Abstract Background Idiopathic pulmonary fibrosis (IPF) is a chronic, irreversible, and fatal disease characterized by progressive interstitial lung fibrosis. Given its insidious onset and poor outcome, there is an urgent need to elucidate the molecular mechanisms underlying IPF and identify effective therapeutic targets and diagnosis and prognosis biomarkers. Ferroptosis is an iron-dependent form of programmed cell death that occurs as lipid peroxides accumulate. Growing evidence suggests that ferroptosis is important in IPF. Methods Human ferroptosis PCR array was performed on IPF and control lung tissue. The differentially expressed ferroptosis-related genes (DE-FRGs) were identified, underwent functional enrichment analyses, protein–protein interaction network construction, and potential drug target prediction. The DE-FRGs were validated and their value as diagnostic and prognostic blood biomarkers were evaluated using the Gene Expression Omnibus dataset [46]GSE28042. Results The array identified 13 DE-FRGs. Gene Ontology enrichment and Kyoto Encyclopedia of Genes and Genomes pathway analyses revealed that the DE-FRGs were mainly related to iron ion transport, blood microparticles, and oxidoreductase activity, and were involved in porphyrin metabolism, necroptosis, and the p53 signaling pathway in addition to ferroptosis. The 13 DE-FRGs were analyzed using the Drug–Gene Interaction Database to explore novel IPF therapeutic agents, yielding 42 potential drugs. Four DE-FRGs (BBC3, STEAP3, EPRS, SLC39A8) in the peripheral blood of IPF patients from the [47]GSE28042 dataset demonstrated the same expression pattern as that observed in the lung tissue array. The receiver operating characteristic analysis demonstrated that the area under the curve of STEAP3 and EPRS were > 0.75. The survival analysis demonstrated that STEAP3 and EPRS were significantly different between the IPF and control groups. Conclusions The FRG expression profiles in IPF and control lung tissue were characterized. The findings provided valuable ideas to elucidate the role of ferroptosis in IPF and aided the identification of novel IPF therapeutic targets and biomarkers. Supplementary Information The online version contains supplementary material available at 10.1186/s12890-025-03555-7. Keywords: Idiopathic pulmonary fibrosis, Ferroptosis, PCR array, Bioinformatics analysis, Drug targets, Biomarkers Background Idiopathic pulmonary fibrosis (IPF) is the most common form of idiopathic interstitial pneumonia and is a chronic, irreversibly progressive, and life-threatening disease with poor clinical outcomes [[48]1, [49]2]. Epidemiological data demonstrated that the global prevalence of newly diagnosed IPF is 20 cases per 100,000 people per year [[50]3]. IPF is a typical age-related disease that occurs in people aged > 50 years [[51]4], where the estimated survival is 3–5 years if untreated after diagnosis [[52]5]. Nevertheless, the exact cause of IPF remains unknown. Several studies reported the potential risk factors for IPF: genetic variations, lifestyle behaviors, environmental exposures, and viral infections [[53]6]. Although IPF pathogenesis has not been fully elucidated, it is accepted that recurrent injury and aberrant repair of alveolar epithelial cells leads to excessive collagen deposition in the lung, which subsequently develops into pulmonary fibrosis [[54]7]. Given the enigmatic etiology, insidious onset, and dire prognosis of IPF, there is an urgent need for reliable biomarkers enabling early diagnosis and prognosis. Current evidence-based guidelines suggest that the antifibrotic agents pirfenidone and nintedanib are approved for treating IPF. However, in addition to tolerability issues, both drugs exhibit limited efficacy in preventing disease progression and enhancing quality of life [[55]8]. Currently, lung transplantation is the only curative treatment for IPF, which is not an option for most patients due to their age and comorbidities [[56]9]. Given these characteristics, the discovery of effective therapeutic targets for IPF also remains fundamentally important. Ferroptosis is a unique form of programmed cell death (PCD) that was identified and named a decade ago [[57]10, [58]11]. Unlike apoptosis, necroptosis, and other forms of PCD, ferroptosis is characterized by the iron-dependent accumulation of hazardous lipid peroxidation products [[59]12]. Ferroptosis has been linked to a wide range of biological properties, including development, immunity, ageing, and cancer [[60]13]. Recently, there has been increasing evidence that ferroptosis is important in lung diseases such as acute lung injury (ALI), chronic obstructive pulmonary disease (COPD), and pulmonary fibrosis [[61]14–[62]16]. Iron accumulation was detected in both in vitro and in vivo models of pulmonary fibrosis and notably in histological samples from IPF patients, revealing some of the underlying mechanisms [[63]17]. For example, TGF-β stimulation upregulated transferrin receptor protein 1 (TFRC) expression in both human and mouse lung fibroblasts, resulting in iron overload, which accelerated fibroblast activation [[64]18]. Additionally, specifically inhibiting UHRF1 arrested ferroptosis formation and impeded pulmonary fibrosis progression in murine models, suggesting UHRF1 as a potential target for treating IPF [[65]19]. However, most studies relating ferroptosis to IPF used cellular and animal models, which are not wholly representative of the process in humans. Therefore, a systematic understanding of the potential role of ferroptosis in IPF pathogenesis was obtained by collecting lung tissue samples from IPF patients and healthy donors during lung transplantation surgery and conducting PCR arrays of the ferroptosis-related genes (FRGs). Differentially expressed genes (DEGs) specifically associated with ferroptosis were identified and functionally analyzed using bioinformatics. Whether these differentially expressed FRGs (DE-FRGs) can be used as potential diagnostic or prognostic indicators in IPF was investigated by testing the expression of DE-FRGs in peripheral blood mononuclear cells (PBMCs) from IPF patients in the Gene Expression Omnibus (GEO) microarray dataset [66]GSE28042. The genes with similar expression patterns to those detected by our array underwent survival analysis. Overall, the analysis revealed the FRG expression profiles in IPF patients’ lung tissue, providing new insights into the underlying mechanism of IPF. We aimed to identify potential diagnosis and prognosis molecular markers and candidate targets for IPF treatment. Methods Patients and tissue samples Fresh lung tissues were obtained from 16 IPF patients who underwent lung transplantation at the Affiliated Wuxi People’s Hospital of Nanjing Medical University (Jiangsu, China) between May and December 2019. The control group consisted of lung tissues from 12 healthy donors who were matched to the above recipients during the same period after donor lung volume reduction. We excluded patients with evidence of a known cause of pulmonary fibrosis, such as occupational exposures, autoimmune diseases, medications, and infections. For each case, the lung tissue was divided into two parts: one part was rapidly frozen in liquid nitrogen and stored at -80 °C for preservation, while the other was fixed in 4% paraformaldehyde, then embedded into a paraffin block. Five cases each from the IPF and control groups were randomly selected for PCR array analysis. The remaining cases from both groups were used for expression validation. The Affiliated Wuxi People’s Hospital of Nanjing Medical University Institutional Research Ethics Committee approved the human sample collection protocols (2023–126). RNA extraction and complementary DNA (cDNA) synthesis The IPF and control lung tissues were transferred from the cryogenic refrigeration system to a liquid nitrogen pre-cooled mortar and ground into powder at low temperature. Total RNA was extracted using TRIzol (Invitrogen, USA). RNA quality was evaluated via agarose gel electrophoresis. RNA concentration and purity were determined using a NanoDrop 2000 spectrophotometer (Thermo Fisher Scientific, USA). The qualifying RNA samples were reverse-transcribed to cDNA in accordance with the reverse transcription kit instructions (Wcgene Biotech, Shanghai, China). The cDNA and remaining RNA were stored at -80 °C until used. PCR array The potential roles of ferroptosis in IPF pathogenesis and progression were explored using a PCR array analysis targeting the ferroptosis-related genes. Gene expression profiles were analyzed using the human ferroptosis PCR array (Wcgene Biotech) according to the manufacturer’s protocol. A total of 90 FRGs were examined, including key regulators involved in biological processes such as iron ion metabolism, mitochondrial function regulation, GSH homeostasis maintenance, and redox signaling. Additionally, the PCR array plate included four housekeeping genes for the subsequent internal reference gene selection and target gene expression analysis. Principal component analysis (PCA) The array data dimensionality was reduced and the similarities and differences between the control and IPF group samples were evaluated on a global scale using PCA. The analysis was performed in R and visualized using the R ggplot2 package. Analyses of DEGs The data were analyzed using Wcgene Biotech software with a screening threshold of |fold change|≥ 2 and P < 0.05 for identifying DEGs. The DEG heat map was generated using the R ComplexHeatmap package. The volcano plot was produced using GraphPad Prism. Functional enrichment analysis Gene Ontology (GO) annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis of DEGs were conducted using the R clusterProfile package. The GO annotation encompassed biological processes (BP), cellular components (CC), and molecular functions (MF). Reactome enrichment analysis was conducted using the R ReactomePA package. The results were visualized using the R ggplot2 package. Literature search for DE-FRGs The molecular mechanisms involved in the effect of DE-FRGs on the pathogenesis of pulmonary fibrosis and their potential role in clinical translational application were explored by systematically searching for each DE-FRG using the PubMed database with the gene name and “pulmonary fibrosis” as keywords. Correlation heat map The interrelationships among the DE-FRGs were visualized using a correlation heat map. Spearman correlation analysis was performed in R. The outcomes were visualized using the R ggplot2 package. Protein–protein interaction (PPI) network construction The GeneMANIA prediction server ([67]www.genemania.org) integrates biological networks to prioritize genes and gene function prediction [[68]20]. A PPI network of DEGs was constructed using GeneMANIA to evaluate their interactions and functions. Drug–gene interactions The Drug–Gene Interaction Database (DGIdb, [69]www.dgidb.org) provides extensive data on drug–gene interactions and druggable genes from credible sources. In this study, potential drugs against the DEGs were predicted using DGIdb to discover new therapeutic targets. The cut-off criterion was a query score > 1 [[70]21]. Subsequently, the drug–gene interaction network was generated using Cytoscape. Real-time PCR expression validation The PCR array results were confirmed using real-time PCR, and individual gene expression was examined. Six of the 13 DEGs identified from the array screening were randomly selected and validated in the remaining collected samples (control: 7, IPF: 11). The RNA, which had been previously extracted and stored at -80 °C, was thawed and re-evaluated for concentration and purity. Subsequently, quantitative PCR (qPCR) was performed using the SYBR Green One-Step RT-qPCR Kit (Beyotime, Shanghai, China) on a real-time PCR instrument (Bio-Rad, USA). Table [71]1 presents the primer sequences. The expression of each gene was normalized to the housekeeping gene GAPDH. The relative gene expression was calculated using the comparative threshold cycle (2^−ΔΔCt) method. Table 1. Real-time PCR primers Primer Sequence (5′–3′) GAPDH F: TCGGAGTCAACGGATTTGGT R: TTCCCGTTCTCAGCCTTGAC BBC3 F: TGCCTGCCTCACCTTCATC R: CTTCAGCCAAAATCTCCCAC FTMT F: ATGCGTCCTACGTGTACTTGTC R: CCTCCTCGCTGGTTCTGC STEAP3 F: TGCGTCCACTGCTGTCTTC R: ATGTCGTTGAGGCACTTTTGTA CDO1 F: TTTGATACATGCCATGCCTTTG R: GGTGCCCCTTAGTTGTTCTCC FTH1 F: GGCTGAATGCAATGGAGTGT R: TTGATGGCTTTCACCTGCTC SLC39A8 F: GTTCAGGCTTAGAGGCTTATCG R: TGCCATTTGTGAGGGGTG [72]Open in a new tab F Forward, R Reverse Immunohistochemical validation of DE-FRGs Immunohistochemical staining was performed on samples from the validation set (control: 7, IPF: 11) to verify the DEG expression at the protein level. The six genes used for the protein expression verification were the same as those selected for the RNA expression validation described earlier. Wax blocks of lung tissue were cut into 5-μm thick sections, then deparaffinized, rehydrated, and processed for antigen retrieval, blocking, and primary antibody incubation at 4 °C overnight. The primary antibodies used were: rabbit anti-BBC3 (1:1000; Immunoway, China, YM8416), rabbit anti-FTMT (1:2000; Abcam, USA, ab124889), rabbit anti-STEAP3 (1:1000; Proteintech, USA, 28,478–1-AP), rabbit anti-CDO1 (1:200; Proteintech, 12,589–1-AP), rabbit anti-FTH1 (1:500; CST, USA, 4393S), and rabbit anti-SLC39A8 (1:200; Proteintech, 20,459–1-AP). After incubation with the secondary antibody, DAB chromogenic detection, and hematoxylin counterstaining, the sections were observed and photographed under a light microscope (Olympus, Japan). GEO database analysis of DE-FRGs in PBMCs The consistency of DE-FRG expression between lung tissue and blood was investigated by evaluating the DEGs identified in our array by analyzing their expression in PBMCs from IPF patients in a larger population using the GEO [73]GSE28042 dataset. The expression of DE-FRGs identified by our array was assessed using the [74]GSE28042 microarray dataset (platform: [75]GPL6480), which contains the clinical information and gene expression profiles of PBMCs from 19 controls and 75 IPF patients. Receiver operating characteristic (ROC) and survival analysis The diagnostic and prognostic potential of DEGs whose expression patterns in [76]GSE28042 PBMCs were concordant with those observed in the lung tissue array were evaluated using ROC and Kaplan–Meier survival analyses. The ROC curve presents the interplay between sensitivity and specificity and is a valuable tool for evaluating diagnostic tests and assessing their predictive accuracy. ROC plots were generated and the relevant parameters (specificity, sensitivity, precision, the Youden index, cutoff value, and the area under the curve [AUC]) were calculated using the R package pROC. The positive and negative likelihood ratios were calculated as follows: positive likelihood ratio = sensitivity/(1—specificity) and negative likelihood ratio = (1—sensitivity)/specificity [[77]22]. Kaplan–Meier analysis was used to compare survival differences between groups. Survival analysis, including the calculation of the number at risk, was performed using Cox regression with the R survival package. The optimal cut-off of the expression value that best separated the survival outcome into two groups was determined for each gene based on the surv-cutpoint function of the R survminer package [[78]23]. The results were visualized using the R survminer and ggplot2 packages. The analyses evaluating the diagnostic and prognostic potential of the DE-FRGs followed the principles outlined in the STARD 2015 (Standards for Reporting Diagnostic Accuracy Studies) guidelines for reporting diagnostic accuracy [[79]24]. Statistical analysis Statistical analysis was performed using R 4.2.1 and GraphPad Prism 8.0. All data conformed to the assumption of normal distribution and are presented as the mean ± standard deviation. Comparisons between two groups were made using two-tailed t-tests. A difference of P < 0.05 was considered statistically significant. Results Baseline characteristics Figure [80]1 depicts the flowchart of the overall research design. This study included 16 IPF patients (female: 4, male: 12; mean age: 58.94 ± 8.13 years, age range: 37–74 years) and 12 controls (female: 3, male: 9; mean age: 43 ± 10.85 years, age range: 13–54 years). Pulmonary function [predicted forced vital capacity percentage (FVC%), predicted diffusing capacity of the lung for carbon monoxide percentage (DLCO%), 6-min walk test (6MWT)] was assessed in the IPF patients. Table [81]2 presents the participants’ clinical characteristics. Fig. 1. [82]Fig. 1 [83]Open in a new tab Flowchart of the overall study design Table 2. The participants’ clinical characteristics IPF (n = 16) Controls (n = 12) Age (years) 58.94 ± 8.13 43 ± 10.85 Sex (female/male) 4/12 3/9 BMI (kg/m^2) 22.82 ± 4.07 26.22 ± 2.92 Former smoker 6 (37.5%) Family history 2 (12.5%) FVC (%) 41.26 ± 18.02 DLCO (%) 18.92 ± 11.30 6MWT (m) 159.5 ± 89.30 [84]Open in a new tab BMI Body mass index, FVC Forced vital capacity, DLCO Diffusing capacity of the lung for carbon monoxide, 6MWT 6-min walk test Identification of DE-FRGs between controls and patients The specific FRGs that contribute to IPF development were identified by examining the expression patterns of 90 FRGs in the lung tissues of five randomly selected samples each from the control and IPF groups. The PCR array data underwent PCA. The first two principal components (PC) accounted for > 50% of the total data variance, with PC1 explaining 36.4% and PC2 explaining 16.6% of the data variance (Fig. [85]2A). This demonstrated a distinct distribution of samples between the control and IPF groups. Fig. 2. [86]Fig. 2 [87]Open in a new tab Identification of DE-FRGs between controls and IPF patients. A PCA of control and IPF groups. B Heatmap of the FRGs. The green box highlights genes with a significant difference in expression. C Volcano plot of the FRG expression profiles The screening threshold for the array data was set at |fold change|≥ 2 and P < 0.05. The heatmap illustrates the clustering distribution of samples and DEGs. Of the 90 FRGs examined in the array, the 13 genes marked with a green box showed a distinct expression pattern between the control and IPF groups (Fig. [88]2B). The volcano plot revealed a consistent pattern with the heatmap, indicating that the IPF group had 13 DE-FRGs compared to the control group. Among the 13 DE-FRGs, seven genes were upregulated (ALOX15, BBC3, CP, FTMT, NOX3, STEAP3, TF) and six genes were downregulated (CDO1, EPRS, FTH1, LPCAT3, SLC39A8, TXNRD1) (Fig. [89]2C). Table [90]3 presents detailed information regarding the DE-FRGs. Table 3. DE-FRGs between the IPF and control groups Gene symbol Description Fold change P-value ALOX15 Arachidonate 15-lipoxygenase ↑2.400 0.040 BBC3 BCL2-binding component 3 ↑2.349 0.030 CP Ceruloplasmin ↑3.593 0.030 FTMT Ferritin mitochondrial ↑3.995 0.010 NOX3 NADPH oxidase 3 ↑2.658 0.045 STEAP3 STEAP3 metalloreductase ↑2.448 0.007 TF Transferrin ↑2.605 0.044 CDO1 Cysteine dioxygenase type 1 ↓0.349 0.019 EPRS Glutamyl-prolyl-tRNA synthetase 1 ↓0.492 0.011 FTH1 Ferritin heavy chain 1 ↓0.483 0.013 LPCAT3 Lysophosphatidylcholine acyltransferase 3 ↓0.496 0.012 SLC39A8 Solute carrier family 39 member 8 ↓0.146 0.000 TXNRD1 Thioredoxin reductase 1 ↓0.304 0.022 [91]Open in a new tab The ↑ and ↓ icons represent upregulated and downregulated genes, respectively Functional enrichment analyses of DE-FRGs and their potential effect on pulmonary fibrosis The biological functions of the DE-FRGs and associated signaling pathways were elucidated using GO, KEGG, and Reactome enrichment analyses. A total of 187 GO items, seven KEGG pathways, and 203 Reactome categories were identified and sorted according to the P-value. The top five items for each GO category are presented in bubble charts (Fig. [92]3A), while the seven KEGG pathways are presented using a chord diagram (Fig. [93]3B). The GO enrichment analysis determined that the DE-FRGs were mainly involved in BP such as iron ion transport, iron ion homeostasis, and transition metal ion transport. The DE-FRGs were primarily related to the CC of blood microparticles, extrinsic component of plasma membrane, and autolysosome. The DE-FRGs were mainly associated with the MF of iron ion binding, ferrous iron binding, and oxidoreductase activity. The KEGG pathway prediction revealed that in addition to ferroptosis, the DE-FRGs were also involved in porphyrin metabolism, mineral absorption, necroptosis, and the p53 signaling pathway. Furthermore, the Reactome analysis indicated that the DE-FRGs were mostly enriched in iron uptake and transport and transport of small molecules (Fig. [94]3C). Additional file 1: Table S1 details the enriched functional items. Fig. 3. [95]Fig. 3 [96]Open in a new tab Functional enrichment analyses of DE-FRGs. A Bubble charts of GO enrichment analysis. B Chord diagram of KEGG pathway analysis. C Bubble chart of Reactome enrichment analysis. Hsa04216, ferroptosis; hsa00860, porphyrin metabolism; hsa04978, mineral absorption; hsa04115, p53 signaling pathway; hsa04217, necroptosis; hsa00430, taurine and hypotaurine metabolism; hsa00450, selenocompound metabolism The literature search on the association between DE-FRGs and pulmonary fibrosis identified seven genes with potential influence on pulmonary fibrosis. Some of the genes have potential biomarker roles, such as CP and TF, others are potential targets for pulmonary fibrosis treatment in in vitro and in vivo models, such as BBC3 and EPRS, and some are downstream targets of certain profibrotic factors or anti-fibrotic agents, such as NOX3 and FTH1. Table [97]4 details the associations between the DE-FRGs and pulmonary fibrosis. Table 4. The influence of identified DE-FRGs on pulmonary fibrosis and their potential role in clinical translational application Gene Potential role in the clinical translational application of pulmonary fibrosis Major effect on pulmonary fibrosis (clinical/preclinical findings) Ref BBC3 Potential novel therapeutic target for pulmonary fibrosis Knockdown of BBC3 expression in vitro and in vivo attenuated silica-induced lung fibrosis by reducing autophagy occurrence [[98]25] CP Potential as a serum biomarker of lung fibrosis Serum levels of CP were significantly correlated with the exposure levels of respirable fibers in occupational exposure-related pulmonary fibrosis [[99]26] NOX3 Downstream target of the profibrotic factor IGFBP-5 IGFBP-5 stimulated transcriptional expression of NOX3 in human fibroblasts. Selective knockdown of NOX3 reduced ROS production by IGFBP-5 [[100]27] TF Potential biomarker of IPF BAL fluid from IPF patients had elevated TF levels compared to controls [[101]28] EPRS Potential therapeutic target for anti-IPF interventions EPRS regulated the expression of mesenchymal markers and extracellular matrix proteins through TGFβ1–STAT signaling in in vitro and in vivo IPF models [[102]29] FTH1 Downstream target of potential pulmonary fibrosis therapeutic DHQ DHQ suppressed ferritinophagy by upregulating FTH1, attenuating silica-induced pulmonary fibrosis in an in vitro model [[103]30] SLC39A8 Potential therapeutic target in pulmonary fibrosis SLC39A8 deficiency reduced AEC2 regeneration and increased lung fibrosis [[104]31] [105]Open in a new tab IGFBP-5 Insulin-like growth factor binding protein-5, ROS Reactive oxygen species, BAL Bronchoalveolar lavage, DHQ Dihydroquercetin, AEC2 Type 2 alveolar epithelial cells Correlation analysis and PPI network of DE-FRGs The relationship between ferroptosis and IPF was examined using expression correlation analysis of the DE-FRGs identified in the array. The correlation heatmap of the 13 DE-FRGs revealed significant negative correlations between CDO1 and ALOX15, CP and FTH1, CP and SLC39A8, NOX3 and LPCAT3, and STEAP3 and TXNRD1. There were significant positive correlations between FTMT and BBC3, SLC39A8 and CDO1, and SLC39A8 and FTH1 (Fig. [106]4A). Furthermore, a PPI network was constructed for the DE-FRGs. The hub nodes represented the 13 DE-FRGs and were surrounded by 20 nodes representing genes that exhibited significant correlations with them (Fig. [107]4B). Fig. 4. [108]Fig. 4 [109]Open in a new tab Correlation and PPI analysis of DE-FRGs. A Correlation analysis of DE-FRGs. B PPI network of DE-FRGs. *P < 0.05; **P < 0.01 Potential drugs targeting the DE-FRGs To investigate potential therapeutics for IPF, the 13 DE-FRGs underwent drug prediction analysis using the DGldb, which identified 42 drugs. Among these drugs, 33 targeted ALOX15, while four targeted CP, one drug each targeted TF and EPRS1, and three drugs targeted TXNRD1. Figure [110]5 depicts the drug–gene interaction network. Additional file 2: Table S2 presents detailed information on the sources, query scores, and relevant papers associated with the potential drugs. Fig. 5. [111]Fig. 5 [112]Open in a new tab The constructed drug–gene interaction network. Orange nodes represent the array-screened DE-FRGs. Green triangles represent the potential drugs identified by the DGIdb database RT-qPCR and immunohistochemistry (IHC) validation of the DE-FRGs The PCR array findings were confirmed using RT-qPCR and IHC. Among the 13 DE-FRGs identified by the array, a six-gene subset (BBC3, FTMT, STEAP3, CDO1, FTH1, SLC39A8) was randomly selected for SYBR Green-based RT-qPCR analysis and immunohistochemical staining. Figure [113]6A demonstrates that the RT-qPCR expression patterns were consistent with those obtained from the PCR array for all six genes. Consistently, the IHC results demonstrated that BBC3, FTMT, and STEAP3 protein levels were generally positively expressed in the IPF group and weakly or negatively expressed in the control group, whereas CDO1, FTH1, and SLC39A8 were positively expressed in the control group and weakly expressed in IPF (Fig. [114]6B). Fig. 6. [115]Fig. 6 [116]Open in a new tab Expression validation of DE-FRGs. A RT-qPCR verified the mRNA expression of six (randomly selected) of 13 DE-FRGs from control (n = 7) and IPF (n = 11) lung tissues. *P < 0.05; **P < 0.01; ***P < 0.001. B Immunohistochemical analysis revealed the protein expression levels of the six genes in the control (n = 7) and IPF (n = 11) lung tissues (× 100) GEO database analysis of the DE-FRGs The expression of DE-FRGs in blood was analyzed using the [117]GSE28042 dataset. Consequently, the expression analysis determined that the BBC3, STEAP3, EPRS, and SLC39A8 expression patterns in the PBMCs were consistent with those observed in our lung tissue array. The opposite was true for the CP and LPCAT3 expression trends, while no significant difference was observed for the remaining genes between the control and IPF groups (Fig. [118]7). Fig. 7. [119]Fig. 7 [120]Open in a new tab GEO database expression analysis of DE-FRGs. The mRNA expression levels of DE-FRGs in PBMCs from controls (n = 19) and IPF patients (n = 75) of the [121]GSE28042 dataset are shown. *P < 0.05; **P < 0.01; ***P < 0.001; ns, no significance Identification of diagnostic and prognostic DE-FRGs in the GEO evaluation set The discriminatory potential of individual genes in differentiating IPF from normal samples was evaluated using ROC analysis of the [122]GSE28042 dataset of the four DE-FRGs that exhibited consistent expression patterns in both lung tissue and PBMCs. Figure [123]8A demonstrates that all four genes had an AUC > 0.6, where the STEAP3 and EPRS AUCs > 0.75. Table [124]5 lists the other major parameters from the ROC analysis. The prognostic value of the four genes was assessed by survival analysis. The Kaplan–Meier survival curves demonstrated a significantly worse prognosis in patients with high STEAP3 expression compared to those with low STEAP3 expression, while patients with high EPRS expression exhibited a significantly better outcome compared to those with low EPRS expression (Fig. [125]8B). Additional file 3: Table S3 presents the data grouping for the Kaplan–Meier survival analysis. Fig. 8. [126]Fig. 8 [127]Open in a new tab Diagnostic and prognostic performance of the DE-FRGs using the GEO database. ROC analysis (A) and Kaplan–Meier survival curves (B) of the four DE-FRGs that demonstrated consistent expression patterns between lung tissue identified by our array and [128]GSE28042 PBMCs are shown Table 5. Relevant parameters of the ROC analysis Variant Specificity Sensitivity Precision Youden index Cut-off value LR +  LR- BBC3 0.53 0.85 0.79 0.38 12.73 1.80 0.28 STEAP3 0.89 0.75 0.78 0.64 8.46 7.09 0.28 EPRS 0.79 0.64 0.67 0.43 12.70 3.04 0.46 SLC39A8 0.79 0.56 0.61 0.35 10.85 2.66 0.56 [129]Open in a new tab LR + Positive likelihood ratio, LR- Negative likelihood ratio Discussion IPF is a life-threatening lung disease predominantly affecting the elderly population, imposing substantial burdens on both human and material resources, and escalating as a growing socio-economic challenge worldwide. Due to the poor prognosis and lack of effective clinical drug treatments, IPF has been characterized as a cancer-like disorder [[130]32]. Given these circumstances, it is imperative to identify efficacious therapeutic targets and reliable prognostic biomarkers and indicators of drug efficacy for IPF. Recent evidence suggested that ferroptosis is important in IPF pathogenesis and that inhibiting ferroptosis effectively attenuated pulmonary fibrosis progression in both in vivo and in vitro models [[131]18, [132]33–[133]35]. Several studies performed bioinformatic analysis by integrating FRGs with DEGs obtained from high-throughput sequencing or microarray data in the GEO database to identify critical FRGs involved in IPF pathogenesis and to elucidate potential prognostic biomarkers [[134]36–[135]38]. In this study, we present for the first time a comprehensive analysis of the FRG expression profiles of IPF patient and healthy donor lung tissue, providing new insights into IPF pathogenesis. The PCR array used in the present study included 90 key genes involved in the ferroptosis pathway of human cells, covering important BP such as iron ion transport, mitochondrial function regulation, GSH homeostasis maintenance, and redox reaction modulation. Array analysis revealed 13 DEGs in the lung tissue from the IPF patients compared to the controls, where seven genes were upregulated (ALOX15, BBC3, CP, FTMT, NOX3, STEAP3, TF) and six genes were downregulated (CDO1, EPRS, FTH1, LPCAT3, SLC39A8, TXNRD1). Most DE-FRGs identified in our array were recently associated with pulmonary fibrosis pathogenesis, where most demonstrated expression patterns consistent with our study. Table [136]4 details the major effects of the DE-FRGs on pulmonary fibrosis pathogenesis and their potential role in clinical translational applications in IPF. The DE-FRG functions were explored using GO, KEGG, and Reactome enrichment analyses to yield insights into the underlying mechanisms and prospective clinical translational applications of ferroptosis in IPF. As the array in the present study was based on detecting the expression of 90 ferroptosis-associated genes, ferroptosis was inevitably the most enriched pathway for the DEGs. The clinical translation of ferroptosis in IPF has progressed as research into ferroptosis in IPF has recently intensified. In a recent review, Hu et al. systematically summarized the potential role of therapeutic agents targeting key molecules in ferroptosis in pulmonary fibrosis treatment. A group of compounds that includes rosiglitazone and empagliflozin offered protection in in vitro and in vivo models of pulmonary fibrosis. These results suggested that these compounds may have clinical translational value in the pharmacological treatment of IPF [[137]39]. Furthermore, Hu et al. provided a comprehensive overview of the potential biomarkers associated with ferroptosis that could be used for the clinical diagnosis and prognosis of pulmonary fibrosis. For example, aconitase 1 in bronchoalveolar lavage fluid and tetrahydrobiopterin in plasma have translational value in assessing the clinical diagnosis and prognosis of IPF [[138]39]. While the porphyrin metabolism pathway has not been directly implicated in IPF, the complex mixture of the porphyrin-containing substance manganese (III) tetrakis (4-benzoate) porphyrin chloride (MnTBAP) attenuated the progression of bleomycin-induced pulmonary fibrosis by acting as a cell-permeable superoxide dismutase (SOD) mimetic and peroxynitrite scavenger [[139]40, [140]41]. These results suggested a potential therapeutic role for porphyrin metabolism in IPF. The p53 signaling pathway has been implicated in cellular senescence. Yao et al. reported abnormal p53 activation and regional alveolar type 2 (AT2) cell deficiency in lung tissue from IPF patients. Additionally, the systemic inhibition of p53 in AT2 cells, selective loss of p53 function, or ablation of senescent cells by systemic administration of senolytic drugs attenuated lung fibrosis [[141]42, [142]43]. Their study suggested that p53-induced AT2 senescence is an important driver and therapeutic target in progressive pulmonary fibrosis. Necrosis was detected in alveolar epithelial type II (AEII) cells from IPF patients associated with SFTPA1 gene mutations. In vivo studies demonstrated that Sftpa1 mutation in mice promoted necrosis in AEII cells through JNK-mediated upregulation of Ripk3, which exacerbated lung fibrosis, highlighting the necroptosis pathway as a therapeutic target in IPF [[143]44–[144]46]. Regarding the KEGG pathway taurine and hypotaurine metabolism, it was reported decades ago that taurine was protective against pulmonary fibrosis in in vivo models of bleomycin- and radiation-induced pulmonary fibrosis. These studies suggested that taurine metabolism may have a potential translational clinical role in pulmonary fibrosis treatment [[145]47–[146]50]. Significantly for the selenocompound metabolism pathway, Lin et al. recently reported that selenite administration prevented and treated lung function decline and pulmonary fibrosis in mice. Selenite exerted protective effects by inhibiting the proliferation of mouse lung fibroblasts, promoting their apoptosis, and upregulating glutathione reductase and thioredoxin reductase in the mouse lung fibroblasts [[147]51]. That study suggested that selenium metabolism may have important clinical translational value in IPF treatment. Due to the irreversible nature of IPF and its poor prognosis and the fact that the two currently approved antifibrotic agents for treating IPF (pirfenidone and nintedanib) only impede disease progression but do not cure it, the search for effective drug targets is vital [[148]52, [149]53]. Therefore, 13 DE-FRGs were analyzed using the DGIdb to identify novel therapeutic agents for IPF, and yielded 42 potential drugs. These included synthetic drugs and natural products, some with palliative effects in in vitro or in vivo models of pulmonary fibrosis. Geraniol is a potential drug targeting ALOX15 and is a monoterpene found in the essential oils of fruits, vegetables, and herbs. Geraniol has a wide range of pharmacological activities, including antimicrobial, anti-inflammatory, antioxidant, and anti-cancer [[150]54]. The essential oil of Cymbopogon winterianus, of which geraniol is a major component, attenuated bleomycin-induced pulmonary fibrosis in a murine model [[151]55]. Tryptophan is another potential drug targeting ALOX15 and is an amino acid commonly used as a component of total parenteral nutrition. Recent reports suggested that 5-methoxytryptophan, an endogenous molecule derived from tryptophan metabolism, might be a good lead compound for developing novel antifibrotic agents [[152]56]. The soy phytoestrogen isoflavone was predicted to target CP and is a biologically active component found in agriculturally important legumes such as soy and peanut. Incorporating isoflavone in mouse diet attenuated pulmonary fibrosis and improved lung function in mice exposed to hydrochloric acid [[153]57]. Dexrazoxane is another predicted drug against CP and is a cytoprotective drug used to prevent and reduce cardiotoxicity in adults and children with cancer receiving anthracyclines [[154]58, [155]59]. The potential role of dexrazoxane in protection against bleomycin-induced pulmonary fibrosis in mice has been suggested for decades [[156]60]. The IL-1β inhibitor anakinra is another predicted drug targeting CP. Anakinra reduced TGF-β1 and collagen accumulation in bleomycin-induced mice [[157]61]. Ademetionine, a potential drug targeting TF, is a glutathione precursor and has been used for treating chronic liver disease. Ademetionine effectively reversed the exacerbation of alcohol-induced lung fibrosis in a mouse model of bleomycin-induced lung fibrosis [[158]62]. Halofuginone is a potential drug targeting EPRS1 and is a low-molecular weight quinazolinone alkaloid that potently inhibits collagen type I gene expression. The role of halofuginone in preventing bleomycin-induced pulmonary fibrosis formation in vivo was indicated decades ago [[159]63]. Spermidine is a predicted drug of TXNRD1 and is a polyamine formed from putrescine and spermine precursor. Spermidine attenuated bleomycin-induced lung fibrosis in mice by inducing autophagy and inhibiting ERS-induced cell death [[160]64]. Arsenic trioxide is another predicted drug against TXNRD1 and is a naturally toxic substance that causes various dangerous side effects. Despite being a poison, arsenic trioxide is important as a chemotherapeutic agent for treating leukemia that does not respond to first-line agents [[161]65, [162]66]. Furthermore, arsenic trioxide inhibited both the functions of TGF-β1-induced lung fibroblasts in vitro and bleomycin-induced lung fibrosis in vivo [[163]67, [164]68]. While blood biomarkers aid the assessment of diagnosis, prognosis, and treatment response, none are currently integrated into clinical decision-making [[165]69–[166]72]. Accordingly, the expression of DE-FRGs in peripheral blood was validated and their value in IPF diagnosis and prognosis was validated via array-derived DE-FRGs in the [167]GSE28042 dataset. Four genes with consistent expression trends between the arrayed lung tissue and the GEO PBMCs were identified: BBC3, STEAP3, EPRS, and SLC39A8. ROC analysis demonstrated that BBC3, STEAP3, and EPRS had AUCs > 0.7, suggesting a diagnostic reference value. Furthermore, survival analysis revealed that patients with high blood STEAP3 expression had a worse prognosis, which was consistent with the upregulation of STEAP3 in the individuals with IPF. In contrast, patients with high blood EPRS expression had a more favorable prognosis, which was also consistent with the downregulation of EPRS in the IPF patients. Taken together, ferroptosis is important in IPF pathogenesis, although the underlying mechanism remains to be elucidated. Our study relied primarily on fresh, native samples obtained from transplant patients and therefore provides a valuable resource for studying the role of ferroptosis in IPF. However, our study has some limitations. First, the lack of patients’ blood samples in this study precluded the evaluation of potential biomarker expression in blood, thereby reducing the clinical translational value of the study. Second, due to the effect of the coronavirus disease 2019 (COVID-19) pandemic and time constraints, we included a small number of patient samples, which reduced statistical power. Furthermore, the donor samples were mostly from brain-dead patients with accidental injuries and who were generally younger, whereas the IPF samples were mostly from elderly patients, which prevented a good age match between the control and IPF groups. The next phase of our work will involve refining the data by increasing patient sample size and validating the expression of potential biomarkers in blood samples. Lastly, the efficacy of the drug candidates and specific mechanisms by which the DEGs regulate IPF progression through ferroptosis will be investigated via in vitro and in vivo experiments. Conclusion We characterized the FRG expression profiles in IPF and control lung tissues via PCR array analysis. We then identified the DEGs and performed functional analyses to elucidate their roles. Furthermore, we predicted potential drug targets based on these findings and integrated GEO data to identify promising blood biomarkers for IPF diagnosis and prognosis. Further experimental studies are needed to confirm our findings and to clarify the role of DE-FRGs in IPF pathogenesis. Supplementary Information [168]12890_2025_3555_MOESM1_ESM.xlsx^ (44KB, xlsx) Additional file 1. Table S1. GO, KEGG, and Reactome enrichment analyses of the DE-FRGs. [169]12890_2025_3555_MOESM2_ESM.docx^ (21.1KB, docx) Additional file 2. Table S2. Drugs predicted to interact with the DE-FRGs according to the DGIdb database. [170]12890_2025_3555_MOESM3_ESM.xlsx^ (20.2KB, xlsx) Additional file 3. Table S3. Data grouping for the Kaplan-Meier survival analysis. Acknowledgements