/pheatmap), sva (Version 3.50.0, https://bioconductor.org/packages/sva), glmnetp (Version 4.1-8, https://glmnet.stanford.edu), ggDCA (Version 1.1, https://cran.r-project.org/package=ggDCA), pROC (Version 1.18.5, https://xrobin.github.io/pROC). For comparisons of continuous variables between the two groups, statistical significance was estimated with the use of the independent Student’s t-test for normally distributed variables and with the use of the Mann-Whitney U test (i.e., the Wilcoxon rank-sum test) for non-normally distributed variables. For comparisons of three or more groups, the Kruskal-Wallis test method was used. Comparisons between categorical variables were performed by the chi-square test or Fisher’s exact test. Spearman correlation analysis was used to calculate the correlation coefficients between different molecules, and the results are presented as correlation coefficients. For the experimental data were all expressed as mean ± standard deviation (mean ± SD), Independent sample T test was used for comparison between two groups, and one-way analysis of variance was used for comparison between three or more groups. Unless otherwise stated, all statistical P values are two-sided and presented as P < 0.05 was considered statistically significant. (* means P < 0.05,* * means P < 0.01,* ** means P < 0.001).

Results

Identification of NRDEGs in IDD

Firstly, we got a Combined Datasets from the combination of GSE56081, GSE70362, and GSE147383 with batch effects removed. To obtain NRDEGs, we divided the data of the Combined Datasets into IDD group and Control group. We performed differential analysis on the Combined Datasets to identify differentially expressed genes (DEGs) between the two groups. A grand total of 1426 genes were pinpointed as DEGs in the Combined Datasets, with 761 genes showing upregulation and 665 genes displaying downregulation (Fig. 1A,B). To obtain NRDEGs, we took the intersection of all DEGs with NRGs and resulted in 18 NRDEGs (BID, CDK9, FAS, FLOT1, GZMB, IFNGR2, IRF1, IRF9, MAP3K7, PELI1, PPIA, SLC25A37, SLC25A6, TNF, TNFAIP3, TNFRSF1A, TRADD, UCHL1) (Fig. 1C). Based on the intersection results, we analyzed the expression differences of NRDEGs between the IDD and Control groups in the Combined Datasets and generated group comparison plots (Fig. 1D) and heatmaps (Fig. 1E). Similarly, we also analyzed the expression differences of NRDEGs in the dataset GSE34095 (Fig. 1F).

Fig. 1.

Analysis of DEGs related to necroptosis. (A) Volcano plot of DEGs in the Combined Datasets, with NRDEGs highlighted for further analysis between the IDD and Control groups. (B) Differential sorting plot of DEGs in the Combined Datasets, with NRDEGs marked for further analysis between the IDD and Control groups. (C) Venn diagram showing the overlap between DEGs and NRGs. (D) Group comparison plot of NRDEGs between the IDD and Control groups in the Combined Datasets.(E) Simplified heatmap displaying the expression of NRDEGs between the IDD and Control groups in the Combined Datasets. (F) Simplified heatmap showing the expression of NRDEGs between the IDD and Control groups in the dataset GSE34095. (ns, no significance; *p < 0.05, **p < 0.01, ***p < 0.001).

Functional and pathway analysis of NRDEGs in intervertebral disc degeneration

To gain insights into the biological functions and molecular pathways associated with NRDEGs, we conducted GO and KEGG enrichment analysis. Notably, NRDEGs were found to be involved in key biological process such as toll-like receptor signaling pathway, toll-like receptor 3 signaling pathway, and cytokine-mediated signaling pathway. In terms of cellular component, NRDEGs were enriched in membrane raft, membrane microdomain, and the external side of the plasma membrane. Additionally, the molecular function analysis highlighted enrichment in functions like tumor necrosis factor receptor superfamily binding, tumor necrosis factor-activated receptor activity, and death receptor activity (Fig. 2A). Furthermore, the KEGG enrichment analysis identified several pathways significantly associated with NRDEGs, including Necroptosis, Influenza A, and TNF signaling pathway (Fig. 2A, B). These findings suggest that NRDEGs play crucial roles in processes related to cell death and immune responses in the context of intervertebral disc degeneration. To provide a comprehensive view of the expression patterns and associations with enriched functions and pathways, we performed joint logFC analysis with GO and KEGG enrichment results for NRDEGs. The circle plots (Fig. 2C) and chord plots (Fig. 2D) effectively demonstrate the relationships between logFC values, enriched functions, and pathways, aiding in the visualization and interpretation of the complex interactions among NRDEGs.

Fig. 2.

Function enrichment analysis of NRDEGs. (A) GO enrichment analysis and KEGG pathway enrichment analysis for NRDEGs. (B) Network graph depicting the relationships between NRDEGs and the results of GO enrichment analysis and KEGG pathway enrichment analysis. (C) Circle plots displaying the results of GO enrichment analysis and KEGG pathway enrichment analysis for NRDEGs. (D) Chord plots representing the results of GO enrichment analysis and KEGG pathway enrichment analysis for NRDEGs.

Identification of diagnostic hub genes for IDD

To determine the diagnostic value of the 18 NRDEGs in the Combined Datasets, we performed LASSO regression analysis to construct the diagnostic model for NRDEGs (Fig. 3A,B). The LASSO diagnostic model consists of 10 genes: BID, IFNGR2, IRF1, MAP3K7, PELI1, SLC25A37, SLC25A6, TNF, TNFRSF1A, and UCHL1 (Fig. 3D). Besides, we constructed a SVM model based on the 18 NRDEGs. We identified the number of genes that resulted in the lowest error rate and the highest accuracy, and showed that when the number of genes was 15, the SVM model achieved the highest accuracy (Fig. 3E,F). The 15 selected genes were TNFRSF1A, MAP3K7, UCHL1, IRF1, SLC25A37, FLOT1, CDK9, PPIA, FAS, TNF TNFAIP3, PPIA, FAS, TNF, and TNFAIP3. Next, we performed a Venn diagram of the genes obtained from the two models, resulting in 10 intersecting genes : BID, IFNGR2, IRF1, MAP3K7, PELI1, SLC25A37, SLC25A6, TNF, TNFRSF1A, and UCHL1, which were defined as hub genes (Fig. 3G). Additionally, to assess the functional importance of the hub genes, we conducted functional similarity analysis (Fig. 3H), and we also generated a chromosome localization plot for the hub genes (Fig. 3C).

Fig. 3.

LASSO analysis and SVM analysis. (A) Variable trajectory plot of LASSO model. (B) Selection factor Plot of LASSO Model. (C) Chromosome localization plot of hub genes. (D) Forest plot of the 10 genes in the LASSO model. (E,F) Visualization of the number of genes with the lowest error rate (E) and the highest accuracy (F) obtained by the SVM algorithm. (G) Venn diagram showing the intersection between the genes obtained from the LASSO and SVM algorithms. (H) Box plot of functional similarity analysis for hub genes.

Diagnostic model construction and validation for hub genes in IDD

To determine the diagnostic value of hub genes in the Combined Datasets, we performed logistic regression analysis to construct a diagnostic model for hub genes (BID, IFNGR2, IRF1, MAP3K7, PELI1, SLC25A37, SLC25A6, TNF, UCHL1, TNFRSF1A). Genes with a p-value less than 0.1 were selected and defined as key genes, resulting in 7 key genes: IFNGR2, IRF1, MAP3K7, PELI1, SLC25A6, TNF, TNFRSF1A (Fig. 4A,B). Additionally, we performed Calibration analysis on the logistic model’s column chart (Fig. 4C). We also used DCA to evaluate the diagnostic performance of the logistic regression model’s predictive values (Fig. 4D). Besides, we explored the relative expression of the 7 genes between the control and IDD groups(Table 3).

Fig. 4.

logistic model and diagnosis of key genes. (A) Forest plot of key genes in the logistic model. (B) Nomogram plot of key genes in the logistic model. (C) Calibration plot of the logistic model. (D) DCA plot of the logistic model. (E–T) ROC curves for the logistic model’s predictive values (E), IFNGR2 (F), IRF1 (G), MAP3K7 (H), PELI1 (I), SLC25A6 (J), TNF (K), and TNFRSF1A (L) in the Combined Datasets. (M), IFNGR2 (N), IRF1 (O), MAP3K7 (P), PELI1 (Q), SLC25A6 (R), TNF (S), and TNFRSF1A (T) in the GSE34095.

Table 3.

List of differential expression analysis of key genes.

Gene symbol	Description	logFC	P. value	Adj.P.Val
IFNGR2	Interferon gamma receptor 2	0.538857	0.002374	0.043107
IRF1	Interferon regulatory factor 1	0.682919	0.002482	0.04408
PELI1	Pellino E3 ubiquitin protein ligase 1	− 0.49307	0.003515	0.053309
SLC25A6	Solute carrier family 25 member 6	− 0.36334	0.004066	0.057224
MAP3K7	Mitogen-activated protein kinase kinase kinase 7	− 0.31893	0.005473	0.067762
TNF	Tumor necrosis factor	− 0.34693	0.005776	0.069715
TNFRSF1A	Tumor necrosis factor receptor superfamily, member 1a	0.430809	0.008797	0.08814

Then, we analyzed the diagnostic value of the logistic model and the 7 key genes in both the Combined Datasets and the validation dataset (GSE34095). Based on the Combined Datasets, we plotted ROC curves for the logistic model and the 7 key genes. The ROC curve results showed that the logistic predictive values (Fig. 4E) had a high degree of accuracy in diagnosing the Combined Datasets. Among the key genes, the expression level of IFNGR2, IRF1, MAP3K7, PELI1, SLC25A6, TNF, and TNFRSF1A (Fig. 4F–L) showed certain diagnostic accuracy in the Combined Datasets. For the GSE34095 dataset, we also plotted ROC curves for the logistic model and each of the 7 key genes. The ROC curve results showed that the logistic predictive values (Fig. 4M) had a certain degree of diagnostic accuracy in the GSE34095 dataset. Among the key genes, the expression level of IFNGR2 (Fig. 4N) showed relatively lower diagnostic accuracy in the GSE34095 dataset, while the expression levels of IRF1, MAP3K7, PELI1, SLC25A6, TNF, and TNFRSF1A (Fig. 4O–T) exhibited certain degrees of diagnostic accuracy.

Enrichment analysis of the combined dataset based on the logistic model

Based on the logistic model, we quantitatively evaluated the information of each IDD patients in the Combined Datasets. First, we performed differential expression analysis on the Combined Datasets with normal samples removed from the logistic model. Additionally, to determine the impact of gene expression levels on IDD occurrence based on the logistic model, we conducted GSEA to analyze the relationship between gene expression in the Combined Datasets and the involved biological process, cellular component, and molecular function (Fig. 5A). The outcomes showed significant enrichment of genes in various pathways, including ferroptosis, hypoxia, apoptosis via trail up, EMT breast tumor up, reactome hedgehog on state, RB1 and TP53 targets up, and IL6 signaling up (Fig. 5B–H).

Fig. 5.

GSEA enrichment analysis based on logistic model. (A) Ridge plot showing the top 7 enriched biological functions in the GSEA of all IDD samples’ genes in the Combined Datasets. (B–H) Differential expression analysis results of all IDD samples in the Combined Datasets, showing significant enrichment of DEGs in the following pathways: ferroptosis (B), hypoxia (C), apoptosis via trail up (D), EMT breast tumor up (E), reactome hedgehog on state (F), RB1 and TP53 targets up (G), and IL6 signaling up (H).

Constructing the interaction network of mRNA-miRNA, mRNA-TFs, mRNA-drugs and functionally similar genes

Next, we predict the miRNAs that interact with the 7 key genes (IFNGR2, IRF1, MAP3K7, PELI1, SLC25A6, TNF, TNFRSF1A) at the mRNA-miRNA level (Fig. 6A). Based on the mRNA-miRNA interaction network, we identified a total of 78 miRNA molecules that interact with the 7 key genes, resulting in 86 pairs of mRNA-miRNA interactions.

Fig. 6.

The interaction network of mRNA-miRNA, mRNA-TF, mRNA-drugs and functionally similar genes. (A) mRNA-miRNA interaction network between key genes and miRNA. (B) mRNA-TF interaction network between key genes and TFs. (C) mRNA-drugs interaction network between key genes and drugs. (D) Functional interactome network of key genes predicting genes with similar functions.

To find TFs that interact with the 7 key genes, we searched the CHIPBase and hTFtarget databases. We downloaded the interaction data and obtained a total of 136 TFs that interact with the 7 key genes, resulting in 306 pairs of mRNA-TF interactions (Fig. 6B).

Furthermore, we identifies potential drugs or small molecules targeting the 7 key genes (Fig. 6C). 67 potential drugs or molecular compounds corresponding to 3 key genes (IRF1, TNF, SLC25A6), resulting in a total of 69 pairs of mRNA-drugs interactions. Among these interactions, the key gene TNF showed the highest number of interactions, targeting 66 different drugs. Besides, we predicted and constructed the functional interactome network of genes with similar functions to these 7 key genes (Fig. 6D).

Analysis of key genes in immune cell infiltration

We calculated the correlation between 22 types of immune cells and the logistic prediction values of IDD samples in the Combined Datasets. Among them, only 16 types of immune cells showed significant correlations. Based on the results of immune cell infiltration analysis, the relative percentage of immune cells in the high and low prediction value groups of the Combined Datasets are shown in Fig. 7A,B. Next, we observed the correlations between immune cells and the 7 key genes and identified the top 2 genes with positive and negative significant correlations with each immune cell type(FIGURE 7C). Subsequently, we illustrated the correlations between TNF and immune cell Macrophages M2 (Fig. 7D), SLC25A6 and immune cell NK cells activated (Fig. 7E), TNFRSF1A and immune cell T cells regulatory (Tregs) (Fig. 7F), and IFNGR2 and immune cell T cells CD8 (Fig. 7G).

Fig. 7.

Immune infiltration analysis based on logistic model. (A) Stacked histogram of infiltrating abundance of immune cells in the Combined Datasets. (B) Simplified value heatmap of the infiltrating abundance of immune cells in the Combined Datasets. (C) Correlation heatmap between immune cells and key genes in the Combined Datasets. (D) Scatter plot of correlation between TNF and immune Macrophages M2. (E) Scatter plot of correlation between SLC25A6 and immune NK cells activated. (F) Scatter plot of correlation between TNFRSF1A and immune T cells regulatory(Tregs). (G) Scatter plot of correlation between IFNGR2 and immune CD8⁺ T cells. (ns, no significance; *p < 0.05, **p < 0.01, ***p < 0.001).

Expression validation of key genes in clinical samples and cellular model of necroptosis in rat nucleus pulposus cells

To further confirm the expression of key genes identified in IDD, we collected human nucleus pulposus tissues with different degrees of degeneration. According to the Pfirrmann classification, grades I and II were classified as the low degeneration group, whereas grades IV and V were classified as the high degeneration group (Fig. 8A). Hematoxylin and eosin (H&E) staining and Safranin O/Fast green staining revealed a disordered arrangement of collagen fibers, decreased cell numbers, and obvious aggregation in the high-degeneration group than in the low-degeneration group (Fig. 8B,C). Immunohistochemical staining showed significantly higher levels of the necroptosis marker p-MLKL and IRF1 protein expression in the high degeneration group compared to the low degeneration group (Fig. 8D–G). RT-qPCR results from human nucleus pulposus tissues demonstrated significantly higher expression of IFNGR2, IRF1, PELI1,CLC25A6, and TNF in the high degeneration group (Fig. 8H).

Fig. 8.

Expression validation of key genes in clinical samples and cellular model of necroptosis in rat nucleus pulposus cells. (A) Low and high degeneration groups defined according to Pfirrmann classification. (B) HE staining of low degeneration group and high degeneration group. (C) Safranin O/Fast green of low degeneration group and high degeneration group. (D) Immunohistochemical staining of necroptosis marker molecule p-MLKL in low degeneration group and high degeneration group. (E) Analysis of Immunohistochemical staining of p-MLKL in low degeneration group and high degeneration group. (F) Immunohistochemical staining of IRF1 in low degeneration group and high degeneration group. (G) Analysis of Immunohistochemical staining of IRF1 in low degeneration group and high degeneration group. (H) The mRNA expression of 7 key genes in human nucleus pulposus tissue with different degrees of degeneration (low degeneration group and high degeneration group, n = 6). (I) The mRNA expression of 7 key genes in the rat nucleus pulposus cell necroptosis model, n = 3. (J) The mRNA expression of necroptosis marker molecules RIPK3 and MLKL in the rat nucleus pulposus cell necroptosis model, n = 3. (K) Western blot analysis of p-MLKL and IRF1 in the rat nucleus pulposus cell necroptosis model, n = 3. (L) Western blot analysis of p-MLKL and IRF1 in the rat nucleus pulposus cell necroptosis model, (n = 3, mean ± SD).

To further verify the expression of these key genes in rat cell models, we used TSZ to construct a classic cell necroptosis model^62,63. RT-qPCR results from the rat cell models showed upregulation of IFNGR2, IRF1, MAP3K7, PELI1, and CLC25A6 in the TSZ group, whereas TNFRSF1 expression was downregulated (Fig. 8I). The mRNA expression levels of necroptosis molecular markers confirmed successful model construction, with RIPK3 and MLKL being significantly upregulated in the degeneration group (Fig. 8J). Western blotting results showed significantly higher expression of p-MLKL and IRF1 proteins in the necroptosis cellular model of rat nucleus pulposus cells (Fig. 8K, L).

Notably, IRF1 was upregulated in both high-degeneration patients and necroptosis cell models, showing a more significant trend than other key genes, consistent with bioinformatics predictions. Therefore, we selected IRF1 for further in-depth analysis.

IRF1 knockdown inhibited necroptosis, mitochondrial ROS levels and promoted ATP in rat nucleus pulposus cells necroptosis model

We constructed the TSZ-induced model group and normal group of rat nucleus pulposus cells treated with both sh-NC and sh-IRF1 viral infection (Supplementary Fig. S3). Western blot analysis of necroptosis markers demonstrated that IRF1 knockdown resulted in a reduction in the protein expression levels of the markers in the TSZ-induced necroptosis model group of rat nucleus pulposus cells (Fig. 9A). Subsequently, protein quantification was performed for the marker molecules (Fig. 9B–G). The expression of p-RIPK1 (Fig. 9B, p = 0.0089), p-RIPK3 (Fig. 9D, p = 0.005), MLKL (Fig. 9F, p = 0.0089), and p-MLKL (Fig. 9G, p = 0.0232) was notably diminished in comparison to the sh-NC-treated group. Mitochondrial ROS assay results indicated that IRF1 knockdown reduced mitROS levels compared to the sh-NC-treated group in both the necroptosis and normal model groups (Fig. 9H). Furthermore, in the TSZ-induced necroptosis model, PI staining demonstrated that IRF1 knockdown markedly impeded necroptotic cell death, with a statistically significant reduction in the percentage of cells undergoing this process compared to the sh-NC group (Fig. 9I,J). ATP level assays revealed that TSZ-induced necroptosis markedly decreased ATP levels, while IRF1 knockdown significantly elevated ATP levels (Fig. 9K).

Fig. 9.

IRF1 knockdown inhibits necroptosis, reduces mitochondrial ROS levels, and promotes ATP production in a rat nucleus pulposus cell necroptosis model. (A–G) Western blot analysis of necroptosis markers in the IRF1 knockdown and control groups (n = 3): (A) Overall protein expression, (B) p-RIPK1, (C) RIPK1, (D) p-RIPK3, (E) RIPK3, (F) p-MLKL, and (G) MLKL. (H) mitochondrial ROS level in the IRF1 knockdown rat nucleus pulposus cell necroptosis model and control group. (I) PI staining in the IRF1 knockdown rat nucleus pulposus cell necroptosis model and control group. (J) The analysis of PI staining in the IRF1 knockdown rat nucleus pulposus cell necroptosis model and control group. n = 3. (K) The ATP level in the IRF1 knockdown rat nucleus pulposus cell necroptosis model and control group. (n = 3, mean ± SD).

IRF1 overexpression promoted necroptosis, mitochondrial ROS levels and reduced ATP in rat nucleus pulposus cells model

We treated rat nucleus pulposus cell groups with both oe-NC and oe-IRF1 viral infection to construct IRF1 overexpression (oe-IRF1) and normal (oe-NC) groups (Supplementary Fig. S4). Western blot analysis of necroptosis markers revealed that IRF1 overexpression increased the protein expression levels of these markers in rat nucleus pulposus cells (Fig. 10A). Specifically, the expression levels of p-RIPK1 (Fig. 10B, p = 0.0446), p-RIPK3 (Fig. 10D, p = 0.0274), MLKL (Fig. 10F, p = 0.0285), and p-MLKL (Fig. 10G, p = 0.0022) were significantly higher than those in the oe-NC group, the expression levels of RIPK1 (Fig. 10C) and RIPK3(FIGURE 10E) were not changing. Mitochondrial ROS detection results indicated that IRF1 overexpression increased mitROS levels compared with those in the oe-NC treatment group (Fig. 10H). Moreover, PI staining showed that IRF1 overexpression significantly promoted necroptosis, with a statistically significant increase in the percentage of cell death compared with that in the oe-NC group (Fig. 10I,J). The ATP level assay revealed that IRF1 overexpression significantly reduced ATP levels compared with those in the oe-NC group (Fig. 10K).

Fig. 10.

IRF1 overexpression promotes necroptosis, increases mitochondrial ROS levels, and reduces ATP levels in the rat nucleus pulposus cell model. (A–G) Western blot of necroptosis markers in the IRF1 overexpression and control groups (n = 3): (A) Overall protein expression, (B) p-RIPK1, (C) RIPK1, (D) p-RIPK3, (E) RIPK3, (F) p-MLKL, and (G) MLKL. (H) mitochondrial ROS level in the IRF1 overexpression rat nucleus pulposus cell model and control group. (I) PI staining in the IRF1 overexpression rat nucleus pulposus cell model and control group. (J) The quantification analysis of PI staining in the IRF1 overexpression rat nucleus pulposus cell model and control group. n = 3. (K) The ATP level in the IRF1 overexpression rat nucleus pulposus cell model and control group. (n = 3, mean ± SD).

Discussion

IDD is a complex pathological process affecting the spinal discs, leading to various clinical manifestations and challenges in the management of patients with spinal-related conditions. Presently, it is acknowledged that numerous genetic and environmental risk factors, including smoking, aging, trauma and occupational exposure, can contribute to IDD^64–68.

Cell death is a complex process that includes various types of PCD, which is Closely related to the development of IDD. Specifically, Yang et al. in SD rats showed that ferritinophagy-mediated ferritin degradation and subsequently lipid peroxidation are associated with ferroptosis in IDD⁶⁹. Our previous study firstly validated the occurrence of disulfidptosis in human nucleus pulposus cells under glucose starvation⁷⁰. And some studies found that pyroptosis is associated with the inflammatory response during IDD⁷¹. Necroptosis, a form of regulated cell death, has garnered increasing attention in various pathological conditions, including neurodegeneration and tissue injury⁷². In the context of IDD, necroptosis appears to be triggered by various stressors, including mechanical loading, oxidative stress, and proinflammatory cytokines, leading to the loss of functional intervertebral disc cells^73,74. Necroptosis contributes to the release of damage-associated molecular patterns (DAMPs) and proinflammatory signals, exacerbating the local inflammatory microenvironment within the disc. Furthermore, necroptotic cell death disrupts tissue homeostasis, impairs extracellular matrix integrity, and accelerates disc degeneration³⁰. However, to date, there are few studies on the role of necroptosis-related genes in IDD. In this study, we conducted bioinformatics analysis to investigate the role of necroptosis related genes in IDD, aiming to shed light on the underlying mechanisms and potential therapeutic targets.

Though comprehensive analysis, we identificated the seven signature genes (IFNGR2, IRF1, MAP3K7, PELI1, SLC25A6, TNF, and TNFRSF1A) using LASSO and SVM-RFE algorithms suggests their potential as key regulators in the necroptosis related pathogenesis of IDD. Several of these genes have been previously associated with inflammatory responses and cell death pathways, supporting their relevance in disc degeneration⁷⁵. Interferon Gamma Receptor 2 (IFNGR2), is a critical protein receptor involved in the immune response within the human body⁷⁶. Although there is no specific information on IFNGR2’s direct involvement in IDD, it’s possible that immune-related factors, including cytokines and signaling pathways similar to those involving IFN-γ (which binds to IFNGR2)⁷⁷, could be relevant in the context of inflammation and tissue degeneration within the intervertebral discs. Even though the functions of MAP3K7, PELI1, SLC25A6, TNFRSF1A in IDD have not been reported, these genes play crucial roles in immunity, inflammation, and cell growth, and they may all play an important role in IDD in which necroptosis participates.

Nevertheless, the mere bioinformatics analysis and literature review cannot substantiate the assertion that these pivotal genes are causally linked to necroptosis in the context of IDD. Consequently, we undertook further experimental validation. we collected degenerated human nucleus pulposus tissues and established a necroptosis model of rat nucleus pulposus cells. In order to demonstrate that the human intervertebral disc samples collected truly reflect the characteristics of low or high degeneration, we not only employed MRI as a reference, but also conducted HE and saffron fast green staining on the samples to further substantiate the accuracy of the clinical sample grouping. Furthermore, in order to ascertain whether necroptosis occurs in degenerated human intervertebral disc tissue and whether the cell model has been successfully established, immunohistochemical staining of the necroptosis marker p-MLKL and PCR verification of RIPK3 and MLKL were performed. The results demonstrated a markedly elevated incidence of necroptosis in the highly degenerated human intervertebral disc tissue in comparison to the low degeneration group. The PCR results for the cell model also demonstrated the successful establishment of a necroptosis cell model. Under this premise, we validated the mRNA expression of 7 key genes through RT-qPCR from both human and rat perspectives. Interestingly, the mRNA expression of IRF1 was significantly up-regulated in both clinical samples and cell models, which was consistent with the expression trend of bioinformatics analysis. At the same time, we were surprised to find that the up-regulation of IRF1 expression was the most significant among all key genes. Therefore, we further verified the expression of IRF1 at the protein level. What is exciting is that the protein expression of IRF1 was significantly up-regulated both in highly degenerated human nucleus pulposus tissue with obvious necroptosis and in rat nucleus pulposus cells with necroptosis. so we followed up with the validation and investigation of IRF1.

IRF1 is a transcription factor that regulated a variety of biological functions, including cellular responses associated with oncogenesis. It has been implicated in the regulation of cell death mechanisms and has been implicated in several types of cancer caused by genetic mutations and functional alterations⁷⁸. In colorectal cancer, IRF1 has been identified as a regulator of PANoptosis (a synergistic and interconnected pathway involving focal death, apoptosis and necrotic apoptosis), which is essential for preventing tumourigenesis. Studies have shown that mice deficient in IRF1 have an increased susceptibility to colon tumours and the cell death defect observed is PANoptosis⁷⁸. In cells infected with influenza A virus (IAV), IRF1 has been shown to promote the activation of NLRP3 inflammasomes and the occurrence of necroptosis. In IRF1-deficient IAV-infected cells, both apoptosis and necroptosis are significantly reduced. Additionally, the authors have demonstrated that IRF1, as a transcriptional regulator of ZBP1, plays a critical role in IAV-induced cell inflammation and necroptosis, further indicating the pivotal role of IRF1 as a key molecule regulating necroptosis in diseases⁷⁹.

In intervertebral disc degeneration, one study used CRISPR epigenome editing to modulate TNFR1/IL1R1 signalling in pathological human IVD cells. The results showed that IRF1 and other transcription factors are key regulators of inflammatory signalling in IVD cells, suggesting IRF1 is a potential new target for gene therapy strategies in disc degeneration⁸⁰. However, there is no direct evidence for a clear function of IRF1 in necroptosis of nucleus pulposus cells. To further elucidate the function of IRF1, we constructed IRF1 knockdown and over-expression models, which revealed that IRF1 promoted necroptosis in rat nucleus pulposus cells, increased mitochondrial ROS level, and decreased ATP level. We found IRF1 knockdown decreased MLKL/p-MLKL expression in TSZ-induced necroptosis rat nucleus pulposus cell model, and IRF1 overexpression promoted MLKL/p-MLKL expression in rat nucleus pulposus cell model. Of interest, Xiong et al. showed that BRD4, IRF1, P-TEFb and RNA polymerase II form a transcriptional complex to regulate MLKL expression, knockdown of IRF1 decreased MLKL expression⁶². Also, Hannes et al. found that type I and II IFNs are known to upregulate MLKL in an IRF1-dependent manner, suggesting that upregulation of IRF1 might affect necroptosis⁸¹. However, the specific molecular mechanism of how IRF1 regulates necroptosis of nucleus pulposus cells needs further experiments to confirm.

Furthermore, our findings suggest a potential association between IRF1 and mitochondrial ROS and cellular ATP levels, which may indirectly influence nucleus pulposus cell senescence. Cellular senescence may represent a significant pathological process in the context of intervertebral disc degeneration⁸². As a consequence of the natural ageing process, intervertebral disc cells may gradually lose their ability to proliferate and differentiate, which can lead to the degeneration and functional decline of disc tissues. These conditions may also result in reduced cell activity and cell cycle arrest. Furthermore, our findings suggest a potential association between IRF1 and mitochondrial ROS and cellular ATP levels, which may indirectly influence nucleus pulposus cell senescence. The precise role of IRF1 in disc degeneration through its impact on the cell cycle remains uncertain; however, it represents a promising avenue for further investigation. So, These studies corroborated our results and we will follow up with an in-depth study of the molecular mechanism of IRF1.

Nevertheless, we must acknowledge certain inherent limitations of our study. Firstly, it should be noted that our study did not utilise our own clinical samples for microarray or RNA sequencing, also due to the number of patients with low-degeneration discs requiring discectomy is very small, so the number of immunohistochemical specimens is not enough. We plan to continue collecting such specimens in future studies. Secondly, the IDD cell model employed may not accurately reflect the intricate in vivo IDD microenvironment. Ultimately, our cellular and molecular biology experiments were unable to elucidate the precise molecular mechanism through which IRF1 promoted necroptosis in IDD. Further research is required in order to gain a more comprehensive understanding.

Conclusion

In conclusion, our comprehensive bioinformatic analysis of necroptosis-related genes in IDD has yielded valuable insights into disease mechanisms. We observed significant upregulation of IRF1 at both the mRNA and protein levels during disc necroptosis, which was subsequently validated using clinical samples and cellular models. This finding highlights the potential of IRF1 as a crucial biomarker of necroptosis in IDD. To further elucidate the function of IRF1, we constructed IRF1 knockdown and overexpression cell models. These experimental systems revealed that IRF1 promotes necroptosis in rat nucleus pulposus cells, enhances mitochondrial ROS levels, and decreases ATP levels. Collectively, these findings establish that IRF1 is a key player in IDD-associated necroptosis, which may facilitate early disease detection and the development of targeted interventions. Further research is warranted to explore the clinical applicability of IRF1 as a diagnostic tool and investigate potential therapeutic approaches targeting IRF1-mediated necroptosis in IDD.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

Haihong Zhang designed and supervised the study. Kaisheng Zhou, Shaobo Wu and Zuolong Wu performed the experiments, analyzed the data, and wrote the manuscript. Rui Ran assisted in cell experiments (RT-PCR and Western blot). Wei Song and Hao Dong assisted in collecting clinical samples and cell culture. All authors contributed to the article and approved the submitted version.

Funding

This work was supported by the National Natural Science Foundation of China (No. 31960175), Natural Science Foundation of Gansu Province (23JRRA0960) and Lanzhou University Second Hospital Cuiying Youth Fund Project (CY2021-QN-A03).

Data availability

The original contributions presented in the study are included in the article. Further inquiries can be directed to the corresponding author.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Ethics Committee of Lanzhou University Second Hospital (No.2024 A-651) for studies involving humans, and all patients signed informed consent. Also, the study was conducted in accordance with the Declaration of ARRIVE guidelines and confirmed that all experimental protocols were approved by the Institutional Ethics Committee of Lanzhou University Second Hospital (No.D2024-609) for studies involving rats, and all methods were carried out in accordance with relevant guidelines and regulations.