Bioinformatics and experimental animal model reveal the prognostic value of immunogenic cell death-related proteins in idiopathic pulmonary fibrosis

Introduction

Idiopathic pulmonary fibrosis (IPF) is a lethal and ongoing respiratory condition of uncertain origin (1). Its etiology is marked by oxidative stress, chemokines, alterations in microbial composition, smoking, sex, and environmental pollution, among other factors (2). Individuals diagnosed with IPF typically have a bleak outlook, as their chances of survival are generally limited to an average of 3–5 years (2). Therefore, more in-depth studies on prognostic markers of IPF are essential.

Immunogenic cell death (ICD), a form of regulated cell death (RCD), is triggered by external stimuli such as chemotherapy drugs (3), physical chemotherapy, photodynamic therapy, oncolytic viruses (4), and radiotherapy. Dying or stressed cells release molecules that can function as either adjuvants or danger signals for the immune system to activate innate or adaptive immune responses (5). ICD has been shown to be involved in malignant diseases, including colorectal cancer, melanoma, non-small cell lung cancer, and breast cancer; certain infectious diseases caused by harmful microorganisms, such as herpes simplex virus-1 and Salmonella enterica; and fibrosis (6). Although a previous study revealed that immune cell infiltration and fibrosis are positively correlated with the degree of cell death, including ICD, in arrhythmogenic cardiomyopathy, no studies have confirmed the relationship between ICD and IPF (7). Notably, several studies (8) demonstrated that innate and adaptive responses play important roles in IPF (9). Alveolar macrophages create fibrosis-promoting substances and chemokines as a result of lung epithelial damage, which thickens the interstitial lining of the lung, promoting fibrosis (10). CD8⁺ T cells appear to suppress fibrosis; however, interleukin-17 (IL-17) produced by T helper cell 17 (Th17) cells has been shown to promote the progression of fibrosis. Immune checkpoint inhibitors have been given to mice to reduce fibrosis (11). A literature review revealed potential associations between several ICD-related gene expression profiles and IPF. Specifically, lycorine has been reported to effectively ameliorate bleomycin (BLM)-induced pulmonary inflammation and fibrosis progression by inhibiting NLRP3 inflammasome activation and pyroptosis in macrophages (12). IL10, recognized for its established role as an anti-inflammatory mediator, represents a focus for potential antifibrotic therapies (13). The elevated expression of interleukin-1 beta (IL-1β) and CASP1 observed in IPF patients suggests that inflammasome activation is associated with this disease (14). Furthermore, tumor necrosis factor-alpha (TNF-α) acts as a key regulator in IPF, mediating myofibroblast differentiation and promoting pulmonary fibrotic remodeling (15). These results suggest that ICD, which activates innate or adaptive immune responses, is more likely to play a crucial role in the development of IPF.

In this study, we aimed to identify ICD-related gene expression signatures that are associated with the progression and prognosis of IPF. Additionally, we constructed a prognostic signature to assess the impact of ICD on IPF. A functional enrichment analysis was subsequently conducted to explore the potential impacts of these risk-associated differentially expressed genes (DEGs) on biological functions and signaling pathways. Various immune microenvironment patterns were subsequently analyzed to gain a deeper understanding of the underlying correlation between the risk score and the immune response. Finally, the mRNA and protein expression levels of all genes in the ICD-related risk signature were assessed in mouse models of BLM-induced pulmonary fibrosis. Through these methods, our aim is to investigate biomarkers that can assist in the diagnosis, treatment, or prognosis of IPF. We present this article in accordance with the TRIPOD and ARRIVE reporting checklists (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-616/rc).

Methods

Data collection and ICD-related gene exploration

This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. The GSE70866 dataset, obtained from the Gene Expression Omnibus database (GEO database) (https://www.ncbi.nlm.nih.gov/geo/), contains BALF gene expression profiles and clinical survival data from 176 patients with IPF and 20 healthy controls. Notably, this is the only GEO dataset with longitudinal clinical follow-up data available for IPF, making it uniquely suited for prognostic model construction. The GSE70867 dataset was used as an independent validation set to evaluate the prognostic value of selected key molecules in a separate IPF cohort with corresponding clinical annotations. In total, 34 ICD-related genes were curated from a published literature source (16) and are listed in Table S1. And we have also added the basic information tables of patients in these two datasets in Table S2. In addition, the term ‘gene’ in this study refers to its functional transcriptional output (mRNA) and encoded protein product, not to genetic variations in its DNA sequence.

Identification of DEGs

The “limma” package was used to explore DEGs with the P<0.05 screening criterion. The results are visualized by the “heatmap” package. Univariate and multivariate Cox regression with the “survival” package was used to identify prognosis-related genes (PRGs); subsequently, two groups of DEGs and PRGs were used to construct a Venn map to search for intersecting genes. The “survival” and “heatmap” packages were used to construct forest maps and expression levels of the intersecting genes, respectively.

PPI network and gene correlation

To explore the interactions among the overlapping genes, we constructed (I) a protein-protein interaction (PPI) network by the STRING database (https://string-db.org/) and (II) a gene coexpression network by the “igraph” R package. The coexpression network was constructed by connecting gene pairs that were significantly positively correlated (Pearson r>0.3 and P<0.05). This dual-network approach facilitated the investigation of both direct protein binding and coordinated gene expression patterns within the gene set of interest.

Construction and validation of the ICD-related risk signature

To further evaluate the prognostic importance of ICD-associated gene expression in IPF, the GSE70866 dataset included a cohort of 176 IPF patients who were divided into training and validation groups at a 7:3 ratio. GSE70867 was also used to validate the model’s effectiveness. A prognostic risk signature was constructed by least absolute shrinkage and selection operator (LASSO)-penalized Cox regression in the training cohort. The optimal λ value, selected through cross-validation to balance model complexity and predictive performance, was used to finalize the model and extract the regression coefficients for the three overlapping genes. The R package “rms” was subsequently employed to construct a nomogram. This nomogram visualizes the combined effect of the three genes (weighted by their coefficients) to compute a patient’s risk score and estimate survival probability. To quantify the nomogram’s calibration accuracy (agreement between the predicted and observed outcomes), calibration curves were generated, and the predicted survival probabilities were compared with the actual Kaplan-Meier estimates within risk-stratified groups.

Prognostic analysis of the ICD-related risk signature

Risk scores were calculated based on the normalized mRNA expression levels of the signature genes and their corresponding coefficients, the medians of which were utilized to divide IPF patients in the training group into high-/low-risk groups. t-distributed stochastic neighbor embedding (t-SNE) and principal component analysis (PCA) were used to depict the two subgroups. Kaplan-Meier (KM) analysis was performed by the “survminer” and “survival” packages to assess overall survival (OS) between risk cohorts. The ability of the signature to predict 1-, 2-, and 3-year survival rates in the two risk groups was evaluated using time-dependent receiver operating characteristic (ROC) curves. mRNA expression levels of the signature genes in the high- and low-risk groups are shown as a heatmap generated by the “pheatmap” package. All of the above analyses were verified in the validation group. Spearman analysis was performed using Prism 9.0 to verify the relationship between gene expression and risk score.

Independent prognostic analysis

Univariate Cox proportional hazards regression was first employed to screen clinical indicators and risk scores for potential associations with IPF prognosis. Variables demonstrating nominal significance (P<0.05) were subsequently entered into multivariate Cox regression models to identify independent prognostic factors. The final model incorporated only covariates that maintained statistical significance (P<0.05) after adjustment for confounders, with the results expressed as hazard ratios (HRs) and 95% confidence intervals. All analyses were implemented using the survival package in R.

Functional enrichment

To systematically identify risk-associated DEGs between the established high- and low-risk groups, we performed comparative transcriptomic analyses in both the training and validation cohorts using the R package “limma”. This empirical Bayes-based framework was applied to model gene expression data, employing moderated t-tests with Benjamini-Hochberg false discovery rate (FDR) correction. DEGs were stringently filtered using two thresholds: an FDR-adjusted P value <0.01 and an absolute log₂-fold change (|log₂FC|) >0.3. Genes meeting these criteria in either cohort were consolidated into a union set of risk-related DEGs for downstream functional annotation. The R package “clusterProfiler” was subsequently utilized to conduct comprehensive enrichment analyses on this gene set. Gene Ontology (GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis were executed to elucidate the biological functions, cellular locations, molecular activities, and signaling pathways significantly associated with the identified risk-related DEGs. The significance of the enrichment results was determined using a conservative FDR threshold of <0.01.

Different immune microenvironment patterns between risk groups

Tumor immune cell infiltration was quantified using CIBERSORT, a deconvolution algorithm that estimates the relative proportions of 22 immune cell types (LM22 signature matrix) in bulk transcriptome data through support vector regression. Samples with CIBERSORT P values <0.01 were retained for high-confidence estimates. To functionally characterize immune pathways, we performed gene set variation analysis (GSVA) using the “GSVA” R package. This method calculates single-sample enrichment scores for curated immune gene sets (e.g., antigen presentation, cytokine signaling) via ssGSEA. Differences in pathway activity between sample groups were assessed using Wilcoxon rank-sum tests with FDR correction (significance threshold: FDR <0.05).

Animal models

Consistent with reported studies (17), BLM, which is capable of inducing histological lung patterns similar to those observed in chemotherapy patients, is one of the most widely used agents for inducing pulmonary fibrosis in animal models. It exerts its cytotoxic effects by cleaving DNA, triggering inflammatory responses, and increasing epithelial apoptosis in a dose-dependent manner, thereby stimulating lung injury and ultimately leading to fibrosis. Considering that C57BL/6 mice have increased sensitivity to bleomycin, the selection of male mice avoids the effect of changes in the estrogen cycle on the construction of animal models of bleomycin. Eight-week-old specific-pathogen-free (SPF)-grade male C57BL/6 mice were raised at the Experimental Animal Center of Xi’an Jiaotong University Health Science Center for at least 2 weeks under controlled conditions (23–25 ℃, relative humidity of 50%, and a 12-h night/day cycle), and food and water were provided before the experiment began. Six mice were randomly divided into a control group (saline) and an intratracheally administered bleomycin group (Selleck, NSC125066, 5 mg/kg). On day 28, the mice were euthanized, and lung tissue was subsequently collected and stored at −80 ℃. All animal experiments were performed under a project license (No. 2024-028) granted by the Animal Ethics Committee of The First Affiliated Hospital of Xi’an Jiaotong University, in compliance with the national guidelines for the care and use of animals and were conducted in full compliance with the U.K. Animals (Scientific Procedures) Act, 1986 and associated guidelines, EU Directive 2010/63/EU for animal experiments. A protocol was prepared before the study without registration.

H&E and Masson staining

The harvested lung tissues were immediately fixed in 4% paraformaldehyde (Meilunbio, MA0192, Dalian, China) overnight, dehydrated, and embedded in paraffin. Sections (4 µm) were stained with H&E (Novland, IH-017 and IH-018; Shanghai, China) or a modified Masson’s trichrome staining kit (Solarbio, G1345; Beijing, China). Images were obtained under a microscope.

Western blotting

Protein extracts from lung tissue were separated by 8–12% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred to polyvinylidene fluoride (PVDF) membranes (Millipore Corp, IPVH00010). The membranes were subsequently blocked with 5% skim milk in Tris Buffered Saline with Tween (TBST) solution for 1 h at room temperature and incubated overnight at 4 ℃ with the primary antibodies shown in Table S3, followed by incubation at room temperature for 1 h with corresponding secondary antibodies. Finally, the immunoreactive protein bands were visualized with an ECL-PLUS/Kit (Thermo, M3121/1859022). Uncropped gel and blot images are shown in Appendix 1.

Real-time reverse transcription polymerase chain reaction (qRT-PCR)

The mRNA levels of the genes were quantified using quantitative real-time reverse transcription polymerase chain reaction (qRT-PCR) and the ChamQ SYBR qPCR Master Mix Kit (Q311‒02; Vazyme). The 2^−∆∆Ct formula was used to quantify the relative expression of the mRNA levels. The forward and reverse sequences of the primers used for qRT-PCR are shown in Appendix 2.

Statistical analysis

Perl, R 4.22, and Prism 9.0 were utilized for data analysis. All experimental and clinical data were analyzed using GraphPad Prism 5 software from GraphPad Software, Inc., located in San Diego, CA, USA. R language (version 4.2.1) was utilized for bioinformatics analysis, excluding descriptive analysis. We utilized various statistical methods, such as t-tests, two-way analysis of variance (ANOVA), and the Benjamini-Hochberg method, to modify the P values, including the FDR, and P<0.05 was considered significant.

Results

Data collection and preprocessing

Figure 1 illustrates the flow of our research. Following differential (P<0.05, logFC =0) and univariate (P<0.05) analyses in IPF patients, we discovered 10 DEGs and 12 PRGs (Figure 2). Six DEGs (IL10, CASP1, NLPR3, IL1R1, CALR, and NT5E) were found to be associated with the prognosis of IPF according to the Venn diagram (Figure 3A). A heatmap and forest diagram revealed the expression and hazard ratios of the six intersecting genes, respectively (Figure 3B,3C). A PPI network was constructed by the STRING database (Figure 3D). To assess the potential interactions among the six prognostic genes, we calculated the Pearson correlation coefficient with a cutoff value of 0.2 (Figure 3E). The results revealed a positive correlation between the mRNA expression level of IL10 and those of the other genes.

Figure 1 Flow chart of this study. AUC, area under the curve; BLM, bleomycin; GEO, Gene Expression Omnibus; ICD, immunogenic cell death; IPF, idiopathic pulmonary fibrosis; LASSO, least absolute shrinkage and selection operator.

Figure 2 Identification of prognostic ICD-related DEGs in IPF. (A) The expression levels of 12 ICD-associated DEGs in IPF patients were shown in the form of heat maps. (B) Forest plot presents the results of univariate Cox regression of the ICD-related prognosis of IPF patients. CI, confidence interval; DEGs, differentially expressed genes; ICD, immunogenic cell death; IPF, idiopathic pulmonary fibrosis.

Figure 3 Prognostic ICD-associated DEGs in IPF and their interrelationships. (A-C) Six ICD-associated DEGs were identified by Venn map and their expression levels were analyzed. (D,E) The correlation of six ICD-associated DEGs in total database and GSE70866 dataset. CI, confidence interval; DEGs, differentially expressed genes; ICD, immunogenic cell death; IPF, idiopathic pulmonary fibrosis.

Construction of the ICD-related risk prognostic signature

The GSE70866 dataset was divided into training and validation groups at a 7:3 ratio. Three ICD-related gene expression products, CASP1, IL10, and NLRP3, were selected by LASSO regression for prognostic signature construction (Figure 4A,4B). In contrast to IL10 and NLRP3, which were identified as risk variables by univariate regression analysis, CASP1 was identified as a protective factor (Figure 4C). Three genes were used to create ICD-dependent risk profiles (ICDRS) with projected 1-, 2-, and 3-year survival probability nomograms (Figure 4D). The calibration curve is near the line of Y=X, demonstrating the model’s high agreement (Figure 4E).

Figure 4 Establishment of the ICDRS by following processes. LASSO Cox analysis identified 3 genes most associated with OS in GEO dataset (A), and the optimal λ was selected by Cross-validation (B). (C) Cox regression highlighted the correlation between ICD-regulators and IPF patients, which included 3 ICD-related genes (***, P<0.001; *, P<0.05). (D) The nomogram model for the prognosis of IPF based on CASP1, IL10, and NLRP3 gene products. (E) Calibration curve for nomogram validation. AIC, Akaike Information Criterion; GEO, Gene Expression Omnibus; ICD, immunogenic cell death; ICDRS, ICD-dependent risk profiles; IPF, idiopathic pulmonary fibrosis; LASSO, least absolute shrinkage and selection operator; OS, overall survival.

Prognostic value of the ICD-related risk prognostic signature

On the basis of the median risk score, IPF patients in the training set were divided into high- and low-risk groups (Figure 5A), with the low-risk group having a greater chance of survival (Figure 5B). PCA (Figure 5C) and t-SNE (Figure 5D) analyses clearly distinguished between the high- and low-risk groups. The Kaplan-Meier survival curve revealed that the low-risk group had a better survival prognosis (Figure 5E). The ROC curve in the training set achieved an area under curve (AUC) value of 0.771 for the 1-year survival rate, 0.734 for the 2-year survival rate, and 0.629 for the 3-year survival rate (Figure 5F). The heatmap displays the differences in the mRNA expression levels of the three genes across the training cohorts (Figure 5G). Similar to the training cohort, this model exhibited the same discriminative performance in the GSE70867 validation cohort and the internal validation cohort samples (Figure 6A-6D), effectively distinguishing the samples in the validation cohort. The AUC in GSE70867 was 0.758 for the 1-year score, 0.731 for the 2-year score, and 0.669 for the 3-year score (Figure 6E,6F). The internal validation cohort had an AUC of 0.733 for the 1-year score, 0.718 for the 2-year score, and 0.750 for the 3-year score. The heatmap displays the differences in the mRNA expression levels of the three genes across the GSE70867 and test cohorts (Figure 6G,6H). Furthermore, the correlation analysis further confirmed that CASP1 was negatively correlated with the risk score, while NLRP3 and IL10 were positively correlated with the risk score (Figure S1).

Figure 5 The training sets’ relationships between ICD-related genes signature and prognosis. (A,B) Risk scores distribution, overall survival were shown. (C,D) The ICDRS model effectively differentiates between the risk groups using PCA and tSNE. (E) Risk model’s prognostic significance in the training set is demonstrated by Kaplan-Meier analyses. (F) ROC curves for prognostic models based on ICDRS, predicting IPF at 1-, 2-, and 3-year intervals. (G) Heatmap of prognostic 3-gene ICDRS in training set. AUC, area under the curve; ICD, immunogenic cell death; ICDRS, ICD-dependent risk profiles; IPF, idiopathic pulmonary fibrosis; PCA, principal component analysis; ROC, receiver operating characteristic; t-SNE, t-distributed stochastic neighbor embedding.

Figure 6 The validation sets’ relationships between ICD-related genes signature and prognosis. (A-D) Risk scores distribution, overall survival between GSE70867 and test group. (E,F) ROC curves of 1-, 2- and 3-year prognostic models based on ICDRS for predicting IPF were identified. (G,H) Heatmaps of prognostic 3-gene ICDRS in validation set. AUC, area under the curve; ICD, immunogenic cell death; ICDRS, ICD-dependent risk profiles; IPF, idiopathic pulmonary fibrosis; ROC, receiver operating characteristic.

Correlations between the ICD risk prognostic model and clinical factors

Univariate and multivariate Cox regression (Figure 7) were used to analyze age, sex, and risk score. After adjusting for these variables, the ICD-derived risk score was identified as an autonomous risk factor for IPF in both the training and validation groups, as indicated by multifactor Cox regression analysis.

Figure 7 Univariate and multivariate Cox analyses evaluate an independent prognostic value of ICDRS in IPF patients in the training set (A,B) and the validation set (C,D). CI, confidence interval; ICD, immunogenic cell death; ICDRS, ICD-dependent risk profiles; IPF, idiopathic pulmonary fibrosis.

Functional enrichment analysis

GO and KEGG analyses were used to investigate the biological processes and pathways connected to risk-related genes (68 from the training group and 121 from the testing group) (Table S4), respectively. In the training group, GO analysis revealed that the risk genes predominantly clustered in the regulation of hormone biosynthetic processes, leukocyte migration, and regulation of hormone metabolic processes, whereas the risk genes clustered in the positive regulation of endothelial cell proliferation, positive regulation of epithelial cell proliferation, and endothelial cell proliferation in the validation group (Figure 8A,8B). The top three KEGG pathways were related to Staphylococcus aureus infection, cytokine-cytokine receptor interactions and viral protein interactions with cytokines in the validation and training groups (Figure 8C,8D).

Figure 8 Bar chart of GO analysis based on the risk score-related DEGs in the training sets (A) and in the validation sets (B). Bar chart of KEGG analysis results based on the risk score-related DEGs in the training sets (C) and in the validation sets (D). BP, biological process; CC, cellular component; DEGs, differentially expressed genes; GO, Gene Ontology; KEGG, Kyoto Encyclopedia of Genes and Genomes; MF, molecular function.

Immune analysis

To further clarify the connection between the risk score and the immune response, as previously explained (18), CIBERSORT analysis was utilized to quantify immune cell infiltration on the basis of gene expression data. Morphologically diverse samples (Figure 9A) exhibited an increase in T-cell count and a reduced percentage of monocyte cells. Figure 9B shows the interactions among immune cells. The activation of dendritic cells and B cells showed a strong positive correlation with each other (r=0.46). On the other hand, monocytes exhibited the most pronounced inverse association with M2 macrophages (r=−0.59). Significant recognition of naïve B cells, NK cells, monocytes, M0 macrophages, dendritic cells, and mast cells was observed in the risk groups during immune cell differential analysis. The low-risk group presented a decrease in activated mast cells and NK cells (Figure 9C). Significant recognition of antigen-presenting cell (APC) costimulation, chemokines (CCRs), checkpoints, cytolytic activity, major histocompatibility complex (MHC) class I, T-cell coinhibition, and T cells was observed in the risk groups according to immune functional differential analysis. However, all of these immune functional results were decreased in the high-risk group (Figure 9D).

Figure 9 The immune infiltration and function analysis in training cohort. (A) Stacking diagram by CIBERSORT algorithm. (B) Correlation matrix of ratios of immune cells. (C) Box-plot showing immune infiltration levels differ in risk groups. (D) The immune function analysis based on ssGSEA. *, P<0.05; **, P<0.01; ***, P<0.001; ****, P<0.0001. APC, antigen-presenting cell; CCR, C-C chemokine receptor; HLA, human leukocyte antigen; MHC, major histocompatibility complex; NK, natural killer; ssGSEA, single sample gene set enrichment analysis.

Validation of signature gene expression levels

H&E and Masson staining revealed thicker alveolar walls in the BLM group than in the saline group. Our mouse fibrosis model was successfully generated using bleomycin-induced fibrosis (Figure 10A). The real-time quantitative PCR results revealed that the mRNA levels of IL10, CASP-1, and NLRP3 were greater in the BLM group than in the negative control group (Figure 10B-10D). Compared with the Western blot result for the negative control group, IL10, CASP-1, and NLRP3 protein expression was upregulated in the BLM-induced pulmonary fibrosis group (Figure 10E,10F).

Figure 10 Verify the mRNA expression levels of the three signature genes. (A) H&E and Masson staining demonstrated that a portion of the alveolar structure was damaged in the lung of the BLM mice compared with the saline group. The qRT-PCR results of CASP1 (B), IL10 (C), and NLRP3 (D). The Western blot images results of CASP1, IL10, and NLRP3 in negative control and BLM:28 Day (E). Depression levels of CASP1, IL10, and NLRP3 in negative control and BLM:28 Day (F). The three signature genes and β-actin protein expression levels by western blotting. Scale bar: 50 μm. *, P<0.05; **, P<0.01; ****, P<0.0001. BLM, bleomycin; H&E, hematoxylin and eosin; qRT-PCR, quantitative real-time reverse transcription polymerase chain reaction.

Discussion

IPF places a significant cost on society because of its limited treatment options and high mortality rate; therefore, the exploration of new potential therapeutic targets and molecular mechanisms of IPF is urgently needed (19). To date, the correlations of various RCD mechanisms such as apoptosis (20), autophagy, and ferroptosis (21) with IPF have been investigated (22). However, research on the function of ICD in benign conditions, especially IPF, is scarce. In the present study, the mRNA expression levels ICD-related genes IL10 and NLPR3 were overexpressed, while CASP1 mRNA was downregulated in the BALF of IPF patients compared with healthy controls. The expression levels of these genes were also found to be correlated with the prognosis of IPF; subsequently, an ICD-related risk signature was constructed and validated based on the expression profiles of these ICD-related molecules. In addition, the protein levels of the signature molecules were assessed in mouse lung tissues with BLM-induced pulmonary fibrosis. These results suggest that ICD plays an important role in the progression of IPF and that this model can be used as a biomarker to assess the prognosis of IPF in patients. A recent study by Xie et al. explored transcriptomic mechanisms linking IPF and lung cancer using several GEO datasets, including GSE70867 (23). In contrast, our study focuses specifically on the prognostic role of ICD-related molecules in IPF. To our knowledge, this is the first work to establish an ICD-based prognostic signature for IPF and to validate it using both bioinformatics and experimental approaches.

NLRP3 (24) proteins are sensors that, when sensing stimuli, together with pre-CASP1 and ASC, constitute NLRP3 inflammatory vesicles that activate CASP1 and produce cytokines such as IL-1β and IL18. IL-1β produces cytokines through the activation of TGF-β, which induces epithelial mesenchymal transition to produce an inflammatory response and promote fibrosis; however, this process can be affected by NLRP3 inhibitors (25). Previous studies have shown that NLRP3 is overexpressed both in lung tissue from patients with IPF and in a BLM-induced mouse model (26), which suggests that NLRP3 is a novel therapeutic target for IPF. NT-0167, an inhibitor of NLRP3, is being evaluated in a phase 1 clinical trial in healthy volunteers with plans for the treatment of IPF (27). Moreover, elevated extracellular ATP, one of the characteristic markers of immunogenic death, was detected in alveolar lavage fluid from IPF patients, suggesting that ICD is possibly associated with IPF (28). In the present study, we found that NLRP3 is highly expressed in both BALF and IPF tissues, which is consistent with the findings of previous studies, but the underlying mechanism needs to be further explored.

CASP1, also known as interleukin-1β converting enzyme, is a cysteine protease that plays a crucial role in the inflammasome signaling pathway (29) and has been implicated in the regulation of ICD, suggesting its potential involvement in the immune response against fibrotic processes. The potential impact of CASP1 in pulmonary fibrosis lies in its ability to modulate the inflammatory milieu and promote fibrotic tissue remodeling. By activating CASP1, the release of proinflammatory cytokines may perpetuate the inflammatory response, leading to the recruitment and activation of immune cells involved in fibrotic processes (30); hence, the use of inhibitors targeting CASP1 may constitute a viable therapeutic strategy for many inflammatory and autoimmune diseases. VX-765, a CASP1 inhibitor, has been reported to inhibit diabetic kidney injury and renal fibrosis in diabetic patients by modulating pyroptosis (31), as well as silica-induced pulmonary injury and pulmonary fibrosis in mice (32). However, CASP1 expression was found to be suppressed in the BALF at the RNA level according to differential analysis but enhanced in the IPF tissue according to our Western blot validation. First, Sang Yeon Kim (33) reported that NXC736 attenuates radiation-induced lung fibrosis by regulating the NLRP3/IL-1β signaling pathway, and Xu et al. (34) reported that statin administration exacerbates lung injury and fibrosis in bleomycin-treated mice and enhances caspase-1 expression in fibrotic lung tissues. All of the above studies reported that CASP1 was expressed at higher levels in IPF tissues than in normal samples. Finally, in the IPF samples in GSE70866, the source of the samples was alveolar lavage fluid, and there were individual sample variations; however, CASP1 in the results of the differential analysis was significant enough to be screened for calculation of the risk score. The mRNA expression trend was consistent with the protein expression trend in the experimental validation of this study. It is not clear whether this difference is due to individual heterogeneity or a feedback mechanism after immune activation, so further experimental studies are warranted to validate the role of CASP1 in pulmonary fibrosis and ICD, which will provide a deeper understanding of its potential as a therapeutic target.

IL10 has previously been shown to exert both anti-inflammatory and fibrosis-inhibiting effects (35). Some studies (36) have demonstrated that IL10 protein levels in IPF vary across samples (37), with endogenous IL10 protein elevated in lung tissue and serum and at lower or slightly increased levels in BALF. Therefore, it remains uncertain whether IL10 has a promoting or inhibiting effect on IPF (38). Exogenous IL10 inhibits alveolar macrophage production and activation of TGF-β1, thereby alleviating pulmonary fibrosis, which has also been shown to prevent BLM-induced fibrosis and prolong survival in mice by downregulating IFN-γ and upregulating TGF-β1 expression. Moreover, hydrogel-based IL10 inhibits lung fibroblast and myofibroblast viability and reduces TGF-β activation, causing a decrease in collagen synthesis (39). In contrast, other studies have shown that high expression levels of IL10 accelerate the fibrosis process by promoting MRC5 cell and primary lung fibroblast activity and collagen synthesis. Long-term overexpression of IL10 can promote fibrosis by activating M2 macrophages. Our study suggests that high levels of IL10 are associated with the promotion of IPF progression and could be an objective basis for disease determination, the mechanism of which needs to be further corroborated.

The functional analysis revealed that the risk-related DEGs were enriched mainly in processes such as hormone synthesis, white blood cell migration, endothelial cell proliferation, and epithelial cell proliferation regulation (39). All of these pathways were identified as the most essential in fibrosis development (40). These findings suggest that the expression of ICD-related genes at the mRNA level might control the progression and growth of IPF by influencing these primary pathways.

In this study, immune-related analysis revealed significant differences in monocytes, M0 macrophages, mast cells, and NK cells between the high- and low-risk groups. Kreuter et al. reported that the monocyte count is a prognostic biomarker for IPF; elevated monocyte counts are associated with increased risk of IPF progression, hospitalization, and death (41). Monocyte infiltration was greater in the high-risk group than in the low-risk group, and when combined with survival analyses, higher monocyte infiltration was accompanied by poorer survival outcomes, which corresponds with previously reported studies (42). Mansouri’s study revealed that mesenchymal stromal cell exosomes can prevent and reverse bleomycin-induced experimental pulmonary fibrosis by modulating the monocyte phenotype (43). The present study revealed that mast cells differed significantly between the high- and low-risk groups, with the high-risk group demonstrating greater mast cell infiltration, a finding that is consistent with Salonen’s findings (44). Not only are mast cells involved in and a contributing factor to the progression of pulmonary fibrosis, but there is also a decrease in mast cell counts during acute exacerbations. NK cells are thought to attenuate the progression of bleomycin-induced pulmonary fibrosis (BIPF) by supplying antifibrotic mediators and cytokines (e.g., IFN-γ), but Monnier’s study (45) suggested that depletion of NK cells with anti-Asialo GM1 does not alter fibrosis progression or affect any measured levels of proinflammatory/profibrotic cytokines. In addition, Galati’s study confirmed that the proportion and absolute number of NK cells were significantly lower in patients with IPF than in healthy controls, suggesting that NK cells may not contribute to the progression of pulmonary fibrosis (46). When differentiated from monocytes, macrophages are crucial components of the immune system and are involved in phagocytosis, pathogen clearance, the regulation of inflammatory responses, and tissue repair. In the context of pulmonary fibrosis, the degree of macrophage infiltration is closely associated with the extent of fibrosis (47). The increased presence of macrophages in the low-risk group observed in this study may reflect their enhanced anti-inflammatory and reparative capabilities, thereby contributing to the attenuation of fibrosis. When tissues are stimulated by injury or inflammation, M0 macrophages can be subjected to specific stimulatory factors, such as the cytokines IL-4 and IL-13, and differentiate into M2 macrophages. M2 macrophages can promote fibroblast proliferation and collagen deposition, which leads to the progression of fibrosis (48). This finding seems to align with the survival analysis results obtained from the training cohort in this study. Therefore, the increased presence of macrophages in the low-risk group may be achieved by enhancing the activity and quantity of M0 macrophages, thereby improving the prognosis in patients with IPF. To provide a clearer overview of the interactions between key immune cells, ICD-related molecules, fibroblasts, and the fibrotic process, a proposed mechanism diagram has been added (Figure 11). Further studies are needed to elucidate the roles of immune cells in IPF, which may help identify relevant potential targets for drugs.

Figure 11 Proposed mechanism of ICD-related proteins regulating immune microenvironment and fibrosis in IPF. In high-risk IPF patients, activation of ICD leads to the release of DAMPs, which stimulate the NLRP3-CASP1-IL1β signaling pathway. This activation is associated with increased infiltration of monocytes and mast cells, along with a reduction in NK cell levels. These immune alterations are linked to enhanced M2 macrophage polarization and elevated production of IL10 and IL13. The resulting immune environment weakens antigen presentation and cytotoxic responses, thereby facilitating fibroblast activation and progressive fibrotic remodeling. APC, antigen-presenting cell; ATP, adenosine triphosphate; DAMPs, damage-associated molecular patterns; ICD, immunogenic cell death; IPF, idiopathic pulmonary fibrosis; MHC, major histocompatibility complex; NK, natural killer.

This study has several strengths. It is the first to systematically investigate the prognostic relevance of ICD-related genes in IPF, and the constructed prognostic model demonstrates superior predictive power compared to previously published models. Additionally, we validated the expression of the three signature genes at the mRNA and protein levels in a bleomycin-induced pulmonary fibrosis mouse model, reinforcing the robustness of our findings. Nonetheless, certain limitations should be acknowledged. First, external validation was limited by the lack of comparable datasets containing adequate clinical follow-up and sufficient healthy control samples. Second, although our protein-level validation using Western blotting provided important confirmatory evidence, immunohistochemistry or immunofluorescence assays were not performed in this study. These techniques could offer further insights into spatial distribution and cell-type specificity of CASP1, IL10, and NLRP3 in lung tissues. However, as our current aim was to focus on prognostic modeling and molecular validation at a systemic level, detailed spatial characterization will be pursued in future investigations. Lastly, expression of the signature genes in BALF from the animal model was not assessed, and additional mechanistic experiments are needed to explore the functional contributions of these molecules to IPF progression. These aspects will be addressed in future prospective studies to further validate and refine the proposed ICD-based prognostic signature.

Conclusions

In conclusion, our study revealed that three ICD-related molecules (IL10, CASP1, and NLRP3) might function as biomarkers for the prognosis of IPF, and additional molecular biology experiments are needed to confirm the functions and mechanisms of these genes in IPF.

Acknowledgments

We thank the editor and the anonymous reviewers for their comments and suggestions.

Footnote

Reporting Checklist: The authors have completed the TRIPOD and ARRIVE reporting checklists. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-616/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-616/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-616/prf

Funding: This study was supported by the Key Clinical Research Projects of The First Affiliated Hospital of Xi’an Jiaotong University (No. XJTU1AF-CRF-2023-006 to G.Z.), Xi’an Jiaotong University Basic-Clinical Integration Innovation Program (No. YXJLRH2022033 to G.Z.), and Institutional Foundation of The First Affiliated Hospital of Xi’an Jiaotong University (No. 2024-MS-15 to Y.Z.).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-616/coif). G.Z. reports that this study was supported by the Key Clinical Research Projects of The First Affiliated Hospital of Xi’an Jiaotong University (No. XJTU1AF-CRF-2023-006) and Xi’an Jiaotong University Basic-Clinical Integration Innovation Program (No. YXJLRH2022033). Y.Z. reports that this study was supported by Institutional Foundation of The First Affiliated Hospital of Xi’an Jiaotong University (No. 2024-MS-15). The other authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments. All animal experiments were performed and approved by the Animal Ethics Committee of The First Affiliated Hospital of Xi’an Jiaotong University (No. 2024-028) in compliance with the national guidelines for the care and use of animals and were conducted in full compliance with the U.K. Animals (Scientific Procedures) Act, 1986 and associated guidelines, EU Directive 2010/63/EU for animal experiments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Richeldi L, Collard HR, Jones MG. Idiopathic pulmonary fibrosis. Lancet 2017;389:1941-52. [Crossref] [PubMed]
2. Lederer DJ, Martinez FJ. Idiopathic Pulmonary Fibrosis. N Engl J Med 2018;378:1811-23. [Crossref] [PubMed]
3. Green DR, Ferguson T, Zitvogel L, et al. Immunogenic and tolerogenic cell death. Nat Rev Immunol 2009;9:353-63. [Crossref] [PubMed]
4. Tesniere A, Panaretakis T, Kepp O, et al. Molecular characteristics of immunogenic cancer cell death. Cell Death Differ 2008;15:3-12. [Crossref] [PubMed]
5. Krysko DV, Garg AD, Kaczmarek A, et al. Immunogenic cell death and DAMPs in cancer therapy. Nat Rev Cancer 2012;12:860-75. [Crossref] [PubMed]
6. Zhou M, Luo C, Zhou Z, et al. Improving anti-PD-L1 therapy in triple negative breast cancer by polymer-enhanced immunogenic cell death and CXCR4 blockade. J Control Release 2021;334:248-62. [Crossref] [PubMed]
7. Lu W, Rao Y, Li Y, et al. The Landscape of Cell Death Processes with Associated Immunogenic and Fibrogenic Effects in Arrhythmogenic Cardiomyopathy. J Cardiovasc Dev Dis 2022;9:301. [Crossref] [PubMed]
8. Desai O, Winkler J, Minasyan M, et al. The Role of Immune and Inflammatory Cells in Idiopathic Pulmonary Fibrosis. Front Med (Lausanne) 2018;5:43. [Crossref] [PubMed]
9. Planté-Bordeneuve T, Pilette C, Froidure A. The Epithelial-Immune Crosstalk in Pulmonary Fibrosis. Front Immunol 2021;12:631235. [Crossref] [PubMed]
10. Shenderov K, Collins SL, Powell JD, et al. Immune dysregulation as a driver of idiopathic pulmonary fibrosis. J Clin Invest 2021;131:e143226. [Crossref] [PubMed]
11. Lettieri S, Oggionni T, Lancia A, et al. Immune Stroma in Lung Cancer and Idiopathic Pulmonary Fibrosis: A Common Biologic Landscape? Int J Mol Sci 2021;22:2882. [Crossref] [PubMed]
12. Liang Q, Cai W, Zhao Y, et al. Lycorine ameliorates bleomycin-induced pulmonary fibrosis via inhibiting NLRP3 inflammasome activation and pyroptosis. Pharmacol Res 2020;158:104884. [Crossref] [PubMed]
13. Steen EH, Wang X, Balaji S, et al. The Role of the Anti-Inflammatory Cytokine Interleukin-10 in Tissue Fibrosis. Adv Wound Care (New Rochelle) 2020;9:184-98. [Crossref] [PubMed]
14. Lin X, Barravecchia M, Matthew Kottmann R, et al. Caveolin-1 gene therapy inhibits inflammasome activation to protect from bleomycin-induced pulmonary fibrosis. Sci Rep 2019;9:19643. [Crossref] [PubMed]
15. Saha P, Talwar P. Idiopathic pulmonary fibrosis (IPF): disease pathophysiology, targets, and potential therapeutic interventions. Mol Cell Biochem 2024;479:2181-94. [Crossref] [PubMed]
16. Ahmed A, Tait SWG. Targeting immunogenic cell death in cancer. Mol Oncol 2020;14:2994-3006. [Crossref] [PubMed]
17. Liu SS, Lv XX, Liu C, et al. Targeting Degradation of the Transcription Factor C/EBPβ Reduces Lung Fibrosis by Restoring Activity of the Ubiquitin-Editing Enzyme A20 in Macrophages. Immunity 2019;51:522-534.e7. [Crossref] [PubMed]
18. Zhang Y, Sun J, Li M, et al. Identification and validation of a disulfidptosis-related genes prognostic signature in lung adenocarcinoma. Heliyon 2024;10:e23502. [Crossref] [PubMed]
19. Hutchinson J, Fogarty A, Hubbard R, et al. Global incidence and mortality of idiopathic pulmonary fibrosis: a systematic review. Eur Respir J 2015;46:795-806. [Crossref] [PubMed]
20. Yue YL, Zhang MY, Liu JY, et al. The role of autophagy in idiopathic pulmonary fibrosis: from mechanisms to therapies. Ther Adv Respir Dis 2022;16:17534666221140972. [Crossref] [PubMed]
21. Yang L, Cao LM, Zhang XJ, et al. Targeting ferroptosis as a vulnerability in pulmonary diseases. Cell Death Dis 2022;13:649. [Crossref] [PubMed]
22. Ptasinski VA, Stegmayr J, Belvisi MG, et al. Targeting Alveolar Repair in Idiopathic Pulmonary Fibrosis. Am J Respir Cell Mol Biol 2021;65:347-65. [Crossref] [PubMed]
23. Xie K, Tan X, Chen Z, et al. Exploring the Mechanisms and Preventive Strategies for the Progression from Idiopathic Pulmonary Fibrosis to Lung Cancer: Insights from Transcriptomics and Genetic Factors. Biomedicines 2024;12:2382. [Crossref] [PubMed]
24. Colunga Biancatelli RML, Solopov PA, Catravas JD. The Inflammasome NLR Family Pyrin Domain-Containing Protein 3 (NLRP3) as a Novel Therapeutic Target for Idiopathic Pulmonary Fibrosis. Am J Pathol 2022;192:837-46. [Crossref] [PubMed]
25. Tian R, Zhu Y, Yao J, et al. NLRP3 participates in the regulation of EMT in bleomycin-induced pulmonary fibrosis. Exp Cell Res 2017;357:328-34. [Crossref] [PubMed]
26. Lasithiotaki I, Giannarakis I, Tsitoura E, et al. NLRP3 inflammasome expression in idiopathic pulmonary fibrosis and rheumatoid lung. Eur Respir J 2016;47:910-8. [Crossref] [PubMed]
27. Schwaid AG, Spencer KB. Strategies for Targeting the NLRP3 Inflammasome in the Clinical and Preclinical Space. J Med Chem 2021;64:101-22. [Crossref] [PubMed]
28. Müller T, Fay S, Vieira RP, et al. The purinergic receptor subtype P2Y2 mediates chemotaxis of neutrophils and fibroblasts in fibrotic lung disease. Oncotarget 2017;8:35962-72. [Crossref] [PubMed]
29. Shi J, Zhao Y, Wang Y, et al. Inflammatory caspases are innate immune receptors for intracellular LPS. Nature 2014;514:187-92. [Crossref] [PubMed]
30. Wang J, Sahoo M, Lantier L, et al. Caspase-11-dependent pyroptosis of lung epithelial cells protects from melioidosis while caspase-1 mediates macrophage pyroptosis and production of IL-18. PLoS Pathog 2018;14:e1007105. [Crossref] [PubMed]
31. Wen S, Deng F, Li L, et al. VX-765 ameliorates renal injury and fibrosis in diabetes by regulating caspase-1-mediated pyroptosis and inflammation. J Diabetes Investig 2022;13:22-33. [Crossref] [PubMed]
32. Tao H, Zhao H, Mo A, et al. VX-765 attenuates silica-induced lung inflammatory injury and fibrosis by modulating alveolar macrophages pyroptosis in mice. Ecotoxicol Environ Saf 2023;249:114359. [Crossref] [PubMed]
33. Kim SY, Park S, Cui R, et al. NXC736 Attenuates Radiation-Induced Lung Fibrosis via Regulating NLRP3/IL-1β Signaling Pathway. Int J Mol Sci 2023;24:16265. [Crossref] [PubMed]
34. Xu JF, Washko GR, Nakahira K, et al. Statins and pulmonary fibrosis: the potential role of NLRP3 inflammasome activation. Am J Respir Crit Care Med 2012;185:547-56. [Crossref] [PubMed]
35. She YX, Yu QY, Tang XX. Role of interleukins in the pathogenesis of pulmonary fibrosis. Cell Death Discov 2021;7:52. [Crossref] [PubMed]
36. Martinez JA, King TE Jr, Brown K, et al. Increased expression of the interleukin-10 gene by alveolar macrophages in interstitial lung disease. Am J Physiol 1997;273:L676-83. [Crossref] [PubMed]
37. Bergeron A, Soler P, Kambouchner M, et al. Cytokine profiles in idiopathic pulmonary fibrosis suggest an important role for TGF-beta and IL-10. Eur Respir J 2003;22:69-76. [Crossref] [PubMed]
38. Tsoutsou PG, Gourgoulianis KI, Petinaki E, et al. Cytokine levels in the sera of patients with idiopathic pulmonary fibrosis. Respir Med 2006;100:938-45. [Crossref] [PubMed]
39. Ye H, Pan J, Cai X, et al. IL-10/IL-10 receptor 1 pathway promotes the viability and collagen synthesis of pulmonary fibroblasts originated from interstitial pneumonia tissues. Exp Ther Med 2022;24:518. [Crossref] [PubMed]
40. Kopchick JJ, Basu R, Berryman DE, et al. Covert actions of growth hormone: fibrosis, cardiovascular diseases and cancer. Nat Rev Endocrinol 2022;18:558-73. [Crossref] [PubMed]
41. Kreuter M, Lee JS, Tzouvelekis A, et al. Monocyte Count as a Prognostic Biomarker in Patients with Idiopathic Pulmonary Fibrosis. Am J Respir Crit Care Med 2021;204:74-81. [Crossref] [PubMed]
42. Zhao W, Wang L, Yang J, et al. Endothelial cell-derived MMP19 promotes pulmonary fibrosis by inducing E(nd)MT and monocyte infiltration. Cell Commun Signal 2023;21:56. [Crossref] [PubMed]
43. Mansouri N, Willis GR, Fernandez-Gonzalez A, et al. Mesenchymal stromal cell exosomes prevent and revert experimental pulmonary fibrosis through modulation of monocyte phenotypes. JCI Insight 2019;4:e128060. [Crossref] [PubMed]
44. Salonen J, Kreus M, Lehtonen S, et al. Decline in Mast Cell Density During Diffuse Alveolar Damage in Idiopathic Pulmonary Fibrosis. Inflammation 2022;45:768-79. [Crossref] [PubMed]
45. Monnier J, Zabel BA. Anti-asialo GM1 NK cell depleting antibody does not alter the development of bleomycin induced pulmonary fibrosis. PLoS One 2014;9:e99350. [Crossref] [PubMed]
46. Galati D, De Martino M, Trotta A, et al. Peripheral depletion of NK cells and imbalance of the Treg/Th17 axis in idiopathic pulmonary fibrosis patients. Cytokine 2014;66:119-26. [Crossref] [PubMed]
47. Kishore A, Petrek M. Roles of Macrophage Polarization and Macrophage-Derived miRNAs in Pulmonary Fibrosis. Front Immunol 2021;12:678457. [Crossref] [PubMed]
48. Tang PM, Nikolic-Paterson DJ, Lan HY. Macrophages: versatile players in renal inflammation and fibrosis. Nat Rev Nephrol 2019;15:144-58. [Crossref] [PubMed]