Identification of Ferroptosis‐Related Hub Genes and Immune Infiltration Landscape in Chronic Kidney Disease via Bioinformatics and Experimental Verification
Yong Luo, Zhong‐ying Huang, Ji‐fang Yang
TL;DR
This study identifies key genes linked to ferroptosis in chronic kidney disease and explores immune cell changes, offering potential new targets for diagnosis and treatment.
Contribution
The study introduces a novel combination of bioinformatics and experimental validation to identify ferroptosis-related hub genes in CKD.
Findings
Seven hub genes (NNMT, GDF15, etc.) were identified and validated as upregulated in CKD.
Hub genes are linked to immune dysfunction and disease progression through specific biological pathways.
In vitro experiments confirmed increased expression of NFE2L2, NNMT, and GDF15 in CKD-related cell models.
Abstract
Chronic kidney disease (CKD) is a serious global health problem with increasing incidence. Ferroptosis plays a crucial role in kidney diseases, but limited studies have elucidated the mechanism and role of ferroptosis in CKD. CKD data sets and ferroptosis‐related genes were acquired from the Gene Expression Omnibus (GEO) database and FerrDB V2. By integrating bioinformatics including weighted gene coexpression network analysis (WGCNA), enrichment analyses, protein‐protein interaction (PPI) network and GeneMANIA analysis, ferroptosis‐related hub genes were identified in CKD. Validation of hub genes was conducted using an external data set, and diagnostic potential capability was evaluated through receiver operating curve (ROC) analysis. Subsequently, the relationship between hub genes and clinical traits was performed using Nephroseq v5 database. Gene set enrichment analysis (GSEA) of…
Genes, proteins, chemicals, diseases, species, mutations and cell lines named across the full text — each resolved to its canonical identifier and authoritative record.
Click any figure to enlarge with its caption.
Figure 1
Figure 2
Figure 3
Figure 4
Figure 5
Figure 6
Figure 7
Figure 8Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsFerroptosis and cancer prognosis · GDF15 and Related Biomarkers · Nuclear Receptors and Signaling
Introduction
1
Chronic kidney disease (CKD) is a serious global health problem with increasing incidence and all‐cause mortality characterized by oxidative stress, renal immune dysregulation and persistent inflammation [1]. CKD can progress to end‐stage renal disease (ESRD), which requires renal replacement therapy such as dialysis or kidney transplantation. In clinical practice, renal function is usually monitored through creatinine levels and glomerular filtration rate (GFR) [2]. However, the underlying pathological mechanism and precise biomarkers behind CKD advancement remain unclear [3]. Therefore, it is crucial to reveal the underlying mechanisms and investigate the diagnostic biomarkers for early screening and treatment of CKD.
Ferroptosis is a form of regulated cell death characterized by lipid peroxidation and iron accumulation [4]. Ferroptosis has shown diagnostic and therapeutic potential in fields such as cancer, degenerative diseases, and organ ischemia reperfusion injury [5]. Furthermore, studies have shown that ferroptosis plays a crucial role in kidney diseases, including acute and chronic nephritis [6]. However, limited studies have elucidated the mechanism and role of ferroptosis in CKD. In the kidneys of CKD patients, iron deposition was observed due to increased iron absorption or decreased iron discharge, which can generate ROS and trigger Fenton‐mediated oxidative damage [7]. Notably, the iron‐catalyzed Fenton reaction, a core chemical process driving lipid peroxidation in ferroptosis, has been demonstrated to be pathologically significant beyond the kidney, as computational modeling shows it critically exacerbates endothelial damage in conditions like atherosclerosis [8]. Regulation of ferroptosis may become promising therapeutic strategies for CKD by clearing lipid peroxidation products, inhibiting lipid peroxidation, and reducing labile iron. However, it is not entirely clear whether ferroptosis‐related genes can serve as diagnostic and therapeutic biomarkers for CKD. Moreover, immune dysregulation and CKD have a strong connection [9]. On one hand, CKD patients are more susceptible to infection due to impaired immune system [10]. On the other hand, immune dysregulation can influence systemic inflammation, thus accelerating the progression of CKD [10, 11]. CKD progression with immune dysregulation can lead to sustained recruitment of immune cells [12]. Dendritic cells (DC), T lymphocytes, B lymphocytes, poly‐morphonuclear leukocytes, monocytes, macrophages, and natural killer (NK) cells are key immune cells involved in the progression of CKD [13]. Furthermore, studies have identified immune‐related genes in CKD [14]. Emerging evidence underscores the crucial role of ferroptosis across kidney diseases, as exemplified by its well‐documented implications in diabetic nephropathy [15]; however, a systematic identification of ferroptosis‐related diagnostic biomarkers for CKD at large is still needed.
To bridge this gap, this study was designed with the following specific objectives: (1) to identify and validate ferroptosis‐related hub genes in CKD through integrated bioinformatics analysis; (2) to evaluate the diagnostic potential of these hub genes (including NNMT, GDF15, ACSL1, DLD, NFE2L2, PARP1, and NR4A1) and their correlation with clinical characteristics; (3) to characterize the landscape of immune cell infiltration in CKD and explore the pro‐fibrotic mechanisms of key immune subsets, such as CD8⁺ T cells and macrophages; and (4) to discuss the potential crosstalk between ferroptosis and the dysregulated immune response, thereby providing new insights for the diagnosis and immunomodulatory therapy of CKD.
Methods
2
Data Source
2.1
GSE66494 and GSE104948 data sets were downloaded from the GEO database (https://www.ncbi.nlm.nih.gov/geo/) and the data sets had the annotation platform of GPL6480 and GPL22945. The GSE66494 samples were derived from renal biopsy specimens, consisting of 8 normal samples and 53 CKD samples and the GSE104948 contained 50 CKD and 18 control human kidney samples. Ferroptosis‐related genes (FRGs) were obtained from the FerrDB V2 (http://www.zhounan.org/ferrdb/current/).
Differentially Expressed Gene Analysis
2.2
Differentially expressed genes (DEGs) were identified with the p‐value < 0.05 and |Log2fold change | > 1 through the “limma” package. Volcano plot and heatmap were generated using the Sangerbox online platform (http://sangerbox.com/home.html#).
Construction of Weighted Gene Co‐Expression Network Analysis (WGCNA)
2.3
WGCNA was used to identify CKD‐related modules and genes. The soft threshold was calculated and then the adjacency matrix was transformed into a topological overlap matrix (TOM). Hierarchical clustering analysis was performed to create the cluster tree and genes were divided into modules. The correlation coefficients and p‐values between modules and clinical features were performed. The most relevant modules were identified based on Module Membership (MM) > 0.8 and Gene Significance (GS) > 0.5.
Construction of Protein–Protein Interaction Network and Identification of Hub Genes
2.4
Hub genes were generated from DEGs, ferroptosis‐related genes, and WGCNA‐related genes. The candidate genes were input into the STRING database (https://cn.string-db.org/) and GeneMANIA analysis to construct the PPI network.
Hub Genes Validation and Correlation Analysis With Renal Function
2.5
The ROC curve of the hub genes was performed and the area under the curve (AUC) was calculated based on the GSE104948 to evaluate the diagnostic performance of the hub genes. The relationship between hub genes and clinical traits was performed using the Nephroseq v5 online database (http://v5.nephroseq.org).
Functional Enrichment Analysis and Gene Set Enrichment Analysis
2.6
Kyoto Encyclopedia of Genes and Genomes (KEGG) and Gene ontology (GO) enrichment analyses of DEGs were conducted using the “clusterProfiler” package of R. Gene set enrichment analysis (GSEA) was performed using the “c2.cp.reactome.v7.4.symbols.gmt” gene set from Molecular Signatures Database (http://www.gsea-msigdb.org/gsea/downloads.jsp). The sample was divided into a high‐expression group and a low‐expression group based on the expression level of hub genes and the signaling pathway was investigated.
Immune Infiltration Analysis
2.7
The CIBERSORT algorithm was employed to examine the infiltration of 22 distinct immune cell types in the samples from GSE66494.
Cell Culture and TGF‐β1 Induced In Vitro Model of CKD
2.8
Human tubular HK‐2 cells were cultured in DMEM‐F12 supplemented with 10% fetal bovine serum (FBS; Gibco, USA) and 1% penicillin‐streptomycin (NCM Biotech, China) at 37°C in 5% CO2. When the HK‐2 cells reached 70% confluence, 48 h of treatment with recombinant TGF‐β1 (10 ng/mL; Peprotech, USA) in the six‐well plates was conducted to establish in vitro model of CKD.
Western Blotting
2.9
HK‐2 cells were lysed in RIPA lysis buffer supplemented with 1% protease inhibitor and 1% PMSF, and then the protein concentration was tested by a BCA kit. Afterwards, the protein (30 μg) was separated and electrically transferred to the PVDF membrane. The membranes were blocked with 5% skim milk for 1 h at room temperature, followed by incubating overnight with antibodies against NFE2L2/NRF2 (T55136, 1:1000) (Abmart, Shanghai, China) and GAPDH (P30008, 1:1000) (Abmart, Shanghai, China). After washing, anti‐Rabbit IgG‐HRP (M21002, 1:5000) (Abmart, Shanghai, China) was incubated, and immunoreactions were visualized by the ECL kit. The protein levels were analyzed by ImageJ software.
NFE2L2 Expression Analysis in Renal Compartments
2.10
The expression distribution of NFE2L2 in renal compartments was observed using data from the Human Protein Atlas (HPA).
RT‐qPCR
2.11
According to the manufacturer's instructions, the total RNA was extracted from cells using the Trizol reagent and the cDNA was obtained using PrimeScript RT reagent kit. The qRT‐PCR experiment was performed using TB Green Premix Ex TaqTM in a StepOnePlusTM RT PCR System (Applied Biosystems; Thermo Fisher Scientific, USA). The qPCR was conducted with the condition of denaturation at 95°C for 30 s, followed by 40 cycles of 95°C for 5 s and 60°C for 30 s. β‐actin served as an internal control. The relative quantification values for NNMT and GDF15 compared to the control were calculated by the 2‐∆∆Ct method.
Statistical Analysis
2.12
The statistical analysis was performed using GraphPad Prism 8 and R software. Two groups were compared using a two‐tailed unpaired t‐test.
Results
3
Identification and Enrichment Analysis of DEGs
3.1
The workflow is shown in Figure 1. A total of 791 DEGs were identified, including 464 up‐regulated and 327 down‐regulated DEGs (Figure 2A). The heatmap showed the top 30 DEGs (Figure 2B). The 791 DEGs were enriched in KEGG pathways, including the cAMP signaling pathway, protein export, fluid shear stress and atherosclerosis, tryptophan metabolism, protein digestion and absorption, and pancreatic secretion (Figure 2C). Additionally, the DEGs were enriched in GO terms such as proteolysis, response to oxygen‐containing compound, secretion, response to cytokine, and response to wounding (Figure 2D).
Flowchart for comprehensive bioinformatics analysis of CKD.
Detection of DEGs and functional enrichment analysis. (A) Volcano plot. (B) Heatmap of the top 30 DEGs in GSE66494. (C) Bar plots of the KEGG enrichment pathways. (D) Bubble plots of the GO enrichment terms.
Weighted Gene Co‐Expression Networks Analysis
3.2
The optimal soft‐threshold was identified as six through the construction of the scale‐free network and the mean connectivity (Figure 3A,B). The dynamic tree cut algorithm successfully delineated 22 distinct gene modules (Figure 3C). Of particular interest, the floralwhite, coral1, mediumpurple3, and ivory modules exhibited a pronounced correlation with CKD, demonstrating a strong correlation coefficient (|R | > 0.5) and significance level (p < 0.01) (Figure 3D). These mod‐ules collectively encompassed 3176 genes for further investigation.
Weighted gene co‐expression networks analysis. (A) Evaluation of fitting indices for scale‐free networks across various soft‐thresholding powers (β). (B) Examination of the mean connectivity across different soft‐thresholding powers. (C) Dendrogram based on a dissimilarity metric for DEGs. (D) Module‐trait relationships between CKD and normal control group.
Construction of PPI Network and Identification of Hub Genes
3.3
By taking the intersection of the gene sets obtained from the DEGs, ferroptosis‐related genes, and WGCNA‐related genes, 11 hub genes were obtained (Nicotinamide N‐methyltransferase (NNMT), Growth differentiation factor‐15 (GDF15), Lysine Demethylase 3B (KDM3B), Long‐chain acyl‐coenzyme A synthases 1 (ACSL1), Selenoprotein S (SELENOS), Dihydrolipoamide Dehydrogenase (DLD), Nuclear factor erythroid 2‐related factor 2 (NFE2L2), Poly (ADP ribose) polymerase 1 (PARP1), Hemoglobin Subunit Alpha 1 (HBA1), TSC22 domain family member 3 (TSC22D3), Nuclear receptor subfamily 4 group A member 1 (NR4A1)) (Figure 4A). Then, the PPI network of 11 hub genes was conducted and 7 hub genes were identified, consisting of exceeding 2 edges (Figure 4B). Furthermore, the GeneMANIA analysis of 7 hub genes was conducted (Figure 4C). Additionally, ROC curves for the 7 hub genes were plotted using the GSE104948 data set. The results demonstrated that the ROC values of NNMT, GDF15, ACSL1, DLD, NFE2L2, PARP1, and NR4A1 were all greater than 0.75 in the validation set of GSE104948. This finding suggested that the 7 hub genes had good diagnostic performance.
Construction of protein‐protein interaction (PPI) networks and identification of hub genes. (A) Venn diagram of DEGs, ferroptosis‐related genes, and WGCNA‐related genes. (B) PPI network of hub genes. (C) The Gene MANIA analysis of hub genes. (D–J) ROC curves of the 7 hub genes in the GSE104948 data set.
Hub Genes Validation and Correlation Analysis With Renal Function
3.4
The expression levels of these hub genes were performed in CKD and normal groups (Figure 5A). NNMT, GDF15, KDM3B, ACSL1, SELENOS, DLD, NFE2L2, PARP1, and TSC22D3 were up‐regulated in CKD group, while HBA1 and NR4A1 were down‐regulated in CKD group. The correlation between the expression of hub genes and clinical data was conducted based on the Nephroseq database. As a result, there was a positive correlation between ACSL1 expression and glomerular filtration rate (GFR) (r ^2^ = 0.25, p < 0.0001) (Figure 5B). There was a negative correlation between DLD expression and serum creatinine levels (r ^2^ = 0.46, p = 0.0026) (Figure 5C). There was a positive correlation between DLD expression and GFR (r ^2^ = 0.57, p = 0.0054) (Figure 5D). Furthermore, there was a positive correlation between these hub genes and serum creatinine levels, while there was a negative correlation between these hub genes and GFR for GDF15 (r ^2^ = 0.45, p = 0.0002; r ^2^ = 0.38, p < 0.0001), PARP1 (r ^2^ = 0.25, p < 0.0001; r ^2^ = 0.30, p < 0.0001), NR4A1 (r ^2^ = 0.57, p = 0.0007; r ^2^ = 0.84, p = 0.0012), NNMT (r ^2^ = 0.32, p < 0.0001; r ^2^ = 0.51, p < 0.0001), and NFE2L2 (r ^2^ = 0.32, p = 0.0009; r ^2^ = 0.32, p = 0.0012) (Figure 5E–N).
The expression of hub genes in GSE66494 (A) and (B–N) correlations between the expression of hub genes and key renal function parameters based on the Nephroseq database.
GSEA for Hub Genes
3.5
GSEA between samples with low and high hub genes expression were conducted. It revealed that “Retrograde transport at the trans‐Golgi network” was enriched gene set in the high ACSL1 expression group (Figure 6A). “Neddylation” and “Peroxisomal lipid metabolism” were enriched gene sets in the high DLD ex‐pression group (Figure 6B). “G1/S‐specific transcription”, “G2/M DNA replication checkpoint”, and “Cell cycle” were enriched gene set in the high GDF15 ex‐pression group (Figure 6C). “Cytokine signaling in immune system” was enriched gene set in the high NNMT expression group (Figure 6D). “tRNA processing in the mitochondrion” was enriched gene set in the low NR4A1 expression group and “Bicarbonate transporters” was enriched gene set in the high NR4A1 expression group (Figure 6E). “Dscam interactions” was enriched gene set in the low NFE2L2 expression group and “TRAIL signaling” was enriched gene set in the high NFE2L2 expression group (Figure 6F). “MeCP2 regulates transcription factors” was enriched gene set in the high PARP1 expression group and “Dscam interactions” was enriched gene set in the low PARP1 expression group (Figure 6G). Taken together, these results indicated that these hub genes affect CKD progress or prognosis through the transport process of inorganic salts, posttranscriptional modifications and transport of pathogenic factors, immune dysfunction, and cell cycle dysregulation.
(A–G) GSEA of hub genes in GSE66494.
Immune Infiltration Analysis
3.6
The CIBERSORT algorithm was employed to assess the infiltration of 22 distinct immune cell types in CKD and normal samples and the proportions of immunocytes were shown in Figure 7A. Comparing with the normal samples, the CKD samples exhibited elevated levels of CD8+ T cells and M0 macrophages, while memory B cells, resting memory CD4+ T cells, Tregs, and resting mast cells showed decreased levels (Figure 7B).
(A) The infiltration of 22 distinct immune cell types in the samples from GSE66494. (B) The landscape of the immune cell infiltration levels in CKD and control groups.
Expression of Hub Genes in Tubular Renal Compartments
3.7
In the in vitro study, the HK‐2 cells were as a tool. The experiments revealed higher expression of NFE2L2 in the TGF‐β1‐treated HK‐2 cells than in the control cells (Figure 8A). We analyzed the distribution and role of NFE2L2 in the kidney based on HPA database, NFE2L2 was predominantly expressed in tubular compartments rather than glomerular compartments (Figure 8B). Subsequently, we conducted RT‐qPCR experiments on the two hub genes, NNMT and GDF15, to further confirm their roles. The results revealed that NNMT and GDF15 expression was significantly increased in the TGF‐β1‐treated HK‐2 cells than in the control cells (Figure 8C,D).
*(A) The representative West blotting images of NFE2L2 protein levels in control and TGF‐β1‐treated HK‐2 cells. (B) Immunohistochemical map of NFE2L2 expression in normal kidney. (C and D) RT‐qPCR experiment results of two hub genes (NNMT and GDF15) in control and TGF‐β1‐treated HK‐2 cells. ***p < 0.001, ***p < 0.0001.
Discussion
4
CKD has become a significant public health issue with increasing incidence and all‐cause mortality [1]. Due to the fact that most CKD patients do not undergo renal biopsy, the diagnosis of CKD mainly relies on clinical features and indicators. However, the number of relevant biomarkers used for clinical diagnosis of CKD is limited. Moreover, CKD has a complex and unclear pathogenesis as a heterogeneous disease. Studies have shown that ferroptosis plays a crucial role in kidney diseases. Consequently, exploring the role of ferroptosis‐related genes in the diagnosis and treatment of CKD is a highly promising research field. In this study, our aim is to identify ferroptosis‐related hub genes in CKD and explore their potential mechanisms of impact on CKD.
In our study, seven hub genes were identified, including NNMT, GDF15, ACSL1, DLD, NFE2L2, PARP1, and NR4A1. NNMT, a nicotinamide (NAM) metabolizing enzyme, regulates both NAD+ and methionine metabolism [16]. A study showed that NNMT deficiency ameliorated renal fibrosis by increasing DNA methylation of connective tissue growth factor and improving renal inflammation [16]. However, another study suggests that NNMT and its metabolite MNAM may be of great therapeutic potential for CKD [17]. According to our study, NNMT was up‐regulated in CKD. There was a positive correlation between NNMT and serum creatinine levels, while there was a negative correlation between NNMT and GFR. “Cytokine signaling in immune system” was enriched gene set in the high NNMT expression. GDF15 is a stress‐responsive cytokine with widely documented protective functions in kidney disease, including anti‐inflammatory and anti‐apoptotic effects, as well as the preservation of renal protective factors like Klotho [18, 19]. Consistent with its role as a biomarker, elevated urinary GDF15 levels predict renal function decline, as seen in diabetic kidney disease [20]. In our study, the upregulation of GDF15 in CKD and its positive correlation with renal impairment parameters align with its recognized status as a sensitive indicator of tissue stress and injury severity. However, emerging perspectives suggest its role may be context‐dependent and dualistic. While acute induction is cytoprotective, chronic, sustained high‐level expression of GDF15 has been implicated in maladaptive processes such as fibrosis and cachexia, potentially contributing to disease progression. Thus, GDF15 may function as a “double‐edged sword”: its elevation primarily signals renal stress and damage rather than directly acting as a causative “risk factor.” Our findings reinforce its value as a robust biomarker and highlight the need for further research to delineate the conditions under which its biological activity shifts from protective to potentially detrimental in the progression of CKD. ACSL1 is a member of ACSL family and one of the hub genes involved in the classic mechanism of promoting ferroptosis [21]. However, there is few studies on the role of ACSL1 in CKD. A study suggests that inhibition of Nrf2 can accelerate renal lipid deposition in obesity associated kidney disease by inhibiting the expression of ACSL1 [22]. In another study, fatty acids and inflammatory stimuli induce expression of ACSL1 to promote lipid remodeling in diabetic kidney disease [23]. In our study, ACSL1 was up‐regulated in CKD and positively correlated with clinical features of renal impairment. Dihydrolipoamide dehydrogenase (DLD or DLDH) is a flavin‐dependent disulfide oxidoreductase and its deficiencies have been linked to numerous metabolic disorders, which may reveal druggable targets for disease interventions or prevention due to its redox and non‐redox features [24]. Multiple studies have shown that DLD can serve as a diagnostic and therapeutic biomarker in various diseases through cuproptosis [25, 26]. However, the role of DLD in CKD remains unclear. In our study, DLD was upregulated in CKD, but there was a negative correlation between DLD expression and serum creatinine levels and there was a positive correlation between DLD expression and GFR. Based on the comprehensive results of GSEA, DLD may act as a CKD protective factor through the “peroxisomal lipid metabolism” pathway. Numerous studies have also shown that peroxisomal fatty acid oxidation plays an important regulatory role in the progression of kidney injury [27, 28, 29]. We look forward to further exploration to confirm the potential of DLD as a protective factor and therapeutic target for CKD. NFE2L2, also known as NRF2, is a key regulator of ferroptosis [30]. Studies have shown that NFE2L2 can serve as a therapeutic target in renal fibrosis and kidney injury and diseases [31, 32]. It can exert its antioxidative ability in inflammatory processes to protect tissues and organs from further damage [32]. In our study, NFE2L2 was up‐regulated in CKD and positively correlated with clinical features of renal impairment. PARP1 is widely involved in the pathological processes of cancer and thrombosis [33, 34]. Moreover, CKD led to enhanced PARP‐1 cleavage associated with cellular apoptosis in a rat model for CKD [35]. NR4A1 belongs to the nuclear receptor superfamily and plays an important role in energy metabolism by regulating cell proliferation, apoptosis, differentiation, development, and immunity [36]. It is reported that loss of NR4A1 enhances macrophage‐mediated renal injury in CKD [37]. Interestingly, another recent study also showed that NR4A1 is a diagnostic biomarker for CKD through a combined bioinformatics analysis of CKD and nonalcoholic fatty liver disease [38]. Overall, the identified core genes associated with ferroptosis have great potential as diagnostic and therapeutic targets for CKD. Furthermore, the renal immune infiltration microenvironment in CKD is also worth paying attention to and exploring. In our study, there is immune heterogeneity between CKD and the normal group. Comparing with the normal samples, the CKD samples exhibited elevated levels of CD8+ T cells and M0 macrophages, while memory B cells, resting memory CD4+ T cells, Tregs, and resting mast cells showed decreased levels. The significant infiltration of CD8⁺ T cells and M0 macrophages suggests their active roles in driving renal injury towards fibrosis. Our bioinformatics analysis revealed a significant increase in CD8⁺ T cell infiltration in CKD, which aligns with emerging evidence establishing a causal and mechanistic link between this immune subset and disease progression. Notably, a recent Mendelian randomization study has provided genetic evidence that an elevated percentage of naïve CD8⁺ T cells (CD28+ CD45RA+) in peripheral blood exerts a causal effect on CKD onset, underscoring its potential role not merely as a biomarker but as a contributing factor in disease etiology [39]. Mechanistically, the pathogenic role of CD8⁺ T cells extends beyond their classical cytotoxic functions. A pivotal study using a mouse model of AKI‐to‐CKD transition elucidated a detailed pathway: CD8⁺ T cells are recruited to the injured kidney interstitium via the CXCL16‐CXCR6 chemokine axis, a process mediated by macrophages. Once localized, these activated CD8⁺ T cells directly drive progressive renal damage by inducing apoptosis of peritubular capillary endothelial cells through Fas ligand (FasL)‐Fas signaling. This results in peritubular capillary (PTC) rarefaction, creating a state of chronic hypoxia that is a potent and fundamental driver of tubular atrophy and interstitial fibrosis. Crucially, the depletion of CD8⁺ T cells was shown to ameliorate both PTC loss and renal fibrosis, confirming their active role in disease pathogenesis [40]. Therefore, the elevated CD8⁺ T cell levels observed in our study likely contribute to CKD progression by instigating microvascular injury and loss, thereby perpetuating a vicious cycle of ischemia, inflammation, and fibrotic scarring. The observed increase in M0 macrophages in our study signifies an expanded reservoir of undifferentiated cells poised for activation within the unique CKD microenvironment. These M0 macrophages are recruited to the injured renal interstitium through chemokine gradients, such as MCP‐1, CCL2, and CCL3, released by activated tubular epithelial cells and other resident cells [41, 42, 43]. Upon arrival, they undergo phenotypic polarization dictated by local signals. In the early and inflammatory phases of CKD, damage‐associated molecules (e.g., KIM‐1 expressed on injured tubules) and pattern recognition receptor (PRR) activation drive M0 macrophages towards a pro‐inflammatory M1 phenotype [41, 42, 43]. These M1 macrophages secrete potent inflammatory cytokines like IL‐1β and TNF‐α, as well as reactive oxygen species (ROS) and matrix metalloproteinases (MMPs) [41, 42, 43]. This sustained inflammatory milieu not only directly damages renal parenchyma but also creates a feedback loop that perpetuates immune cell recruitment and activation, setting the stage for fibrogenesis [41, 42, 43]. As the disease progresses, the cytokine milieu shifts, promoting a transition towards a pro‐fibrotic M2 phenotype. Although possessing anti‐inflammatory functions via IL‐10, M2 macrophages are pivotal drivers of fibrosis through the robust production of transforming growth factor‐beta (TGF‐β). TGF‐β is the master regulator of fibrosis, directly activating resident fibroblasts and inducing their differentiation into α‐smooth muscle actin (α‐SMA)‐positive myofibroblasts [41, 42, 43]. These myofibroblasts are the primary effector cells responsible for the excessive synthesis and deposition of extracellular matrix (ECM) proteins, such as collagens, leading to glomerulosclerosis, tubular atrophy, and interstitial fibrosis [41, 42, 43]. Therefore, the accumulation of M0 macrophages represents a critical upstream event in a dynamic pathological cascade. Their context‐dependent polarization into M1 and subsequently M2 subsets sequentially fuels chronic inflammation and then directly executes the fibrotic program, ultimately culminating in the irreversible destruction of normal renal architecture and the decline of kidney function. The above findings suggest that CKD patients may see more possibilities and dawn from immunomodulatory treatment strategies.
Our study provides a new perspective for studying the role of ferroptosis in CKD and identifies ferroptosis‐related diagnostic biomarkers. However, our research also has certain limitations, including a limited number of available CKD samples and lack of cellular animal experiments. Our analysis of immune cell infiltration was performed exclusively using the CIBERSORT computational algorithm on publicly available transcriptomic data. While CIBERSORT is a widely adopted and validated tool for deconvoluting immune cell fractions from bulk RNA‐seq data, this approach remains an inference and lacks direct experimental validation, such as flow cytometry or immunofluorescence staining of actual kidney tissue samples from CKD patients. Therefore, the specific alterations in immune cell subsets reported here require further confirmation through histopathological or cytometric techniques in future studies. In order to explore our findings more comprehensively, we will expand our data set in the future and confirm our research through cell and animal experiments. In conclusion, recent advances underscore ferroptosis as a pivotal and druggable mechanism in renal pathology. Its role extends beyond direct tubular injury to involve complex interactions, such as endothelial damage triggered by hemolytic microparticles via the miR‐130a/ACSL4 axis, highlighting context‐specific triggers in kidney disease [44]. Critically, mitochondrial dysfunction and oxidative stress are established as key upstream events driving ferroptotic cell death, offering precise intervention points [45]. The consolidation of these pathways reaffirms ferroptosis as a promising therapeutic target for halting the progression of various kidney diseases, including CKD [46]. Our identification of ferroptosis‐related hub genes aligns with this evolving framework, suggesting their potential involvement in these core regulatory networks that connect iron metabolism, redox homeostasis, and inflammatory‐fibrotic responses in CKD.
Conclusions
5
Our comprehensive bioinformatics analysis identified seven ferroptosis‐related hub genes—NNMT, GDF15, ACSL1, DLD, NFE2L2, PARP1, and NR4A1—as potential diagnostic biomarkers for CKD. Each gene presents a unique mechanistic link to disease progression: NNMT upregulation correlates with renal function indices, suggesting its role in metabolic reprogramming and fibrosis in CKD. GDF15, a sensitive stress‐response marker, may reflect and participate in a vicious cycle of renal injury and fibrosis upon chronic elevation. ACSL1, a canonical ferroptosis promoter, likely drives lipid peroxidation damage in tubular cells. The expression pattern of DLD and its association with the “peroxisomal lipid metabolism” pathway imply a potential protective role via modulating cellular redox homeostasis. The compensatory upregulation of NFE2L2, a master antioxidant transcription factor, may represent a key response to intrarenal oxidative stress. The aberrant expressions of PARP1 and NR4A1 point to dysregulated DNA damage response and immunometabolic regulation in CKD, respectively. Furthermore, we revealed a distinct immune landscape in CKD, characterized by increased infiltration of pro‐fibrotic CD8⁺ T cells and M0 macrophages. The interplay between these hub genes and the dysregulated immune microenvironment likely forms a vicious cycle, perpetuating ferroptosis, inflammation, and ultimately renal fibrosis. Collectively, these findings not only advance our understanding of ferroptosis in CKD pathogenesis but also pinpoint specific molecular targets (e.g., ACSL1, NFE2L2) and immune pathways for developing novel diagnostic strategies and therapeutic interventions. However, our research also has certain limitations, including a limited number of available CKD samples and lack of cellular animal experiments.
Author Contributions
Conceptualization: Ji‐Fang Yang. Methodology: Yong Luo and Zhong‐Ying Huang. Software: Zhong‐Ying Huang. Validation: Yong Luo. Formal analysis: Yong Luo. Investigation: Yong Luo. Resources: Yong Luo and Ji‐Fang Yang. Data curation: Yong Luo. Writing – original draft preparation: Yong Luo. Writing – review and editing: Ji‐Fang Yang. Visualisation: Zhong‐Ying Huang. Supervision: Ji‐Fang Yang. Project administration: Ji‐Fang Yang. All authors have read and agreed to the published version of the manuscript.
Funding
The authors received no specific funding for this work.
Ethics Statement
The authors have nothing to report.
Consent
The authors have nothing to report.
Conflicts of Interest
The authors declare no conflicts of interest.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1A. C. Webster , E. V. Nagler , R. L. Morton , and P. Masson , “Chronic Kidney Disease,” Lancet 389 (2017): 1238–1252.27887750 10.1016/S 0140-6736(16)32064-5 · doi ↗ · pubmed ↗
- 2J. S. Lees , C. E. Welsh , C. A. Celis‐Morales , et al., “Glomerular Filtration Rate by Differing Measures, Albuminuria and Prediction of Cardiovascular Disease, Mortality and End‐Stage Kidney Disease,” Nature Medicine 25 (2019): 1753–1760.10.1038/s 41591-019-0627-8PMC 685887631700174 · doi ↗ · pubmed ↗
- 3R. Huang , P. Fu , and L. Ma , “Kidney Fibrosis: From Mechanisms to Therapeutic Medicines,” Signal Transduction and Targeted Therapy 8 (2023): 129.36932062 10.1038/s 41392-023-01379-7PMC 10023808 · doi ↗ · pubmed ↗
- 4B. R. Stockwell , “Ferroptosis Turns 10: Emerging Mechanisms, Physiological Functions, and Therapeutic Applications,” Cell 185 (2022): 2401–2421.35803244 10.1016/j.cell.2022.06.003PMC 9273022 · doi ↗ · pubmed ↗
- 5F. Chen , R. Kang , D. Tang , and J. Liu , “Ferroptosis: Principles and Significance in Health and Disease,” Journal of Hematology & Oncology 17 (2024): 41.38844964 10.1186/s 13045-024-01564-3PMC 11157757 · doi ↗ · pubmed ↗
- 6Z. Han , Y. Luo , H. Chen , et al., “A Deep Insight Into Ferroptosis in Renal Disease: Facts and Perspectives,” Kidney Diseases 10 (2024): 224–236.38835406 10.1159/000538106 PMC 11149998 · doi ↗ · pubmed ↗
- 7S. Li , Q. Han , C. Liu , et al., “Role of Ferroptosis in Chronic Kidney Disease,” Cell Communication and Signaling 22 (2024): 113.38347570 10.1186/s 12964-023-01422-8PMC 10860320 · doi ↗ · pubmed ↗
- 8D. Formanowicz , M. Radom , A. Rybarczyk , and P. Formanowicz , “The Role of Fenton Reaction in ROS‐Induced Toxicity Underlying Atherosclerosis‐Modeled and Analyzed Using a Petri Net‐Based Approach,” Biosystems 165 (2018): 71–87.29317313 10.1016/j.biosystems.2018.01.002 · doi ↗ · pubmed ↗