Skip to main content
NCBI home page
Scientific Reports logoLink to Scientific Reports
. 2026 Apr 3;16:16036. doi: 10.1038/s41598-026-47184-4

Comprehensive analysis of programmed cell death and oridonin target identification in primary lung cancer for prognostic prediction and therapeutic strategies

Xiaoqin Xiang 1,✉, Xiang Yu 2, Jing Li 1, Wei Cao 1, Rongrong Miao 1, Xin Zhao 1,✉
PMCID: PMC13199421  PMID: 41933146

Abstract

This study investigated the involvement of Oridonin-related prognostic genes in programmed cell death (PCD) in primary lung cancer and constructed a corresponding prognostic model. Transcriptomic, mutation, and clinical data from the TCGA database, including the TCGA-LUSC (Lung Squamous Cell Carcinoma) and TCGA-LUAD (Lung Adenocarcinoma) cohorts with 989 primary lung cancer samples, were analyzed. A total of 38 potential Oridonin-related target genes were identified from the PubChem database, among which 15 demonstrated prognostic significance. Weighted Gene Co-expression Network Analysis (WGCNA) revealed correlations between these genes and 13 types of PCD. LASSO and random forest algorithms constructed risk score models and identified FLNC and FOSL1 as independent prognostic biomarkers. Molecular docking analysis confirmed strong binding affinities between FLNC, FOSL1, and Oridonin, suggesting their potential as therapeutic targets. The results showed that the tumor mutation burden (TMB) of the high-risk group was significantly higher than that of the low-risk group. Analysis of immune profiles revealed that the heterogeneity in risk scores was closely linked to distinct patterns of immune cell infiltration and differential activation of immune responses in the tumor microenvironment. Additionally, experimental validation through Western blot demonstrated elevated protein expression of FLNC and FOSL1 in tumor cell lines, and immunohistochemistry (IHC) confirmed their high expression in tumor tissues. The results highlight the critical role of FLNC and FOSL1 as prognostic markers in lung cancer, while also providing new perspectives on the development of Oridonin-based therapeutic interventions.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-47184-4.

Keywords: Oridonin, Primary Lung Cancer, FLNC, FOSL1, Molecular Docking, WGCNA, Immune Microenvironment, Nomogram

Subject terms: Lung cancer, Tumour biomarkers

Background

Globally, lung cancer continues to be the leading cause of both incidence and death among all cancer types1. It is broadly divided into non-small cell lung cancer (NSCLC) and small cell lung cancer (SCLC), with NSCLC accounting for the majority of cases. NSCLC can be subdivided into adenocarcinoma, squamous cell carcinoma, and large cell carcinoma2,3. Patients with lung adenocarcinoma are sensitive to EGFR inhibitors and ALK inhibitors, while patients with lung squamous cell carcinoma are more suitable for PD-L1 immune checkpoint inhibitors4–7. However, molecular targeted therapy is not effective for all lung cancer patients, and some patients may have adverse reactions and are prone to develop drug resistance. Further research and innovation are therefore needed to improve diagnostic accuracy, treatment outcomes, and patient survival.

Oridonin, a diterpenoid derived from the medicinal plant Rabdosia serra (referred to as Donglingcao in traditional Chinese medicine), demonstrates a range of pharmacological effects, including anti-inflammatory, antiviral, antibacterial, detoxifying, and heat-clearing activities8,9. Moreover, Oridonin has demonstrated broad-spectrum antitumor activity by regulating the NF-κB signaling pathway, thereby inhibiting cell proliferation and promoting apoptosis and autophagy in cancer cells10,11. Studies have shown that oridonin affects gasdermin (GSDM) cleavage and thus inhibits pyroptosis by inhibiting caspase-1 and NLRP3, activating caspase-3 and caspase-8, and promoting ROS accumulation12. Oridonin enhances the antitumor activity of RSL3 against breast cancer cells by modulating the JNK-mediated signaling pathway, thereby promoting ferroptosis in tumor cells13. Furthermore, oridonin inhibits the growth, migration, and invasive potential of lung cancer cells through the suppression of lncRNA AFAP1-AS1 and the inactivation of the FAK-ERK1/2 signaling cascade14,15. Studies have also found that oridonin can enhance radiation-induced cytotoxicity and anti-tumor activity, enhancing radiation response by increasing apoptosis16. However, the regulatory mechanisms by which oridonin influences programmed cell death (PCD) in lung cancer remain unclear. Therefore, we reviewed 13 PCD signature molecules and systematically investigated the effect of oridonin on cell death in primary lung cancer.

PCD is a regulated process essential for tissue homeostasis and immune function, distinguishing it from non-programmed necrosis17. Apoptosis, the most common form, maintains cell renewal and immune balance17. Ferroptosis arises from iron accumulation and lipid peroxidation-induced toxicity18,19, while pyroptosis, also iron-dependent, disrupts the cell membrane20,21. Necroptosis shares features of necrosis and apoptosis but is mediated by RIPK1/RIPK3 activation instead of caspases22. Neutrophil extracellular traps (NETs) contribute to necrotic cell death23,24. Entotic cell death involves one cell engulfing another25, whereas lysosome-dependent cell death results from lysosomal enzyme release26. Parthanatos is induced by excessive PARP activation, leading to mitochondrial dysfunction27,28. Autophagy-dependent cell death occurs in response to cellular stress29, and oxeiptosis, triggered by oxidative stress, leads to non-inflammatory cell demise30. Alkaliptosis results from alkaline exposure, causing cellular damage31. Copper-induced cell death involves copper ion accumulation and mitochondrial dysfunction32, while disulfide-induced cell death stems from abnormal disulfide bond formation affecting protein function33. Nevertheless, the mechanisms by which oridonin regulates PCD and influences lung cancer progression are not fully understood, highlighting the need for additional research to elucidate its molecular pathways and clinical applications.

We conducted a comprehensive screening and analysis of multi-omics data from primary lung cancer patients within the TCGA-LUAD and TCGA-LUSC cohorts, categorizing individuals into distinct molecular subtypes according to the expression profiles of genes targeted by Oridonin. The ssGSEA algorithm was utilized to assess the association between 13 PCD biological processes and patient prognosis. Subsequently, WGCNA was employed to identify PCD-related genes regulated by oridonin targets. Based on these key genes, a prognostic model and clinical nomogram were developed to evaluate their impact on patient prognosis and response to immunotherapy. Additionally, molecular docking simulations were performed to assess whether these independent prognostic biomarkers could be targeted by Oridonin in primary lung cancer treatment. Validation studies, utilizing techniques such as Western Blot and immunohistochemistry (IHC), demonstrated alterations or dysregulation of critical genes in both tumor cell lines and tissue samples. As the first comprehensive investigation of PCD in primary lung cancer patients, this research provides a foundational framework for understanding the molecular mechanisms through which oridonin modulates PCD. The results not only provide valuable insights into potential therapeutic targets for precision medicine but also hold promise for advancing the development of novel diagnostic and prognostic biomarkers for primary lung cancer.

Methods

Data acquisition

This study utilized transcriptomic data, mutation profiles, copy number variation (CNV) data, and clinical information from The Cancer Genome Atlas (TCGA) database (https://portal.gdc.cancer.gov/). The analysis targeted patients with primary lung cancer, specifically the TCGA-LUSC (Lung Squamous Cell Carcinoma) and TCGA-LUAD (Lung Adenocarcinoma) cohorts, encompassing 989 samples (501 LUSC and 488 LUAD). Gene expression profiles underwent preprocessing to eliminate batch effects and standardize data using the “sva” package, followed by log2 transformation for normalization.

Identification of oridonin-related prognostic genes

A search in the PubChem database34(Compound CID: 5321010) for “Chemical-Target Interactions” identified 38 potential Oridonin-related target genes. These genes were then analyzed for their prognostic significance in lung cancer using univariate Cox regression analysis.

Genomic analysis, correlation, and PPI network

The genomic locations of the 15 prognostic genes identified from univariate Cox regression were mapped to their respective chromosomes. CNV analysis targets the parts of the chromosome where cnv occurs the most and assesses the frequency of increased copy numbers. The associations among the 15 genes were further investigated, and Pearson correlation analyses were performed to evaluate their potential relevance as risk factors in lung cancer. To investigate potential functional relationships, protein-protein interaction (PPI) networks were constructed using interaction data obtained from the STRING database35(confidence score < 0.4), and the resulting network was visualized using Cytoscape (version 3.9.1).

Consensus clustering

Consensus clustering was performed on the 15 Oridonin-related prognostic genes in primary lung cancer using the ConsensusClusterPlus package in R36. The optimal selection of clusters was determined by evaluating cluster numbers from 2 to 9 (maxK = 9), with 50 iterations (reps = 50) and 80% of samples randomly selected in each iteration (pItem = 0.8). K-means clustering (clusterAlg="km”) with Euclidean distance (distance="euclidean”) was applied. PCA was used to assess cluster distribution, and Kaplan-Meier (KM) analysis evaluated overall survival (OS) differences between clusters. Gene Set Enrichment Analysis (GSEA) was conducted to identify significantly enriched pathways within each cluster, providing insights into the molecular mechanisms driving these differences.

Differential expression and enrichment analysis

The genetic analysis differs from absolute logFC > 0585 and p < 0.05 to identify genes which indicate significant differences. Functional enrichment analysis (GO and KEGG) of these genes was carried out to uncover significant biological processes and pathways.

WGCNA analysis of oridonin-related genes and PCD

WGCNA was used to investigate the association between Oridonin-related genes and 13 distinct types of PCD in primary lung cancer, using differentially expressed genes identified between clusters from consensus clustering. Gene expression data was preprocessed with the limma package in R, merging duplicate genes and log2 transforming the data. Genes with the highest variables were selected for analysis, and quality control removed genes and samples with excessive missing values.

The pickSoftThreshold function is used to determine the optimal soft threshold power for network construction, based on topological index (R²) and average connectivity. A similarity matrix (TOM) was constructed, followed by hierarchical clustering and dynamic tree cutting to identify gene modules. Module eigengenes (MEs) were calculated, and highly correlated modules were merged. The relationship between gene modules and 13 PCD processes was then analyzed.

LASSO and random forest for prognostic biomarker selection

In this study, 974 primary lung cancer patients with complete gene expression profiles were randomly assigned to training or testing groups. The analysis was conducted in two steps: (1) Univariate Cox regression analysis was conducted to identify hub genes within the module that exhibited significant associations with patient prognosis, and (2) feature selection using LASSO and Random Forest algorithms to refine potential prognostic biomarkers.

Gene expression data from prognostic module hub genes were analyzed, with survival status as the dependent variables. For LASSO, the `glmnet` package in R was applied to build a regression model, applying cross-validation (`cv.glmnet`) to select the optimal regularization parameter. Genes with non-zero coefficients were selected as significant. For Random Forest, a model was trained with 500 trees (`ntree = 500`) and a feature subset (`mtry = 5`). Gene importance was evaluated based on the mean decrease in Gini score, and important genes were selected for further analysis. Both algorithms were used to narrow down the set of potential prognostic biomarkers.

Construction of oridonin-related clinical nomogram

A clinical nomogram was constructed by integrating key clinical variables, including tumor stage, radiation therapy, age, and smoking history, as well as the risk score, to estimate overall survival (OS) in patients with primary lung cancer. To assess the independent prognostic significance of these variables, multivariate Cox regression analysis was performed.Survival calibration curves were drawn for patients to compare observed and predicted survival. The predictive performance of the nomogram was validated by calculating the concordance index (C-index), while its discriminative power was evaluated through time-dependent ROC curves at 1-, 3-, and 5-year intervals, with the area under the curve (AUC) determined for each time point.Significant clinical factors were first identified through univariate Cox regression and subsequently incorporated into multivariate analyses to determine independent prognostic factors, with a particular focus on the risk level.

Immune cell composition analysis

To quantify the proportions of immune cell types present in the samples, the CIBERSORT algorithm37 was applied. Gene expression data were examined to quantify different immune cell subsets, followed by an assessment of their relationships with risk scores and model genes.

Immune and stromal scores

The ESTIMATE algorithm was employed to compute immunescores and stromal scores for individual samples, representing the extent of immune cell infiltration and stromal content within the tumor microenvironment, respectively. These scores were then compared between different risk groups to identify potential variations.

Immune response assessment

The TIDE database38(http://tide.dfci.harvard.edu) was employed to evaluate immune evasion and predict sample responsiveness to immunotherapy. Comparative analysis of immune response levels was conducted between risk groups, and TIDE scores were calculated to assess immune escape potential.

Immune checkpoint inhibition response

The Immune Phenotype Score (IPS) was applied to estimate the likelihood of response to immune checkpoint inhibitors39. IPS scores were compared between the two groups for different immune checkpoint treatments to assess potential response to immunotherapy.

Tumor mutation burden (TMB) calculation and correlation with risk score

TMB for primary lung cancer patients was calculated using somatic mutation data from two lung cancer databases, and processed with the “maftools” R package. Patients were stratified into two subgroups according to the risk scores generated from the gene expression-based prognostic model. Differences in tumor mutation burden (TMB) between the subgroups were evaluated using the t-test, and the association between TMB and risk scores was analyzed through Pearson correlation.Boxplots and scatter plots were used for visualization, and waterfall plots depicted mutation landscapes for each risk group.

Molecular docking of oridonin with independent prognostic genes using AutoDock vina

The 3D structure of Oridonin (CID: 5321010) was retrieved from the PubChem database and optimized using Chem3D software. The structure was then optimized and unnecessary molecules were removed. For potential target proteins, the structures were sourced from the Protein Data Bank (PDB), or if experimental structures were unavailable, predicted using AlphaFold. Protein structures were processed to remove water molecules, ions, and ligands, followed by the addition of hydrogen atoms using PyMOL. Molecular docking simulations were performed by preparing protein structures and converting them into PDBQT format. Oridonin was docked to the binding sites of these target proteins using AutoDock Vina, which calculated the binding affinities for various poses. The pose with the lowest binding energy was selected for further analysis.

Western blot

To evaluate the differential expression of FLNC and FOSL1, Western Blot analysis was performed using two cell lines: BEAS-2B, representing normal bronchial epithelial cells, and A549, derived from lung adenocarcinoma. Both cell lines were cultured under standard conditions, and cell proteins were extracted using lysates. Protein samples (30 µg per lane) were resolved using gel electrophoresis and subsequently transferred onto PVDF membranes. The membranes were probed with primary antibodies targeting FLNC (ab clone, A13018) and FOSL1 (Immunoway, YT1772), followed by incubation with corresponding secondary antibodies. Protein bands were visualized using ECL reagents, and the relative expression levels of FLNC and FOSL1 were quantified with ImageJ software.

HE staining and immunohistochemistry

Primary lung cancer tissues and matched adjacent normal tissues from three patients were processed for paraffin embedding, sectioning, and subsequent staining with hematoxylin-eosin (HE) and immunohistochemical (IHC) methods. The staining results were then systematically analyzed. Tissue structure and quality were evaluated by HE staining, and FLNC and FOSL1 protein expressions were detected by IHC. Sections were incubated with primary antibodies for FLNC (Immunoway, YN3834) and FOSL1 (Immunoway, YT1772) for 1 h at room temperature, followed by color development and counterstained with hematoxylin using DAB reagent. Quantitative analysis of protein expression was conducted using ImageJ software to calculate integrated optical density (IOD mean) from five random high-power fields. Statistical differences in FLNC and FOSL1 expression between tumor and adjacent normal tissues were assessed usingWilcoxon signed-rank test.

Statistical analysis

Statistical analysis was performed in R (version 4.3.3). Differential expression was assessed using the “limma” package (logFC > 0.585, P < 0.05). Pearson correlation evaluated gene-risk score relationships. Univariate Cox regression identified survival-associated genes. Prognostic biomarkers were selected via LASSO and Random Forest, with feature selection based on cross-validation and Gini scores. Differences between two risk groups in clinical factors, immune scores, and TMB were analyzed with t-tests, and TMB-risk score correlation was examined using Pearson correlation. A clinical nomogram was built via multivariate Cox regression, and its accuracy was validated with calibration curves, C-index, and ROC curves (1-, 3-, and 5-year). Statistical significance was set at P < 0.05.

Results

Identification of Prognostic Oridonin Target Genes

Through univariate Cox regression analysis applied to 38 Oridonin-related genes’ expression data in lung cancer patients, 15 genes showed statistically significant prognostic value (P < 0.05). Chromosomal mapping revealed that MYC, CCND1, and BIRC5 exhibited the most frequent copy number amplifications, predominantly localized to chromosomes 8, 11, and 18 (Fig. 1A and B). Pearson correlation analysis revealed that most of these genes were significantly positively correlated with each other and acted as risk factors, with hazard ratios greater than 1 (HR > 1, P < 0.05) (Fig. 1C). Furthermore, PPI network analysis demonstrated that the proteins encoded by these genes interact with each other, forming a network centered around key nodes such as CCND1, SRC, and MYC (Fig. 1D). The results demonstrate that these 15 genes hold notable therapeutic promise as Oridonin-responsive targets for lung cancer management. Their strong association with prognosis suggests that Oridonin may modulate these genes to exert therapeutic effects, potentially enhancing treatment efficacy and patient outcomes.

Fig. 1.

Fig. 1

Open in a new tab

Analysis of 15 Oridonin-related prognostic genes in primary lung cancer. (A) Chromosomal locations of the 15 genes. (B) Copy number variations (CNV) of the genes, with MYC, CCND1, and BIRC5 showing the highest gains. (C) Pearson correlation analysis indicating positive correlations with most genes being risk factors. (D) Protein-protein interaction (PPI) network, highlighting key genes like CCND1, SRC, and MYC.

Molecular characterization and prognostic implications of oridonin-related genes

Application of consensus clustering to 15 Oridonin-associated prognostic genes within primary lung tumors revealed two molecular subgroups, categorized as subgroup A and B. The optimal stability of the clustering was determined at k = 2, as indicated by the cumulative distribution function (CDF) plot (Fig. 2A, B). Kaplan-Meier analysis revealed significantly better OS in cluster A compared to cluster B (Fig. 2C, P = 0.030, HR = 1.253 (95% CI: 1.022–1.536)). Principal component analysis (PCA) validation demonstrated distinct segregation between molecular subgroups (previously labeled clusters), correlating with unique transcriptional profiles (Fig. 2D). Comparative analysis revealed upregulated expression levels of most Oridonin-associated prognostic genes in subgroup B versus subgroup A, with statistical significance confirmed (Fig. 2E, P < 0.05).

Fig. 2.

Fig. 2

Open in a new tab

Consensus clustering and survival analysis of Oridonin-related prognostic genes. (A) CDF curve determining cluster stability maxima at k = 2. (B) Cluster-defining heatmap (k = 2) visualizing molecular subgroup segregation. Expression values are normalized to z-scores. (C) Stratified survival curves demonstrating cluster A/B prognostic divergence (log-rank P = 0.030). (D) PCA dimensionality reduction mapping cluster-specific transcriptomic profiles. (E) Oridonin-target gene expression dichotomy across molecular subgroups. Expression values are log2-transformed (FPKM + 1) and normalized. (F-H) Enrichment topology: GSEA pathway activation (F), GO biological processes (G), and KEGG oncogenic networks (H) across clusters. Normalized enrichment scores (NES) are shown for each pathway.

GSEA identified multiple KEGG pathways with significant enrichment in cluster A relative to cluster B, including alpha-linolenic acid metabolism, arachidonic acid metabolism, cytochrome P450-mediated drug metabolism, and linoleic acid metabolism (Fig. 2F, P < 0.05). The findings indicate that cluster A is strongly linked to specific metabolic processes, particularly those involved in lipid metabolism and drug metabolism pathways. Pathway enrichment observed in cluster A suggests a characteristic metabolic profile with potential implications for cancer progression dynamics, therapeutic efficacy optimization, and metabolism-targeted treatment strategy development in lung malignancies.

Differential expression analysis, with a threshold of absolute logFC > 0.585 and P < 0.05, revealed a total of 727 genes that were differentially expressed between clusters A and B. GO analysis of these genes showed significant enrichment in biological processes such as chromosomal region, chromosome condensation, and mitotic nuclear division (Fig. 2G), while KEGG pathway analysis revealed significant associations with pathways like small cell lung cancer, cell cycle, and DNA replication (Fig. 2H). The data indicate that inter-cluster variations are associated with cell cycle control and mitotic progression, potentially modulating apoptotic pathways and alternative cell death modalities in pulmonary malignancies.

Selection and analysis of the blue module for programmed cell death

To explore the Oridonin-PCD interaction in lung cancer pathogenesis, WGCNA was performed using 727 cluster A/B differential genes. The dataset underwent preprocessing and clustering to maintain quality, with no significant outliers observed. Genes were classified into distinct modules according to TOM similarity, with the primary groups identified as blue, brown, gray, and turquoise. These modules were further analyzed to explore their potential associations with PCD processes and Oridonin’s effects in lung cancer (Fig. 3A). The optimal soft-thresholding power for network construction was identified as 4, with an R² value of 0.885 (Fig. 3B). The eigengenes (MEs) for each module were calculated, and hierarchical clustering showed relationships between the modules. Highly correlated modules were merged into the final blue, brown, grey, and turquoise modules. Among these, the blue module demonstrated a significant correlation with various types of PCD, particularly apoptosis (cor = 0.52) and autophagy-dependent cell death (cor = 0.52) (Fig. 3C). This module initially contained 167 genes, and after applying filters for module membership (MM ≥ 0.4) and gene significance (GS for apoptosis ≥ 0.2), 116 genes were selected (Fig. 3D). To refine the list further, univariate Cox regression analysis was performed, identifying 83 genes with significant associations to prognosis. These 83 prioritized genes underwent functional annotation based on their predicted involvement in Oridonin-triggered PCD within primary pulmonary malignancies, demonstrating dual potential as druggable targets and prognostic biomarkers.

Fig. 3.

Fig. 3

Open in a new tab

WGCNA of Differentially Expressed Genes between Two Clusters. (A) Gene dendrogram and module colors, showing the clustering of genes into blue, brown, grey, and turquoise modules. (B) Soft-thresholding power selection, with the optimal power of 4 (R² = 0.885) for network construction. (C) Correlation between gene modules and programmed cell death (PCD) types. (D) Module membership vs. gene significance for apoptosis.

Prognostic biomarker identification using LASSO and random forest

A cohort of 974 primary lung cancer patients was randomly allocated into training (487) and testing (487) sets. Within the 116 hub genes of the blue module, univariate Cox regression analysis identified 83 with significant prognostic relevance (P < 0.05). These genes were selected for further investigation due to their potential relevance in predicting outcomes in primary lung cancer.

In the training group, the normalized gene expression matrix of these 83 genes served as independent variables, while OS.status was considered the dependent variable. The LASSO method, utilizing 10-fold cross-validation, determined that the model achieved the minimum binomial deviance (mean square error) when five variables were included (Fig. 4A and B). This resulted in the identification of four genes: ITGA6, DCBLD2, FOSL1, FLNC, and PCDH7.

Fig. 4.

Fig. 4

Open in a new tab

Gene Selection and Prognostic Model Construction. (A) Cross-validation plot for LASSO showing optimal lambda selection. (B) Selection of five variables with minimum binomial deviance. (C) Model error plot from Random Forest with 294 trees. (D) Gene importance from Random Forest based on Gini score. (E) Venn diagram showing intersecting genes (FLNC, FOSL1) from LASSO and Random Forest. (F) Forest plot of multivariate Cox regression. (G) Model coefficients for FOSL1 and FLNC. (H) Heatmap showing that the expression levels of FLNC and FOSL1 increase with higher risk in patients.

To further narrow down the list of prognostic biomarkers, Random Forest (RF) analysis was performed. The model was trained with 500 trees (ntree = 500) and a feature subset of five variables (mtry = 5). The optimal number of trees in the RF algorithm was determined to be 294, as it resulted in the lowest error rate (Fig. 4C). Gene importance was assessed by evaluating the mean decrease in the Gini score, and genes with a Gini score greater than 7 were selected for further analysis, highlighting their potential significance in prognosis prediction. The RF analysis identified five key genes: CYP24A1, MYO1E, EREG, FLNC, and FOSL1 (Fig. 4D).

The intersection between the LASSO and RF methods identified FLNC and FOSL1 as common key genes (Fig. 4E). These genes were further validated through multivariate Cox regression analysis, confirming them as independent prognostic factors (P < 0.001 for FLNC and P = 0.042 for FOSL1). Both FLNC and FOSL1 were associated with poor prognosis, with higher expression correlating with worse survival outcomes. Specifically, for FLNC, the HR = 1.227 (95% CI: 1.104–1.365) and for FOSL1, the HR = 1.111 (95% CI: 1.004–1.229) (Fig. 4F).To evaluate patient risk, a prognostic risk score was computed using the following formula:

Inline graphic

Based on training cohort-derived median risk thresholds, subjects were classified into high/low-risk subgroups. Risk stratification heatmaps revealed marked upregulation of FLNC and FOSL1 in high-risk versus low-risk patient cohorts (Fig. 4H). These findings imply that elevated expression of FLNC and FOSL1 could act as reliable prognostic indicators for unfavorable outcomes in primary lung cancer patients. Subsequent analysis further validated the robustness of these findings in predicting patient prognosis.

Development and validation of a clinical nomogram

We constructed a predictive nomogram combining clinical parameters with prognostic risk stratification to estimate survival probabilities in lung cancer cases (Fig. 5A). Calibration plots at 1-/3-/5-year intervals (Fig. 5B) showed strong alignment of model-predicted survival probabilities with actual observations, validating predictive consistency. The nomogram achieved higher C-index values than standalone clinical variables or risk scores (Fig. 5C), demonstrating improved prognostic precision. Time-dependent ROC analysis showed the nomogram’s AUCs of 0.689 (1-year), 0.682 (3-year), and 0.656 (5-year), outperforming comparator clinical predictors (Fig. 5D). DCA analysis at 1/3/5 years revealed the nomogram’s superior clinical net benefit across threshold probabilities compared to individual clinical parameters (Fig. 5E). Univariate screening identified Pathologic_M/N/T stage, tumor staging, and Risk_score as significant prognostic determinants (Fig. 5F, P < 0.05), while multivariate validation established Risk_score as an independent predictor with HR > 1 (Fig. 5G, P < 0.05). This integrated model demonstrates clinical utility for personalized outcome prediction in lung oncology practice.

Fig. 5.

Fig. 5

Open in a new tab

Development and Validation of a Clinical Nomogram for Integrated Prognostication. (A) Nomogram incorporating 1-, 3-, and 5-year overall survival (OS) probabilities. (B) Calibration curves validating prediction-actual survival concordance at 1/3/5 years. (C) Diagnostic accuracy comparison: Nomogram vs. individual clinical/risk predictors. (D) Time-stratified ROC curves assessing 1-/3-/5-year mortality prediction. (E) Decision curve analysis (DCA) for 1-/3-/5-year clinical utility evaluation. (F) Univariate Cox regression of survival-associated clinicopathological variables (P < 0.05). (G) Multivariate confirmation of nomogram-derived risk stratification’s independence.

Immune microenvironment profiling via multi-algorithm integration

CIBERSORT deconvolution showed significant positive associations linking risk stratification to M0 macrophage infiltration and resting NK cell abundance (Fig. 6A, cor > 0, P < 0.05), while inversely correlating with follicular helper T-cell and CD8 + T-lymphocyte populations (Fig. 6A, cor < 0, P < 0.05). ESTIMATE algorithm quantification identified elevated stromal component scores in high-risk patients versus low-risk counterparts (Fig. 6B, P < 0.05), though immune microenvironment scoring showed inter-group equivalence. TIDE computational modeling revealed the low-risk cohort (41%) displayed enhanced anti-tumor immune reactivity compared to high-risk subjects (19%) (Fig. 6C, P < 0.05). Concurrently, elevated TIDE prediction indices in high-risk patients implied heightened immune evasion capacity (Fig. 6D, P < 0.05). IPS profiling demonstrated high-risk patients demonstrated elevated IPS_ctla4_neg_pd1_neg index values, predictive of resistance to CTLA-4/PD-1 dual blockade therapies (Fig. 6E). Contrastingly, reduced IPS_ctla4_pos_pd1_neg scores in this subgroup indicated impaired CTLA-4 inhibitor responsiveness under PD-1 suppression (Fig. 6F). These immunogenomic patterns delineate risk-specific checkpoint inhibitor vulnerabilities, informing precision immunotherapy approaches.

Fig. 6.

Fig. 6

Open in a new tab

Immune Landscape Analysis. (A) Correlation between risk score and immune cell types. (B) Comparison of stromal scores between two groups. (C) Comparison of immune response levels between two groups based on TIDE (Tumor Immune Dysfunction and Exclusion). (D) TIDE scores comparison between two groups. (E) Comparison of IPS_CTLA4_neg_PD1_neg (immune response in the absence of CTLA-4 and PD-1 inhibitors) between two groups. (F) Comparison of IPS_CTLA4_pos_PD1_neg (immune response with CTLA-4 inhibition but without PD-1 inhibition) between two groups.

Mutation landscape and its association with risk scores

Risk-stratified analysis revealed substantially elevated TMB in high-risk patients versus low-risk counterparts (Fig. 7A, P < 0.05). Bivariate correlation analysis demonstrated significant positive associations between risk stratification indices and TMB quantification (Fig. 7B, R = 0.11, P = 0.0022), with Cluster B preferentially distributed in high TMB/high-risk coordinate zones, suggesting synergistic prognostic detriment. Mutation spectrum visualization via waterfall plots identified increased mutational load in the high-risk cohort (95.03%) compared to low-risk subjects (92.84%) (Figs. 7C-D). Recurrently mutated oncogenes including TP53, TTN, and MUC16 exhibited amplified alteration frequencies in high-risk patients, dominated by missense mutation subtypes. This mutational landscape implies mechanistic involvement in lung carcinogenesis and prognostic deterioration within high-risk demographics.

Fig. 7.

Fig. 7

Open in a new tab

Tumor Mutation Landscape and Correlation with Risk Score. (A) Risk-stratified TMB (tumor mutation burden) quantification demonstrating elevated mutational burden in high-risk cohort (P < 0.05). (B) Scatter plot depicting the positive correlation between TMB and risk score, indicating that higher TMB is associated with higher risk scores. (C) Mutation spectrum visualization (high-risk subjects) showing 95.03% alteration frequency. (D) Mutation spectrum visualization (low-risk subjects) showing 92.84% alteration frequency.

Molecular docking of oridonin with FLNC and FOSL1

The molecular docking results revealed significant binding interactions between Oridonin and both FLNC and FOSL1. The structure of FOSL1 was predicted using AlphaFold (AF), while the structure of FLNC was obtained from the Protein Data Bank (PDB). For FLNC, the best docking pose showed a binding affinity of -6.7 kcal/mol, with additional poses ranging from − 6.5 kcal/mol to -5.1 kcal/mol. Similarly, the optimal docking pose for FOSL1 demonstrated a binding affinity of -6.0 kcal/mol, with other poses exhibiting binding energies between − 5.6 kcal/mol and − 4.8 kcal/mol. The binding sites corresponding to the lowest binding energy poses were visualized using PyMOL 2.5.5, providing insight into the molecular interactions between these proteins and potential therapeutic implications (Fig. 8). These results suggest that Oridonin forms stable binding interactions with both FLNC and FOSL1, with favorable docking poses observed for both proteins.

Fig. 8.

Fig. 8

Open in a new tab

Molecular Docking Sites of FLNC and FOSL1 with Oridonin. (A) Docking site of FLNC with Oridonin at the lowest binding energy. (B) Docking site of FOSL1 with Oridonin at the lowest binding energy.

Differential expression of FLNC and FOSL1 in A549 and BEAS-2B cell lines

Western blot densitometry demonstrated differential expression profiles of FLNC and FOSL1 across cell models. As depicted in Fig. 9, both FLNC and FOSL1 showed significantly higher expression in A549 cells (lung adenocarcinoma) compared to BEAS-2B cells (normal bronchial epithelial cells). Quantitative analysis further confirmed that the protein levels of FLNC and FOSL1 in A549 cells were significantly increased (P < 0.05). The expression level of FLNC in A549 cells was approximately 2.17 times that in BEAS-2B cells, and the expression level of FOSL1 in A549 cells was approximately 2.80 times that in BEAS-2B cells. This indicates that compared with the normal control, these proteins were abnormally upregulated in the tumor cell line.

Fig. 9.

Fig. 9

Open in a new tab

Protein level of FLNC and FOSL1 in BEAS-2B and A549 cell lines.

Western blot analysis showing the protein levels of FLNC and FOSL1 in both BEAS-2B and A549 cells. Quantitative analysis reveals that the expression of FLNC and FOSL1.

Differential expression of FLNC and FOSL1 in tumor versus adjacent normal tissues

HE staining revealed distinct histological differences between tumor tissues and adjacent normal tissues, with tumor samples displaying characteristic disorganized architecture and increased nuclear atypia (Fig. 10). IHC analysis further demonstrated a significant increase in the expression levels of FLNC and FOSL1 in tumor tissues compared to adjacent normal tissues. Quantitative evaluation using integrated optical density (IOD mean) confirmed that the expression of FLNC was upregulated by approximately 2.05-fold and FOSL1 by approximately 1.89-fold in tumor samples relative to normal tissues (Fig. 10, P < 0.05), highlighting their potential as prognostic biomarkers in primary lung cancer.

Fig. 10.

Fig. 10

Open in a new tab

HE staining and IHC analysis of FLNC and FOSL1 in primary lung cancer and adjacent normal tissues. IHC shows significantly higher FLNC and FOSL1 expression in tumor tissues, supported by IOD (mean) quantification. Scale bar = 60 μm. P < 0.05.

Discussion

In this study, we identified FLNC and FOSL1 as independent prognostic biomarkers in primary lung cancer using a comprehensive multi-omics approach that integrated transcriptomic, mutation, and clinical data from the LUSC and LUAD cohorts. The analysis aimed to identify Oridonin-related prognostic genes and construct a model for patient stratification and therapeutic targeting. Both FLNC and FOSL1 expression levels were strongly associated with unfavorable survival outcomes. Their participation in multiple PCD pathways, such as apoptosis and autophagy-mediated cell death, underscores their pivotal contributions to lung cancer progression. Experimental validation further confirmed their involvement in PCD-related pathways, reinforcing their potential as key biomarkers for prognosis and therapeutic intervention in lung cancer.

Molecular subtyping and the role of oridonin-related genes

Molecular subtyping based on Oridonin target genes offered profound insights into the heterogeneity of lung cancer. We identified distinct patient subgroups with varying survival outcomes by analyzing the expression patterns of FOSL1 and FLNC. Genes regulated by Oridonin were identified as key regulators of biological pathways linked to cancer development, indicating their promise as targets for therapeutic intervention. FOSL1, a member of the Fos family of transcription factors, has been implicated in promoting tumor cell proliferation and survival40, while FLNC, which plays a key role in cytoskeletal remodeling, is essential for cell migration and metastasis41,42. These findings suggest that Oridonin-regulated genes could not only serve as biomarkers for prognosis but also offer therapeutic targets for precision medicine in lung cancer.

The impact of programmed cell death in lung cancer prognosis

This research further elucidates the significance of PCD mechanisms in predicting lung cancer outcomes, offering novel perspectives on their prognostic relevance. Previous studies have linked different PCD mechanisms, including apoptosis, ferroptosis, and necroptosis43–45, to various survival outcomes in primary lung cancer, underscoring the importance of PCD in regulating tumor progression and patient survival. Therefore, to explore how Oridonin influences the onset and progression of primary lung cancer through specific PCD pathways, we performed ssGSEA and WGCNA analyses to examine the close relationship between Oridonin target genes, downstream regulatory genes, and PCD. Additionally, we identified key genes, such as FOSL1 and FLNC, that significantly impact PCD pathways, particularly apoptosis and autophagy-dependent cell death. Furthermore, the dysregulation of cell cycle processes, such as chromosome separation and mitosis, observed in lung cancer cells, may contribute to defects in PCD pathways. This disruption could further promote tumor progression and negatively impact therapeutic responses46.

Mechanistic insights into FOSL1 and FLNC in lung cancer

FOSL1 and FLNC are essential in the pathogenesis of lung cancer, significantly influencing critical processes such as immune modulation, metastasis, and tumor proliferation. FOSL1 is involved in regulating genes related to epithelial-mesenchymal transition (EMT) and immune response modulation, both critical for lung cancer progression41,42. FOSL1 also plays a central role in promoting tumor survival and cisplatin resistance by regulating DNA repair and apoptosis, thus contributing to therapy resistance in lung cancer cells47. Additionally, recent studies have shown that FOSL1 can drive the expression of genes such as amphiregulin, which are crucial for cell proliferation and survival in KRAS-driven lung adenocarcinoma, further emphasizing its importance in tumorigenesis48.

On the other hand, FLNC is involved in regulating cell motility and maintaining cytoskeletal architecture, both of which are essential for the metastatic spread of lung cancer cells41. FLNC’s interaction with the cytoskeleton enables cell shape changes necessary for tumor cell invasion and migration, processes vital for metastasis. Moreover, FLNC has been implicated in regulating autophagy and mitophagy in muscle cells, highlighting its broader role in cell survival under stress conditions. However, its precise involvement in autophagy regulation in lung cancer remains to be fully elucidated49,50.

These roles suggest that FOSL1 and FLNC may not only influence tumor biology but also affect immune responses within the tumor microenvironment (TME), potentially making them key players in determining patient responses to immunotherapy. FLNC’s dysregulation in various cancers correlates with poor prognosis, further indicating its importance in the tumor immune microenvironment and as a potential therapeutic target for immune modulation51. Moreover, FOSL1 has been shown to influence immune cell dynamics within the TME, particularly in its role in regulating immune checkpoints, which may contribute to immune escape mechanisms in tumors47. This suggests that FOSL1 plays a crucial role in immune escape mechanisms by regulating immune checkpoint expression, which may facilitate tumor progression and resistance to immunotherapy.

Risk stratification and immunotherapy implications

Our findings suggest that integrating FLNC and FOSL1 into a prognostic model offers a robust tool for predicting clinical outcomes in lung cancer patients. High-risk patients, characterized by elevated expression of these genes, were associated with poorer prognosis, thus enabling more precise risk stratification. Furthermore, our molecular docking analysis revealed that both FLNC and FOSL1 exhibit strong binding affinities with Oridonin, a promising anti-cancer agent, suggesting that Oridonin may directly modulate the function of these proteins48,52. These findings provide a mechanistic basis for integrating Oridonin with complementary treatment strategies, including immune checkpoint inhibitors, particularly in high-risk lung cancer patients.

Analysis of mutation profiles revealed that patients in the high-risk subgroup exhibited a markedly elevated TMB, a recognized biomarker associated with improved responses to immunotherapy53. This increased TMB implies its potential as an additional biomarker for guiding the application of immune checkpoint inhibitors in personalized treatment strategies. Moreover, the correlation between risk scores and the immune cell landscape in the (TME) implies that modulating immune activity may play a critical role in improving therapeutic outcomes for high-risk individuals, particularly those with dysregulated immune function.

Comparison with previous studies

This study builds on existing research by providing the first comprehensive analysis of Oridonin-regulated PCD genes in primary lung cancer. Although earlier research has highlighted the importance of PCD in tumor development46,54, our study extends this work by identifying FOSL1 and FLNC as critical factors in lung cancer prognosis and Oridonin’s potential as a therapeutic agent. We performed experimental validation through Western Blot and IHC to confirm the significant upregulation of these genes in tumor tissues. Our findings indicate that the PCD processes regulated by these genes may also contribute to immune evasion, a key factor in tumor progression and treatment resistance, thus offering a distinct perspective compared to previous studies.

Limitations and future directions

While this study provides valuable insights into the roles of FOSL1 and FLNC in lung cancer, several limitations should be considered. First, the retrospective nature of the analysis introduces potential biases that may impact the generalizability of our findings. Additionally, the relatively small sample size, particularly in the IHC and Western Blot experiments, limits the statistical power and the broader applicability of the conclusions. To enhance the robustness of future studies, it is crucial to include a larger sample size and expand the range of lung cancer cell lines examined. This should include more than two cell lines, incorporating various histological subtypes, such as squamous cell carcinoma and large cell carcinoma, as well as normal lung tissue as negative controls.

Furthermore, although we observed promising interactions between FLNC, FOSL1, and Oridonin, the binding of these proteins with Oridonin requires further validation. In this study, fluorescent antibody assays were not performed; however, we plan to employ techniques such as co-immunoprecipitation or immunofluorescence in future experiments to confirm these interactions.

Lastly, while our analysis was based on publicly available data from the TCGA database, the lack of validation with an independent cohort of human lung carcinoma samples limits the external validity of our findings. Future research should aim to validate these results using real patient cohorts, which would strengthen the clinical relevance and confirm the reproducibility of the observed relationships in diverse clinical settings.

Conclusion

In conclusion, our study identifies FLNC and FOSL1 as independent prognostic biomarkers in primary lung cancer, providing valuable insights for precision medicine and potential immunotherapy strategies. The significant expression of these two potential biomarkers in tumors was validated at the protein level in both cell and patient tissue samples. By integrating molecular subtyping with PCD-related biomarkers, a prognostic scoring model was established to forecast clinical outcomes and enhance diagnostic and therapeutic approaches. Molecular docking further supports Oridonin’s potential as a therapeutic agent targeting FLNC and FOSL1. Furthermore, our examination of the mutation landscape and immune response uncovers distinct immune signatures and TMB variations across risk groups, emphasizing the significance of risk stratification in forecasting immune modulation and genetic instability. These findings underscore Oridonin’s therapeutic potential, particularly in combination with immunotherapies, although further clinical validation is necessary.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary Material 1 (517.5KB, pdf)
Supplementary Material 2 (899.4KB, pdf)

Author contributions

Xiaoqin Xiang: Conceptualization, Data curation, Project administration, Writing–review and editing. Xiang Yu and Jing Li: Data curation, Formal Analysis, Methodology.Wei Cao: Data curation, Software, Visualization, Writing–original draft, Writing–review and editing. Rongrong Miao: Data curation, Software, Visualization, Writing–original draft, Writing–review and editing. Xin Zhao: Conceptualization, Data curation, Project administration, Writing–review and editing.

Funding

This work was supported by the Scientific Research Project of Hunan Provincial Health Commission (Development of Oridonin Modified by TPGS and Its Efficacy in Primary Lung Cancer in Rats, Grant No. D2023 13028318).

Data availability

The data used in this study were sourced from publicly available databases. Transcriptomic, mutation, and clinical data for the TCGA-LUSC (Lung Squamous Cell Carcinoma) and TCGA-LUAD (Lung Adenocarcinoma) cohorts were accessed from The Cancer Genome Atlas (TCGA) database ( [https://portal.gdc.cancer.gov/](https:/portal.gdc.cancer.gov) ). The potential Oridonin-related target genes were identified from the PubChem database ( [https://pubchem.ncbi.nlm.nih.gov/](https:/pubchem.ncbi.nlm.nih.gov) ). Protein-protein interaction data were obtained from the STRING database ( [https://string-db.org/](https:/string-db.org) ). The protein structures for FOSL1 were predicted using AlphaFold ( [https://alphafold.ebi.ac.uk/](https:/alphafold.ebi.ac.uk) ). Tumor Immune Dysfunction and Exclusion (TIDE) scores were calculated using the TIDE database ( [http://tide.dfci.harvard.edu/](http:/tide.dfci.harvard.edu) ), and Immunophenoscore (IPS) data were accessed from the IPS database ( [https://www.immport.org](https:/www.immport.org) ). These resources provide comprehensive and publicly available data that were utilized for the analysis and validation in this study. All datasets used are available in the respective public repositories.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

This study was conducted in accordance with the ethical principles outlined in the Declaration of Helsinki. The research was reviewed and approved by the Ethics Committee of Hunan Children’s Hospital, with the approval reference number HCHLL-2025-38. All participants (or their legal guardians) provided informed consent prior to inclusion in the study. The study adhered to all relevant guidelines and regulations to ensure ethical compliance in all aspects of the research.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Contributor Information

Xiaoqin Xiang, Email: xiangxq2024@sina.com.

Xin Zhao, Email: zhaox202303@163.com.

References

  • 1.Siegel, R. L., Miller, K. D., Wagle, N. S. & Jemal, A. Cancer statistics, 2023. CA Cancer J. Clin.73, 17–48. 10.3322/caac.21763 (2023). [DOI] [PubMed] [Google Scholar]
  • 2.Li, Q. et al. A new hope: the immunotherapy in small cell lung cancer. Neoplasma63, 342–350. 10.4149/302_151001N511 (2016). [DOI] [PubMed] [Google Scholar]
  • 3.Travis, W. D. et al. The 2015 World Health Organization Classification of Lung Tumors: Impact of Genetic, Clinical and Radiologic Advances Since the 2004 Classification. J. Thorac. Oncol.10, 1243–1260. 10.1097/JTO.0000000000000630 (2015). [DOI] [PubMed] [Google Scholar]
  • 4.Thai, A. A., Solomon, B. J., Sequist, L. V., Gainor, J. F. & Heist, R. S. Lung cancer. Lancet398, 535–554. 10.1016/S0140-6736(21)00312-3 (2021). [DOI] [PubMed] [Google Scholar]
  • 5.Tan, A. C. & Tan, D. S. W. Targeted Therapies for Lung Cancer Patients With Oncogenic Driver Molecular Alterations. J. Clin. Oncol.40, 611–625. 10.1200/JCO.21.01626 (2022). [DOI] [PubMed] [Google Scholar]
  • 6.Duma, N., Santana-Davila, R. & Molina, J. R. Non-Small Cell Lung Cancer: Epidemiology, Screening, Diagnosis, and Treatment. Mayo Clin. Proc.94, 1623–1640. 10.1016/j.mayocp.2019.01.013 (2019). [DOI] [PubMed] [Google Scholar]
  • 7.Gridelli, C. et al. Non-small-cell lung cancer. Nat. Rev. Dis. Primers. 1, 15009. 10.1038/nrdp.2015.9 (2015). [DOI] [PubMed] [Google Scholar]
  • 8.Vasaturo, M. et al. The anti-tumor diterpene oridonin is a direct inhibitor of Nucleolin in cancer cells. Sci. Rep.8, 16735. 10.1038/s41598-018-35088-x (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Xu, S. et al. Synthesis and antimycobacterial evaluation of natural oridonin and its enmein-type derivatives. Fitoterapia99, 300–306. 10.1016/j.fitote.2014.10.005 (2014). [DOI] [PubMed] [Google Scholar]
  • 10.Li, C. Y., Wang, E. Q., Cheng, Y. & Bao, J. K. Oridonin: An active diterpenoid targeting cell cycle arrest, apoptotic and autophagic pathways for cancer therapeutics. Int. J. Biochem. Cell. Biol.43, 701–704. 10.1016/j.biocel.2011.01.020 (2011). [DOI] [PubMed] [Google Scholar]
  • 11.Liu, Q. Q. et al. Oridonin derivative ameliorates experimental colitis by inhibiting activated T-cells and translocation of nuclear factor-kappa B. J. Dig. Dis.17, 104–112. 10.1111/1751-2980.12314 (2016). [DOI] [PubMed] [Google Scholar]
  • 12.Li, G. Q. et al. Anticancer mechanisms on pyroptosis induced by Oridonin: New potential targeted therapeutic strategies. Biomed. Pharmacother. 165, 115019. 10.1016/j.biopha.2023.115019 (2023). [DOI] [PubMed] [Google Scholar]
  • 13.Ye, S. et al. Oridonin promotes RSL3-induced ferroptosis in breast cancer cells by regulating the oxidative stress signaling pathway JNK/Nrf2/HO-1. Eur. J. Pharmacol.974, 176620. 10.1016/j.ejphar.2024.176620 (2024). [DOI] [PubMed] [Google Scholar]
  • 14.Chen, M. & Shen, Y. Oridonin represses lung cancer cell proliferation and migration by modulating AFAP1-AS1/IGF2BP1. Cell. Mol. Biol. (Noisy-le-grand). 70, 108–113. 10.14715/cmb/2024.70.6.17 (2024). [DOI] [PubMed] [Google Scholar]
  • 15.Xu, L. et al. Oridonin inhibits the migration and epithelial-to-mesenchymal transition of small cell lung cancer cells by suppressing FAK-ERK1/2 signalling pathway. J. Cell. Mol. Med.24, 4480–4493. 10.1111/jcmm.15106 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Park, H., Jeong, Y. J., Han, N. K., Kim, J. S. & Lee, H. J. Oridonin Enhances Radiation-Induced Cell Death by Promoting DNA Damage in Non-Small Cell Lung Cancer Cells. Int. J. Mol. Sci. (2018). 10.3390/ijms19082378 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Renehan, A. G., Booth, C. & Potten, C. S. What is apoptosis, and why is it important? BMJ. 322, 1536–1538 , (2001). 10.1136/bmj.322.7301.1536 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Jiang, X., Stockwell, B. R. & Conrad, M. Ferroptosis: mechanisms, biology and role in disease. Nat. Rev. Mol. Cell. Biol.22, 266–282. 10.1038/s41580-020-00324-8 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Li, J. et al. Ferroptosis: past, present and future. Cell. Death Dis.11, 88. 10.1038/s41419-020-2298-2 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Yu, P. et al. Pyroptosis: mechanisms and diseases. Signal. Transduct. Target. Ther.6, 128. 10.1038/s41392-021-00507-5 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Rao, Z. et al. Pyroptosis in inflammatory diseases and cancer. Theranostics12, 4310–4329. 10.7150/thno.71086 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.Yan, J., Wan, P., Choksi, S. & Liu, Z. G. Necroptosis and tumor progression. Trends Cancer. 8, 21–27. 10.1016/j.trecan.2021.09.003 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Fuchs, T. A. et al. Novel cell death program leads to neutrophil extracellular traps. J. Cell. Biol.176, 231–241. 10.1083/jcb.200606027 (2007). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Brinkmann, V. et al. Neutrophil extracellular traps kill bacteria. Science303, 1532–1535. 10.1126/science.1092385 (2004). [DOI] [PubMed] [Google Scholar]
  • 25.Hamann, J. C., Kim, S. E. & Overholtzer, M. Methods for the Study of Entotic Cell Death. Methods Mol. Biol.1880, 447–454. 10.1007/978-1-4939-8873-0_28 (2019). [DOI] [PubMed] [Google Scholar]
  • 26.Nakamura, H. et al. Lysosome-Associated Membrane Protein 3 Induces Lysosome-Dependent Cell Death by Impairing Autophagic Caspase 8 Degradation in the Salivary Glands of Individuals With Sjogren’s Disease. Arthritis Rheumatol.75, 1586–1598. 10.1002/art.42540 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Zheng, D. et al. ROS-triggered endothelial cell death mechanisms: Focus on pyroptosis, parthanatos, and ferroptosis. Front. Immunol.13, 1039241. 10.3389/fimmu.2022.1039241 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Robinson, N. et al. Programmed necrotic cell death of macrophages: Focus on pyroptosis, necroptosis, and parthanatos. Redox Biol.26, 101239. 10.1016/j.redox.2019.101239 (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Denton, D. & Kumar, S. Autophagy-dependent cell death. Cell. Death Differ.26, 605–616. 10.1038/s41418-018-0252-y (2019). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Holze, C. et al. Oxeiptosis, a ROS-induced caspase-independent apoptosis-like cell-death pathway. Nat. Immunol.19, 130–140. 10.1038/s41590-017-0013-y (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 31.Chen, F., Kang, R., Liu, J. & Tang, D. Mechanisms of alkaliptosis. Front. Cell. Dev. Biol.11, 1213995. 10.3389/fcell.2023.1213995 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 32.Kahlson, M. A. & Dixon, S. J. Copper-induced cell death. Science375, 1231–1232. 10.1126/science.abo3959 (2022). [DOI] [PubMed] [Google Scholar]
  • 33.Altonsy, M. O., Habib, T. N. & Andrews, S. C. Diallyl disulfide-induced apoptosis in a breast-cancer cell line (MCF-7) may be caused by inhibition of histone deacetylation. Nutr. Cancer. 64, 1251–1260. 10.1080/01635581.2012.721156 (2012). [DOI] [PubMed] [Google Scholar]
  • 34.Kim, S. et al. PubChem in 2021: new data content and improved web interfaces. Nucleic Acids Res.49, D1388–D1395. 10.1093/nar/gkaa971 (2021). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 35.Szklarczyk, D. et al. The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest. Nucleic Acids Res.51, D638–D646. 10.1093/nar/gkac1000 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 36.Wilkerson, M. D. & Hayes, D. N. ConsensusClusterPlus: a class discovery tool with confidence assessments and item tracking. Bioinformatics26, 1572–1573. 10.1093/bioinformatics/btq170 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 37.Newman, A. M. et al. Robust enumeration of cell subsets from tissue expression profiles. Nat. Methods. 12, 453–457. 10.1038/nmeth.3337 (2015). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 38.Fu, J. et al. Large-scale public data reuse to model immunotherapy response and resistance. Genome Med.12, 21. 10.1186/s13073-020-0721-z (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 39.Charoentong, P. et al. Pan-cancer Immunogenomic Analyses Reveal Genotype-Immunophenotype Relationships and Predictors of Response to Checkpoint Blockade. Cell. Rep.18, 248–262. 10.1016/j.celrep.2016.12.019 (2017). [DOI] [PubMed] [Google Scholar]
  • 40.Vallejo, A., Valencia, K. & Vicent, S. All for one and FOSL1 for all: FOSL1 at the crossroads of lung and pancreatic cancer driven by mutant KRAS. Mol. Cell. Oncol.4, e1314239. 10.1080/23723556.2017.1314239 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 41.Luo, X. et al. Melatonin inhibits EMT and PD-L1 expression through the ERK1/2/FOSL1 pathway and regulates anti-tumor immunity in HNSCC. Cancer Sci.113, 2232–2245. 10.1111/cas.15338 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 42.Hu, Y. Y. et al. Jinfeng pills ameliorate premature ovarian insufficiency induced by cyclophosphamide in rats and correlate to modulating IL-17A/IL-6 axis and MEK/ERK signals. J. Ethnopharmacol.307, 116242. 10.1016/j.jep.2023.116242 (2023). [DOI] [PubMed] [Google Scholar]
  • 43.Yuan, R. et al. Cucurbitacin B inhibits non-small cell lung cancer in vivo and in vitro by triggering TLR4/NLRP3/GSDMD-dependent pyroptosis. Pharmacol. Res.170, 105748. 10.1016/j.phrs.2021.105748 (2021). [DOI] [PubMed] [Google Scholar]
  • 44.Wang, L. et al. Mechanisms of PANoptosis and relevant small-molecule compounds for fighting diseases. Cell. Death Dis.14, 851. 10.1038/s41419-023-06370-2 (2023). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 45.Chen, S., Jiang, J., Li, T., Huang, L. & PANoptosis Mechanism and Role in Pulmonary Diseases. Int. J. Mol. Sci. (2023). 10.3390/ijms242015343 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 46.Ocansey, D. K. W. et al. Current evidence and therapeutic implication of PANoptosis in cancer. Theranostics14, 640–661. 10.7150/thno.91814 (2024). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 47.Tong, W. et al. Transcription Factor FOSL1 Promotes Cisplatin Resistance in Non-Small Cell Lung Cancer Cells by Modulating the Wnt3a/beta-Catenin Signaling through Upregulation of PLIN3 Expression. Front. Biosci. (Landmark Ed). 30, 26898. 10.31083/FBL26898 (2025). [DOI] [PubMed] [Google Scholar]
  • 48.Elangovan, I. M. et al. FOSL1 Promotes Kras-induced Lung Cancer through Amphiregulin and Cell Survival Gene Regulation. Am. J. Respir Cell. Mol. Biol.58, 625–635. 10.1165/rcmb.2017-0164OC (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 49.Han, S. et al. Filamin C regulates skeletal muscle atrophy by stabilizing dishevelled-2 to inhibit autophagy and mitophagy. Mol. Ther. Nucleic Acids. 27, 147–164. 10.1016/j.omtn.2021.11.022 (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 50.Goliusova, D. V. et al. Role of Filamin C in Muscle Cells. Biochem. (Mosc). 89, 1546–1557. 10.1134/S0006297924090025 (2024). [DOI] [PubMed] [Google Scholar]
  • 51.Ni, P. et al. Unveiling the multifaceted role of the FLNC gene: implications for cancer diagnosis and prognosis. Eur. J. Med. Res.30, 608. 10.1186/s40001-025-02876-x (2025). [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 52.Roman, M. et al. Inhibitor of Differentiation-1 Sustains Mutant KRAS-Driven Progression, Maintenance, and Metastasis of Lung Adenocarcinoma via Regulation of a FOSL1 Network. Cancer Res.79, 625–638. 10.1158/0008-5472.CAN-18-1479 (2019). [DOI] [PubMed] [Google Scholar]
  • 53.Wang, D. et al. LCN2 secreted by tissue-infiltrating neutrophils induces the ferroptosis and wasting of adipose and muscle tissues in lung cancer cachexia. J. Hematol. Oncol. (2023). 10.1186/s13045-023-01429-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 54.Peng, F. et al. Regulated cell death (RCD) in cancer: key pathways and targeted therapies. Signal. Transduct. Target. Ther.7, 286. 10.1038/s41392-022-01110-y (2022). [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary Material 1 (517.5KB, pdf)
Supplementary Material 2 (899.4KB, pdf)

Data Availability Statement

The data used in this study were sourced from publicly available databases. Transcriptomic, mutation, and clinical data for the TCGA-LUSC (Lung Squamous Cell Carcinoma) and TCGA-LUAD (Lung Adenocarcinoma) cohorts were accessed from The Cancer Genome Atlas (TCGA) database ( [https://portal.gdc.cancer.gov/](https:/portal.gdc.cancer.gov) ). The potential Oridonin-related target genes were identified from the PubChem database ( [https://pubchem.ncbi.nlm.nih.gov/](https:/pubchem.ncbi.nlm.nih.gov) ). Protein-protein interaction data were obtained from the STRING database ( [https://string-db.org/](https:/string-db.org) ). The protein structures for FOSL1 were predicted using AlphaFold ( [https://alphafold.ebi.ac.uk/](https:/alphafold.ebi.ac.uk) ). Tumor Immune Dysfunction and Exclusion (TIDE) scores were calculated using the TIDE database ( [http://tide.dfci.harvard.edu/](http:/tide.dfci.harvard.edu) ), and Immunophenoscore (IPS) data were accessed from the IPS database ( [https://www.immport.org](https:/www.immport.org) ). These resources provide comprehensive and publicly available data that were utilized for the analysis and validation in this study. All datasets used are available in the respective public repositories.

Articles from Scientific Reports are provided here courtesy of Nature Publishing Group

RESOURCES