Examining the Involvement of Ferroptosis-Related Genes in Ankylosing Spondylitis and the Infiltration of Immune Cells

doi:10.21203/rs.3.rs-4231978/v1

Download PDF

Research Article

Examining the Involvement of Ferroptosis-Related Genes in Ankylosing Spondylitis and the Infiltration of Immune Cells

https://doi.org/10.21203/rs.3.rs-4231978/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Ankylosing Spondylitis (AS) is a chronic inflammatory disease which is characterized by pain and progressive stiffness and which spinal and sacroiliac joints are mainly affected, with insidious onset, high rates of disability among patients, unknown pathogenesis, and no effective treatment. Ferroptosis is a regulated form of cell death that is important for normal development and tissue homeostasis. However, its relation to AS is not clear. In this study, we identified two potential therapeutic targets for AS based on genes associated with ferroptosis and explored their association with immune cell infiltration (ICI) and immune cells. We studied gene expression profiles of two cohorts of patients with AS (GSE73754 and GSE41038) derived from the gene expression omnibus database at NCBI, and ferroptosis-associated genes (FRGs) were obtained from the FerrDb database. LASSO regression analysis was performed to estimate predictive factors for AS based on FRGs, and the ferroptosis level in each sample was performed via single-sample gene set enrichment analysis. Weighted gene co-expression network analysis (WGCNA) and protein-protein interaction (PPI) network analyses were assessed. The relationship between key genes and ICI levels was assessed using the CIBERSORT algorithm, followed by gene ontology and Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses. These results suggest that ALKBH5 and NDUFA12 might serve as potential diagnostic biomarkers and targets for AS. And both was negatively correlated with the infiltration levels of several different types of immune cells. In conclusion, ALKBH5 and NDUFA12 may induce ferroptosis in the cells of patients with AS via changes in the inflammatory response in the immune microenvironment, and these genes could serve as molecular targets for AS therapy.

Ankylosing spondylitis

Ferroptosis

immunity

ALKBH5

NDUFA12

Our study found that ferroptosis plays a crucial role in AS.
ALKBH5 and NDUFA12 was negatively correlated with the infiltration levels of several different types of immune cells.
ALKBH5 and NDUFA12 may induce ferroptosis in the cells of patients with AS via changes in the inflammatory response in the immune microenvironment.
ALKBH5 and NDUFA12 serve as potential diagnostic biomarkers and targets for AS.

Ankylosing spondylitis (AS) is a chronic inflammatory disease characterized by inflammation of the sacroiliac joints and spinal entheses(1). It is often beginning before age 40, with male predominance, and a prevalence of roughly 0.5% in the United States(2). Extra-axial manifestations include acute uveitis, peripheral arthritis, enthesitis (inflammation of where tendons insert on bones), psoriasis, aortic root, and gut inflammation(3). Erosions and sclerosis are early signs of radiographic progression that may precede the development of syndesmophytes, which can eventually grow and bridge adjacent vertebrae, ultimately leading to spinal fusion or the characteristic bamboo spine in some patients (4, 5). Due to the insidiously progressive nature of AS, patients may fail to notice the disease in its early stages, which may result in disease progression and missing the best treatment window, years of pain and the rigid stooped posture incur significant disability and economic cost(6–8). The early treatments for AS include non-steroidal anti-inflammatory drugs (NSAIDs) combined with physical therapy and exercise. Moderate to high level quality evidence supports a clinically important benefit of anti-TNF compared with placebo for improvement in disease activity and function and achieving partial remission in ankylosing spondylitis in the short term. Surgical interventions are necessary for patients experiencing severe symptoms(9–12). The prognosis individuals with AS can differ greatly. It would be difficult to cure the disease without having a complete understanding of its pathogenesis, and the progression of the disease can be slowed only with targeted treatment. Therefore, it is essential to have a thorough comprehension of the AS pathogenesis.

Despite AS having a strong hereditary component, the specifics of its development remain unclear. HLA-B27 has been identified as the primary genetic contributor to the development of AS since its initial discovery in 1973. Despite this, the polygenic aspect of AS has been confirmed and scientists are gradually uncovering new molecular pathways. The discovery of the IL-23/IL-17 pathway, revealed new treatment targets along with molecules and signaling pathways that proved pivotal in the pathophysiology, leading to further understanding of the pathogenesis of AS. (13–16). Cellular demise is a process that takes place at various points in life, independent of any physical or pathological stimuli. Maintaining a balance between cell death and survival is crucial for proper development and maintaining in vivo balance. The process of cell death is diverse and can exhibit various characteristics depending on the changes in the physio-pathological environment(17, 18). Various approaches to cell death, including apoptosis, necrosis, pyroptosis, autophagy, and ferroptosis, have been defined by the Nomenclature Committee on Cell Death, taking into account the physical characteristics and biochemical and functional elements of the cell(19). Certain pathological conditions may be prevented using non-apoptotic cell death, which can selectively eliminate and prevent the activation of potential cancerous cells. Ferroptosis is a novel mechanism of regulated cell death (RCD). The concept of ferroptosis, a distinct form of cell death, connects various components of cell metabolism such as iron, selenium, amino acids, lipids, and redox chemistry. The induction of cell death in ferroptosis relies on an excess of free ferrous ions and the catalysis of lipid peroxidation of abundant unsaturated fatty acids on the cell membrane, aided by lipoxygenase(20–23). Ferroptosis is characterized by a decrease in mitochondrial size and an increase in membrane density, but no visible alterations in the nucleus, setting it apart from other types of cell death. It also differs from apoptosis, necrosis, and pyroptosis with respect to biochemistry and gene expression(24, 25). A growing body of evidence suggests that malfunctions in ferroptosis are linked to multiple diseases, although no connection has been established with AS.

Using multiple bioinformatic techniques, our study explored the biological processes involved in AS pathogenesis. The gene expression patterns of individuals with AS and those in good health were obtained from the gene expression omnibus (GEO) database. The AS transcriptome data was then examined and studied in detail. The treatment and control groups were compared, and differential expression of genes (DEGs) was observed. Predictive models were built using LASSO regression analysis, focusing on ferroptosis-related genes (FRGs). The level of ferroptosis in each sample was evaluated using single sample gene enrichment analysis (ssGSEA), and weighted gene co-expression network analysis (WGCNA) was utilized to identify potential biomarkers or therapeutic targets. A protein-protein interaction (PPI) network was established to elucidate prospective target genes involved in AS. The connection between key genes and levels of immune cell infiltration in AS was evaluated using CIBERSORT, allowing for a better understanding of their function and correlation in the immune microenvironment. Ultimately, the key genes were subjected to gene ontology (GO) and pathway analyses through the Kyoto Encyclopedia of Genes and Genomes (KEGG) to anticipate their biological functions. Through these studies, we have sought to bring attention to the transcriptome distinctions between phenotypes and disease, revealing a potential connection between ferroptosis and AS. The findings of this study could inform the development of strategies for AS treatment and the evaluation of potential molecular targets.

Data acquisition and pre-processing

The Gene expression profile data of patients with AS were obtained from datasets GSE 73754(26)(https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE73754) and GSE41038 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE41038) in the GEO database (27). Dataset GSE73754 (GPL10558 Illumina HumanHT-12 V4.0 expression beadchip) comprised data corresponding to 72 samples in total, with fifty-two AS samples and twenty normal control samples. Dataset GSE41038 (GPL6883 Illumina HumanRef-8 v3.0 expression beadchip) comprised the data of 15 samples, among which 2 were AS samples, and the other 13 were controls. We normalized and merged the downloaded expression matrices using the normalizeBetweenArrays function in the limma package of R (28). Ferroptosis genes were obtained from the FerrDb database (http://www.zhounan.org/ ferrdb) (29). The FerrDb database comprises 784 articles on ferroptosis downloaded from the PubMed database and the information extracted from the papers on regulatory factors and biomarkers of ferroptosis and corresponding diseases.

Random Forest Model Screening for Characterized Genes

In order to anticipate the impact of oxidative stress genes on AS development, we developed a Random Forest (RF) predictive model utilizing oxidative stress genes that were differentially expressed in AS. The Random Forest technique, used for both classification and regression in machine learning, offers variable importance measures in contrast to a single decision tree. Additionally, it excels at managing datasets with numerous predictor variables, leading to more comprehensive and understandable model outcomes(30, 31).

Construction of predictive model

Differentially expressed FRGs in patients with AS were used to construct predictive models. LASSO regression analysis was performed on the training cohort using the Glmnet package in R(32). The LASSO algorithm could reduce the dimensionality of high- dimensional data and depict the features of the data with a model comprising fewer variables(33). To avoid overfitting the model design of the training set, tenfold cross validation is performed. Finally, the regression coefficients generated using LASSO regression analysis are used to create a scoring system. We split all AS patients into high-risk and low-risk groups using the R package "SurvMiner" threshold, and a univariate COX regression analysis were performed to screen RNA modifier genes that play significant roles, and results were visualized using the package “forestplot” in R.

Characteristic gene enrichment analysis

Gene ontology (GO) functional annotation analysis, which includes biological process (BP), molecular function (MF), and Cellular Component (CC), is a common method for large-scale gene functional enrichment. GO functional annotation analysis of the signature genes was performed using the package ClusterProfiler in R(34), and the significant GO results were further visualized using the package treemap in R. A p-value < 0.05 was considered statistically significant in this study.

Single sample gene enrichment analysis

The principle of ssGSEA is similar to that of GSEA. The aforementioned method is an implementation proposed mainly for single samples that cannot be analyzed with GSEA. Five FRG sets were obtained from the FerrDb database(29) (http://www.zhounan.org/ferrdb/index.html), and the level of ferroptosis in each sample was assessed based on the expression levels of ferroptosis-specific marker genes. ssGSEA analysis was performed using the “GSVA” package in R.

Weighted gene Co-expression network analysis

Weighted correlation network analysis (WGCNA) is a statistical method that can be used to describe the patterns of gene association networks between different samples. It can also be used to identify highly synergistic sets of genes and candidate biomarkers, or therapeutic targets based on the endogeneity of these gene sets and the association between the gene sets and corresponding phenotypes. In contrast to focusing solely on DEGs, in WGCNA, data of thousands of the most highly variable genes or complete genes are utilized to identify gene sets of interest, and their association with phenotypes is examined using significant association analysis(35). We used the WGCNA package in R to establish a weighted gene correlation network and identify the clustering modules based on clinicopathological features and ferroptosis gene sets(36). The correlation among module eigengenes, clinicopathological features, and ferroptosis gene sets was evaluated to identify highly correlated modules. Gene significance (GS) > 0.5 and module membership (MM) > 0.7 were used as the screening parameters.

Construction of the protein-protein interaction (PPI) network

PPI networks are composed of interactions between different individual proteins. Individual proteins interact with one another in protein interaction networks, which involves various aspects of biological signal transmission, gene expression management, energy and chemical metabolism, and cell cycle control. A systematic analysis of the interaction between many proteins in biological systems is required to understand the proteins co-function in biological systems, the reaction mechanism of biological signals, and energy metabolism under special physiological states such as diseases, as well as the functional relationships between proteins. Spearman correlation analysis was conducted between characteristic FRGs and screened module genes with the screening conditions of cor > 0.4 and p < 0.01. The ferroptosis related AS characteristic gene-module gene network was visualized using Cytoscape (version: 3.9.0). The gene with the highest node degree in the network was selected as the hub gene for this study.

GSEA enrichment analysis

GSEA (Gene Set Enrichment Analysis) was used to evaluate the distribution trend of genes in a predefined Gene Set in the Gene list sorted by phenotypic relevance, and thus to judge its contribution to phenotype(37). We obtained “C2.kegg.v7.4.symbols” and “C5.go.v7.4.symbols” gene sets from the MSigDB database. GSEA of hub genes was performed with two gene sets separately, and used the "ClusterProfiler" R package for GSEA analysis(34). P value < 0.05 was considered to be statistically significant (38).

Key genes and immune infiltration analysis

CIBERSORT is a deconvolution technique of expression matrices of immune cell subtypes based on the premise of linear support vector regression. CIBERSORT was originally used for the analysis of tumor microenvironment (TME) and is now being increasingly applied in the characteristic analysis of immune infiltration in non-tumor tissues (39). Immune cell infiltration (ICI) analysis has an important guiding role in disease research and prognosis prediction. RNA-Seq data were used to estimate the infiltration of immune cell in the AS and control groups(40). CIBERSORT algorithm was performed to analysis ICI between AS an control samples. Pearson correlation coefficients between the expression of key genes and immune cell infiltration levels were calculated to assess the relationship between key genes and ICI.

Identification of characteristic immune cell subtypes and hub gene validation

Based on the results of CIBERSORT, “ConsensusClusterPlus “package in R (41)(http://www.bioconductor.org/packages/release/bioc/html/ ConsensusClusterPlus.html) was used for clustering analysis of AS samples, with repetitions set to 50 (reps = 50) and a resampling rate of 80% (pItem = 0.8). Samples were classified into different groups based on the number of immune cells present in each sample. To validate these groupings, principal component analysis (PCA) was performed to examine the expression of all genes, and results were visualized with the “ggplot2” package in R. The expression of hub genes among different groups was also evaluated.

General flow of experimental design

Samples were downloaded from the GEO database from ankylosing spondylitis and normal tissues, differential genes were obtained and analyzed for functional enrichment, further oxidative stress genes among them were selected for expression profiling, screened and analyzed for correlation with functional enrichment using Random Forests, in addition, we used the WGCNA algorithm to identify and correlate with genes related to the iron death pathway of the strong spine. We further took the intersection of iron death genes with oxidative stress genes and used Lasso regression modelling to screen key genes to identify diagnostic markers, and correlation analysis of key genes with molecular subtypes and infiltrating immune cells.

Analysis of different expression characteristics of RNA modification factors in AS and control samples

Differential gene expression analysis was firstly performed on integrated data (Figs. 1A, B) using the limma package to evaluate the effect of gene expression levels in the AS tissue compared to in normal tissues.

To compare the functional variations between ankylosing spondylitis and normal samples, we conducted GSEA and GSVA enrichment analyses on each group separately (Fig. 2). GSEA enrichment results show that the GO enriched in AS patients include inner ear morphogenesis, potassium channel complex, chemokine activity, hormone activity, and voltage gated cation channel activity (Fig. 2A), GO enriched in normal samples include biological process involved in symbiotic interaction, immune effector process, response to biotic stimulus, response to endogenous stimulus, and secretion (Fig. 2B), KEGG pathways that are predominantly enriched in AS patients include alzheimers disease, huntingtons disease, pathways in cancer, and regulation of actin cytoskeleton (Fig. 2C). GSVA enrichment results showed that the main GO enriched in AS patients were phosphatidylinositol 3 kinase complex, phosphatidylinositol phosphorylation, lipid phosphorylation, cellular response to virus, T cell homeostasis, and leukocyte homeostasis (Fig. 3D), the KEGG in AS patients were butanoate metabolism, valine leucine and isoleucine degradation, cardiac muscle contraction, pyruvate metabolism, and propanoate metabolism (Fig. 2E).

Thirteen differentially expressed FRGs were identified in the AS tissue compared with those in the normal tissues (Figs. 3A, B, D) including four genes with upregulated expression and nine genes with downregulated expression. The following genes showed upregulated expression: GCLC, PTGS1, SESN3 and SOD2, while the following genes showed downregulated expression: ALKBH3, ALKBH5, COX4l1, GLRX2, GPX7, NDUFA12, PARK7, PRDX1 and PRDX3. Co-expression analysis showed that the expression of ALKBH3 was significantly negatively correlated with the expression of most DEGs, including PRDX3, PARK7, SESN3, PTGS1, GLRX2, and GCLC (Fig. 3C).

Construction of AS prediction model

By employing a random forest approach, we created a predictive model for ankylosing spondylitis that incorporated gene expression data linked to oxidative stress. We developed predictive models and performed functional enrichment analyses using the expression of 13 genes each (Figs. 4 and 5). As shown in Fig. 5A, the main GO enrichment for differentially expressed RNA modifiers was in DNA repair, electron transfer activity, electron transport chain, oxidative demethylation, and oxidative phosphorylation. And shown in Fig. 5B and Fig. 5C, the main pathway enrichment for differentially expressed RNA modifiers was in Oxidative phosphorylation, Pathways of neurodegeneration-multiple diseases, Amyotrophic lateral sclerosis, Arachidonic acid metabolism and Huntington disease. We further analyzed oxidative phosphorylation pathway in detail (Fig. 5).

Weighted gene Co-expression network and module enrichment analyses

WGCNA was performed on the expression profiles of the combined GEO dataset, and the correlation of each color module with ferroptosis was evaluated (Fig. 6A). The turquoise module was significantly negatively correlated with each indicator of ferroptosis, with the drive marker function demonstrating the strongest correlation (− 0.75, P = 3e − 14, Fig. 6C). Therefore, the study was conducted with the blue module of the marker function, which was selected as the subject. In this module, genes (n = 87) with GS > 0.5 and MM > 0.7 were selected for enrichment analysis (Fig. 6B).

Firstly, we used the LASSO regression algorithm to constructed a of based on the expression levels of FRGs (Figs. 7A, B). Secondly, COX regression was used to screen the genes associated with AS, and nine genes was verified, including COX4I1, ALKBH5, PRDX1, SOD2, NDUFA12, PRDX3, PARK7, SESN3, and PTGS1 (Fig. 7C). To narrow down the key genes, we cross-referenced the signature genes, significant module genes from WGCNA, LASSO regression algorithm, and genes with P < 0.05, and obtained the key genes ALKBH5 and NDUFA12 (Fig. 7D).

Screening key genes and single sample gene enrichment analysis

A PPI network of FRGs was constructed to further analyze genes with potential interactions with key genes ALKBH5 and NDUFA12 (Fig. 8). The results showed that 145 genes interacted with ALKBH5, and 58 genes interacted with NDUFA12, of which RPL15, NDUFA4 and HNRNPA2B1 interacted with both key genes.

Correlation analysis of immune cell infiltration and key genes

To further investigate the function of key genes in AS, correlation was analyzed between key genes and the immune microenvironment. The results of the correlation of CIBERSORT results with hub gene expression levels indicated that the expression of ALKBH5 was significantly positively correlated with the infiltration levels of many innate and acquired immune cells, including T cells follicular helper, naive T cell CD4, neutrophils, mast cells resting, M2 macrophages, dendritic cells activated, and naive CD4 T cells. The expression of ALKBH5 was significantly correlated with the infiltration levels of many innate and acquired immune cells, including T cells regulatory, T cells follicular helper, T cells CD4 memory activated, neutrophils, mast cells resting, macrophages M2, eosinophils, dendritic cells activated and naive B cells (Fig. 9).

Ankylosing spondylitis immuno-molecular typing

To further explore the biological characteristics of gene expression in AS tissues, unsupervised consensus clustering (UCC) was applied to all samples based on their immune cell contents. All samples were divided into two subtypes (subtype 1: n = 6; subtype 2: n = 33, Figs. 10A–C), and the results of PCA indicated high separation quality (Fig. 10D). Further differential expression analysis of the hub genes in group 1 and group 2 showed that ALKBH5 and NDUFA12 were significantly differentially expressed in different subgroups (p < 0.001) (Figs. 10E, F). Accordingly, we speculated that ALKBH5 and NDUFA12 could be used as key genes for AS.

Ferroptosis is a recently uncovered form of controlled cell death, distinguished by the oxidation of lipids through iron dependency(20). The idea of ferroptosis has sparked a multitude of studies, all aiming to understand its involvement in different diseases. Numerous studies have specifically shown the correlation between ferroptosis and the management or outcome of different types of cancer(42–44).

Ferroptosis has recently been highlighted by Rong et al. as a potential key player in the pathogenesis and molecular regulation of AS.(45). The revelation of the specific pathways may have been withheld, but it still assists researchers in their subsequent investigations. Our efforts yielded positive results as we successfully detected diagnostic biomarkers of ferroptosis in AS using various datasets and demonstrated its clinical applicability through experimental testing. Nonetheless, the association between ferroptosis and AS has not been brought to light. In the past few years, significant progress has been made in identifying new AS-associated genes, thanks to advancements in high-throughput technologies. This includes genes encoding cytokine receptors, transcription factors, and signaling molecules(14, 46). The pathogenesis of AS is complex. Hence, we employed various bioinformatic and statistical techniques in this study to uncover the molecular and signaling mechanisms involved in AS progression. To investigate the potential role of FRGs in AS and their associated biological functions, we developed predictive models for both AS and FRGs. A functional enrichment analysis was also carried out on the identified key genes. GO-GSEA results revealed a strong connection between DEGs and biological processes like cell death and biotic stimulus responses. Differential expression analysis revealed 13 DEGs related to ferroptosis, consisting of 4 upregulated genes and 9 downregulated genes. Utilizing FRGs and LASSO regression, we developed a predictive model and applied COX regression to pinpoint genes associated with AS. Nine of these genes have been previously linked to AS. As an illustration, ACSL4 is a protein that plays a critical role in the metabolism of fatty acids by facilitating β-oxidation. High levels of fatty acid oxidation contribute to the reduced body fat in individuals with AS. This result could be attributed to the shift from alternative metabolic pathways to fatty acid metabolism in the ligaments of AS patients, as insulin signaling plays an important role(47). Another example e is PRDX1. The presence of proteins responsible for cell death may have a significant impact on the development of AS, and PRDX1 has a significant impact on both osteogenesis and the pathogenesis of AS(48). Patients with AS demonstrate elevated PTGS1 expression, indicating a significant connection between inflammation and bone resorption in the osteoproliferation of the disease and potential treatment implications (49). Additionally, SOD2 has a crucial function in suppressing inflammation and oxidative stress in RA and AS(50, 51).

Utilizing WGCNA, we detected gene clusters within the GEO dataset that exhibited strong correlations with ferroptosis. Specifically, the turquoise module showed significant positive associations with all indicators of ferroptosis, encompassing a total of 87 genes. A PPI network of ferroptosis was constructed based on the STRING database. ALKBH5 and NDUFA12, among other genes, were identified as important RNA modifiers in the network.

ALKBH5, an important participant in m6A methylation modification, has been reported in many tumors (52–56). ALKBH5 is a member of the well-conserved Alkb family of non-heme Fe(II)/α-KG-dependent dioxygenases, which mediate the repair of N-alkylated nucleobases by oxidative demethylation(57). It plays a critical role in regulating the m6A modification and governing multiple cellular functions(58). The demethylation of m6A by ALKBH5 plays a crucial role in regulating gene expression by impacting various aspects of RNA metabolism, such as pre-mRNA processing, mRNA degradation, and translation.

The demethylation enzyme ALKBH5 plays a crucial role in the development of OLF, its overexpression resulted in p-AKT activation, and the ALKBH5/AKT signaling pathway regulated BMP2. This led to the enhancement of osteogenesis in ligamentum flavum cells, as ALKBH5 demethylated BMP2 and activated AKT(59). The research revealed a decline in ALKBH5 mRNA levels in PBMC among AS patients and low expression of ALKBH5 can enhance the osteogenesis of MSCs through activation of the PI3K/AKT pathway(60). The presence of HLA-B27 receptors on T cells was significantly higher in individuals with AS, and these cells also underwent a change to the Th17 phenotype, which is known to play a pivotal role in the development of AS (61). The role of NK cell cytokines in immune response control and potential involvement in the development of AS and other immune-mediated diseases is significant(62). We propose that it is highly likely to have a significant impact on AS.

In addition, NDUFA12 serves as a small, hydrophobic accessory subunit in the transmembrane region of Mitochondrial complex I (CI) but is unlikely to be involved in CI's catalytic activity (63). The NDUFA12-associated disease is characterized by a clinical picture with psychomotor delay/regression, dystonia, and visual impairment(64). Overexpression of NDUFA12 in thyroid cancer was found associated with prolonged survival (65). The link between NDUFA12 and AS has not been extensively studied in existing literature, making this study the first to propose a possible correlation.

One of the main characteristics of AS is the development of pathological osteogenesis. Evidence continues to support the idea that inflammation has a direct impact on the process of pathological osteogenesis, by promoting the generation of osteoinductive proteins and the multiplication of osteoprogenitor cells. In this study, we investigated the correlation between ALKBH5 and NDUFA12 and their impact on the immune response, to gain a deeper understanding of the inflammatory and immune microenvironments in individuals with AS, and to examine the potential roles of these important genes. Initial examination indicated that AS-related DEGs were concentrated in pathways related to responding to living organisms. Increased levels of reactive oxygen species (ROS) have been noted in AS patients' white blood cells. The data suggests that ROS may act as a mediator in AS pathogenesis. The influence of ROS on normal cell growth can vary, and even small amounts of oxygen radicals can trigger RCD, including apoptosis, autophagy, and ferroptosis (66).

Further examination of the immune system showed increased presence of immune cells, including neutrophils, in AS samples. The activation of the innate immune response, a crucial factor in AS, was evident in patients with the disease. HLA-B27-positive controls exhibited heightened sensitivity in their polymorphonuclear leukocytes towards specific chemoattractants. The presence of hyperreactive neutrophils at sites of inflammation could potentially amplify the inflammatory response in individuals with AS (67). Correlation analysis of ICI and gene expression performed using CIBERSORT revealed that the two key genes showed a significant positive correlation with the degree of infiltration of various immune cells, such as naive B cells, eosinophils, M1 macrophages, and naive CD4 T cells. AKLBH5 and Mast cells resting had the highest positive correlation coefficient (R = 0.34), while HSPB1 showed the strongest negative correlation (R = -0.68) with neutrophils. There was a significant positive correlation between the key gene ALKBH5 and Mast cells resting, while a negative correlation was observed with the infiltration of Neutrophils, Dendritic cell activated, and T cells CD4 memory activated in both innate and acquired immunity cells. The degree of infiltration by naive B cells, Macrophages M2, and Mast cells resting showed a significant positive correlation with the key gene NDUFA12, while a significant negative correlation was found with the infiltration of Dendritic cells activated, Neutrophils, and T cells CD4 naive. Therefore, we propose that the role of key genes in AS may be associated with neutrophil infiltration.

In summary, our study found that ferroptosis plays a crucial role in AS. We identified ferroptosis-associated markers in AS diagnosis such as ALKBH5 and NDUFA12. However, further experiments are required for in-depth study of the mechanism to support our conclusion.

Acknowledgments

This is a short text to acknowledge the contributions of specific colleagues, institutions, or agencies that aided the efforts of the authors.

Conflict of Interest

The authors have no conflict of interest.

Data Availability Statement

The datasets generated and/or analyzed during the current study are available in the Gene Expression Omnibus (GEO) DataSets (https://www.ncbi.nlm.nih.gov/geo/).

Funding

This study was supported by the Health Commission of Mianyang City(Grant No. 2020303).

Ethics statement

The studies involving human participants were reviewed and approved by the Ethics Committee of theThird Hospital of Mianyang, Sichuan Mental Health Center (Approval Number: 2023-162). Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements. Written informed consent was not obtained from the individual(s) for the publication of any potentially identifiable images or data included in this article.

Authors’ Contributions

Study concepts: Y.Z., B.T.; Study design: Y.Z., B.T.; Data acquisition: Y.Z., S.Z., Y.W.; Quality control of data and algorithms: B.T., Q.Y.; Data analysis and interpretation: Y.Z., B.T.; Statistical analysis: Y.Z., S.Z., Y.W.; Manuscript preparation: Y.Z., S.Z., Y.W.; All authors have reviewed and edited the manuscript.

Ahn Y, Keum HJ, Shin SH. Percutaneous Endoscopic Cervical Discectomy Versus Anterior Cervical Discectomy and Fusion: A Comparative Cohort Study with a Five-Year Follow-Up. J Clin Med 2020;9.
Smith JA. Update on ankylosing spondylitis: current concepts in pathogenesis. Curr Allergy Asthma Rep 2015;15:489.
Stolwijk C, Essers I, van Tubergen A, et al. The epidemiology of extra-articular manifestations in ankylosing spondylitis: a population-based matched cohort study. Ann Rheum Dis 2015;74:1373–8.
Boonen A, Severens JL. Ankylosing spondylitis: what is the cost to society, and can it be reduced? Best Pract Res Clin Rheumatol 2002;16:691–705.
van der Heijde D, Braun J, Deodhar A, et al. Modified stoke ankylosing spondylitis spinal score as an outcome measure to assess the impact of treatment on structural progression in ankylosing spondylitis. Rheumatology (Oxford) 2019;58:388–400.
Fallahi S, Jamshidi AR. Diagnostic Delay in Ankylosing Spondylitis: Related Factors and Prognostic Outcomes. Arch Rheumatol 2016;31:24–30.
Dean LE, Jones GT, MacDonald AG, et al. Global prevalence of ankylosing spondylitis. Rheumatology (Oxford) 2014;53:650–7.
El Maghraoui A. Extra-articular manifestations of ankylosing spondylitis: prevalence, characteristics and therapeutic implications. Eur J Intern Med 2011;22:554–60.
Ritchlin C, Adamopoulos IE. Axial spondyloarthritis: new advances in diagnosis and management. Bmj 2021;372:m4447.
Haroon N, Inman RD, Learch TJ, et al. The impact of tumor necrosis factor α inhibitors on radiographic progression in ankylosing spondylitis. Arthritis Rheum 2013;65:2645–54.
Taurog JD, Chhabra A, Colbert RA. Ankylosing Spondylitis and Axial Spondyloarthritis. N Engl J Med 2016;374:2563–74.
Pécourneau V, Degboé Y, Barnetche T, et al. Effectiveness of Exercise Programs in Ankylosing Spondylitis: A Meta-Analysis of Randomized Controlled Trials. Arch Phys Med Rehabil 2018;99:383-9.e1.
Brown MA, Wordsworth BP. Genetics in ankylosing spondylitis - Current state of the art and translation into clinical outcomes. Best Pract Res Clin Rheumatol 2017;31:763–76.
Ranganathan V, Gracey E, Brown MA, et al. Pathogenesis of ankylosing spondylitis - recent advances and future directions. Nat Rev Rheumatol 2017;13:359–67.
Paine A, Ritchlin CT. Targeting the interleukin-23/17 axis in axial spondyloarthritis. Curr Opin Rheumatol 2016;28:359–67.
van der Heijde D, Cheng-Chung Wei J, Dougados M, et al. Ixekizumab, an interleukin-17A antagonist in the treatment of ankylosing spondylitis or radiographic axial spondyloarthritis in patients previously untreated with biological disease-modifying anti-rheumatic drugs (COAST-V): 16 week results of a phase 3 randomised, double-blind, active-controlled and placebo-controlled trial. Lancet 2018;392:2441–51.
Bertheloot D, Latz E, Franklin BS. Necroptosis, pyroptosis and apoptosis: an intricate game of cell death. Cell Mol Immunol 2021;18:1106–21.
Wang Y, Kanneganti TD. From pyroptosis, apoptosis and necroptosis to PANoptosis: A mechanistic compendium of programmed cell death pathways. Comput Struct Biotechnol J 2021;19:4641–57.
Kroemer G, Galluzzi L, Vandenabeele P, et al. Classification of cell death: recommendations of the Nomenclature Committee on Cell Death 2009. Cell Death Differ 2009;16:3–11.
Ursini F, Maiorino M. Lipid peroxidation and ferroptosis: The role of GSH and GPx4. Free Radic Biol Med 2020;152:175–85.
Jiang X, Stockwell BR, Conrad M. Ferroptosis: mechanisms, biology and role in disease. Nat Rev Mol Cell Biol 2021;22:266–82.
Stockwell BR, Friedmann Angeli JP, Bayir H, et al. Ferroptosis: A Regulated Cell Death Nexus Linking Metabolism, Redox Biology, and Disease. Cell 2017;171:273–85.
Yuan H, Pratte J, Giardina C. Ferroptosis and its potential as a therapeutic target. Biochem Pharmacol 2021;186:114486.
Fang X, Ardehali H, Min J, Wang F. The molecular and metabolic landscape of iron and ferroptosis in cardiovascular disease. Nat Rev Cardiol 2023;20:7–23.
Tang D, Chen X, Kang R, Kroemer G. Ferroptosis: molecular mechanisms and health implications. Cell Res 2021;31:107–25.
Gracey E, Yao Y, Green B, et al. Sexual Dimorphism in the Th17 Signature of Ankylosing Spondylitis. Arthritis Rheumatol 2016;68:679–89.
Sun S, Fang K, Zhao Y, et al. Increased expression of alpha 1-anti-trypsin in the synovial tissues of patients with ankylosing spondylitis. Clin Exp Rheumatol 2012;30:39–44.
Bolstad BM, Irizarry RA, Astrand M, Speed TP. A comparison of normalization methods for high density oligonucleotide array data based on variance and bias. Bioinformatics 2003;19:185–93.
Zhou N, Bao J. FerrDb: a manually curated resource for regulators and markers of ferroptosis and ferroptosis-disease associations. Database (Oxford) 2020;2020.
Chen X, Ishwaran H. Random forests for genomic data analysis. Genomics 2012;99:323–9.
Speiser JL, Miller ME, Tooze J, Ip E. A Comparison of Random Forest Variable Selection Methods for Classification Prediction Modeling. Expert Syst Appl 2019;134:93–101.
Friedman J, Hastie T, Tibshirani R. Regularization Paths for Generalized Linear Models via Coordinate Descent. J Stat Softw 2010;33:1–22.
Gui J, Li H. Penalized Cox regression analysis in the high-dimensional and low-sample size settings, with applications to microarray gene expression data. Bioinformatics 2005;21:3001–8.
Yu G, Wang LG, Han Y, He QY. clusterProfiler: an R package for comparing biological themes among gene clusters. Omics 2012;16:284–7.
Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics 2008;9:559.
Botía JA, Vandrovcova J, Forabosco P, et al. An additional k-means clustering step improves the biological features of WGCNA gene co-expression networks. BMC Syst Biol 2017;11:47.
Subramanian A, Tamayo P, Mootha VK, et al. Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 2005;102:15545–50.
Liberzon A, Birger C, Thorvaldsdóttir H, et al. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst 2015;1:417–25.
Ge Y, Chen Z, Fu Y, et al. Identification and validation of hub genes of synovial tissue for patients with osteoarthritis and rheumatoid arthritis. Hereditas 2021;158:37.
Newman AM, Steen CB, Liu CL, et al. Determining cell type abundance and expression from bulk tissues with digital cytometry. Nat Biotechnol 2019;37:773–82.
Wilkerson MD, Hayes DN. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinformatics 2010;26:1572–3.
Koppula P, Zhuang L, Gan B. Cystine transporter SLC7A11/xCT in cancer: ferroptosis, nutrient dependency, and cancer therapy. Protein Cell 2021;12:599–620.
Mou Y, Wang J, Wu J, et al. Ferroptosis, a new form of cell death: opportunities and challenges in cancer. J Hematol Oncol 2019;12:34.
Zhang C, Liu X, Jin S, et al. Ferroptosis in cancer therapy: a novel approach to reversing drug resistance. Mol Cancer 2022;21:47.
Rong T, Jia N, Wu B, et al. New Insights into the Regulatory Role of Ferroptosis in Ankylosing Spondylitis via Consensus Clustering of Ferroptosis-Related Genes and Weighted Gene Co-Expression Network Analysis. Genes (Basel) 2022;13.
Venetsanopoulou AI, Anagnostou NE, Tziortzioti Z, et al. "Long-term MRI findings in Ankylosing spondylitis patients treated with TNF inhibitors for a decade". Rheumatol Int 2024.
Xu WD, Yang XY, Li DH, et al. Up-regulation of fatty acid oxidation in the ligament as a contributing factor of ankylosing spondylitis: A comparative proteomic study. J Proteomics 2015;113:57–72.
Cai A, Qi S, Su Z, et al. Quantitative Proteomic Analysis of Peripheral Blood Mononuclear Cells in Ankylosing Spondylitis by iTRAQ. Clin Transl Sci 2015;8:579–83.
Cortes A, Maksymowych WP, Wordsworth BP, et al. Association study of genes related to bone formation and resorption and the extent of radiographic change in ankylosing spondylitis. Ann Rheum Dis 2015;74:1387–93.
Chang X, Cui Y, Zong M, et al. Identification of proteins with increased expression in rheumatoid arthritis synovial tissues. J Rheumatol 2009;36:872–80.
Sun Y, Liu J, Xin L, et al. Xinfeng capsule inhibits inflammation and oxidative stress in rheumatoid arthritis by up-regulating LINC00638 and activating Nrf2/HO-1 pathway. J Ethnopharmacol 2023;301:115839.
Hu Y, Gong C, Li Z, et al. Demethylase ALKBH5 suppresses invasion of gastric cancer via PKMYT1 m6A modification. Mol Cancer 2022;21:34.
Shen C, Sheng Y, Zhu AC, et al. RNA Demethylase ALKBH5 Selectively Promotes Tumorigenesis and Cancer Stem Cell Self-Renewal in Acute Myeloid Leukemia. Cell Stem Cell 2020;27:64–80.e9.
Malacrida A, Rivara M, Di Domizio A, et al. 3D proteome-wide scale screening and activity evaluation of a new ALKBH5 inhibitor in U87 glioblastoma cell line. Bioorg Med Chem 2020;28:115300.
Tang B, Yang Y, Kang M, et al. m(6)A demethylase ALKBH5 inhibits pancreatic cancer tumorigenesis by decreasing WIF-1 RNA methylation and mediating Wnt signaling. Mol Cancer 2020;19:3.
Zhu Z, Qian Q, Zhao X, et al. N(6)-methyladenosine ALKBH5 promotes non-small cell lung cancer progress by regulating TIMP3 stability. Gene 2020;731:144348.
Trewick SC, Henshaw TF, Hausinger RP, et al. Oxidative demethylation by Escherichia coli AlkB directly reverts DNA base damage. Nature 2002;419:174–8.
Zheng G, Dahl JA, Niu Y, et al. ALKBH5 is a mammalian RNA demethylase that impacts RNA metabolism and mouse fertility. Mol Cell 2013;49:18–29.
Wang HF, Kuang MJ, Han SJ, et al. BMP2 Modified by the m(6)A Demethylation Enzyme ALKBH5 in the Ossification of the Ligamentum Flavum Through the AKT Signaling Pathway. Calcif Tissue Int 2020;106:486–93.
Li Z, Wang P, Li J, et al. The N(6)-methyladenosine demethylase ALKBH5 negatively regulates the osteogenic differentiation of mesenchymal stem cells through PRMT6. Cell Death Dis 2021;12:578.
Bowness P, Ridley A, Shaw J, et al. Th17 cells expressing KIR3DL2 + and responsive to HLA-B27 homodimers are increased in ankylosing spondylitis. J Immunol 2011;186:2672–80.
Kucuksezer UC, Aktas Cetin E, Esen F, et al. The Role of Natural Killer Cells in Autoimmune Diseases. Front Immunol 2021;12:622306.
Rak M, Rustin P. Supernumerary subunits NDUFA3, NDUFA5 and NDUFA12 are required for the formation of the extramembrane arm of human mitochondrial complex I. FEBS Lett 2014;588:1832–8.
Torraco A, Nasca A, Verrigni D, et al. Novel NDUFA12 variants are associated with isolated complex I defect and variable clinical manifestation. Hum Mutat 2021;42:699–710.
Ren S, Cao W, Ma J, et al. Correlation evaluation between cancer microenvironment related genes and prognosis based on intelligent medical internet of things. Front Genet 2023;14:1132242.
Ye G, Xie Z, Zeng H, et al. Oxidative stress-mediated mitochondrial dysfunction facilitates mesenchymal stem cell senescence in ankylosing spondylitis. Cell Death Dis 2020;11:775.
Pease CT, Fennell M, Brewerton DA. Polymorphonuclear leucocyte motility in men with ankylosing spondylitis. Ann Rheum Dis 1989;48:35–41.

No competing interests reported.

Download PDF

Version 1

posted

You are reading this latest preprint version

Examining the Involvement of Ferroptosis-Related Genes in Ankylosing Spondylitis and the Infiltration of Immune Cells

Status:

Version 1

Abstract

Figures

Highlights

Introduction

Materials and methods

Data acquisition and pre-processing

Random Forest Model Screening for Characterized Genes

Construction of predictive model

Characteristic gene enrichment analysis

Single sample gene enrichment analysis

Weighted gene Co-expression network analysis

Construction of the protein-protein interaction (PPI) network

GSEA enrichment analysis

Key genes and immune infiltration analysis

Identification of characteristic immune cell subtypes and hub gene validation

Results

General flow of experimental design

Analysis of different expression characteristics of RNA modification factors in AS and control samples

Construction of AS prediction model

Weighted gene Co-expression network and module enrichment analyses

Screening key genes and single sample gene enrichment analysis

Correlation analysis of immune cell infiltration and key genes

Ankylosing spondylitis immuno-molecular typing

Discussion

Conclusion

Declarations

References

Additional Declarations

Status:

Version 1