Unveiling Ferroptosis-Driven Mechanisms in Liver Fibrosis: A Bioinformatics Approach to Identify Novel Therapeutic Targets

By Chao Wang

Affiliations

1. Department of Hepatobiliary and Pancreatic Surgery, Affiliated Hangzhou First People’s Hospital, West Lake University School of Medicine, Hangzhou, China

doi: 10.29271/jcpsp.2026.08.1004

ABSTRACT
Objective: To explore the potential biological significance and diagnostic value of ferroptosis-related differentially expressed genes (DEGs) in liver fibrosis (LF).
Study Design: Bioinformatic analysis of a publicly available microarray dataset.
Place and Duration of the Study: Department of Hepatobiliary and Pancreatic Surgery, Affiliated Hangzhou First People’s Hospital, West Lake University School of Medicine, Hangzhou, China, from June 2 to July 2, 2024.
Methodology: Gene expression data from GSE139602 and ferroptosis-related genes from the GeneCards database were analysed to identify ferroptosis-related DEGs in LF. Functional enrichment, gene set enrichment, immune infiltration, protein-protein interaction, and ROC analyses were performed to identify key genes and their potential roles in LF.
Results: Forty-one DEGs were associated with ferroptosis, and LF were identified. Among these, CEBPA, MYC, SCD, and SREBF1 were highlighted as important genes with significant diagnostic ability, confirmed through the receiver operating characteristic curve. Notably, CEBPA and SCD were implicated in fibrosis progression via their roles in lipid metabolism and oxidative stress modulation, which are essential components of ferroptosis.
Conclusion: The bioinformatics analysis identified four key ferroptosis-related genes—CEBPA, MYC, SCD, and SREBF1—as diagnostic biomarkers for LF. This study found the ferroptosis-related pathways as potential therapeutic targets for LF.

Key Words: Biomedicine, Liver fibrosis, Ferroptosis, Differentially expressed genes.

INTRODUCTION

Liver fibrosis (LF) is a condition caused by excessive downfall of extracellular matrix (ECM)  proteins, primarily collagen, as a consequence of liver damage. This progressive process can lead to cirrhosis and liver cancer, representing a major global health concern.^1,2 Hepatic injury activates the hepatic stellate cells, which differentiate into myofibroblast-like cells, resulting in ECM production and ultimately fibrosis. Therefore, knowing the mole- cular pathways of LF is important for the development of targeted therapy.

Ferroptosis is induced by iron-mediated lipid peroxidation.³ This process is marked by the buildup of harmful lipid-based reactive oxygen species (ROS), which overwhelm and disable the antioxidant system that depends on glutathione. It leads to a lot of pathology, including neurodegenerative diseases, ischaemia- reperfusion injury, and carcinoma.⁴

Recent research found that ferroptosis may be connected to LF, with iron overload and oxidative stress playing an important role.⁵ The connection between ferroptosis and LF is getting attention because both processes involve similar mediators such as ROS, lipid peroxidation, and iron meta-bolism pathways.⁶

Given the clinical outcomes associated with LF and ferrop- tosis, the mechanisms underlying this relationship need to be explored. Bioinformatics can be used to analyse existing data-sets to identify ferroptosis-related differentially expressed genes (DEGs) in LF and to perform functional enrichment and pathway analyses of ferroptosis-associated pathways. Identifying the mechanism through this study may lead to new therapeutic strategies for LF.

Therefore, the present study aimed to identify ferroptosis- related DEGs in liver fibrosis using the GSE139602 dataset and the GeneCards database and explore their potential bio- logical functions, immune associations, and diagnostic value through integrated bioinformatics analyses.

METHODOLOGY

This bioinformatics study was conducted at the Department of Hepatobiliary and Pancreatic Surgery, Affiliated Hangzhou First People’s Hospital, West Lake University School of Medicine, Hangzhou, China, from June 2 to July 2, 2024.

Patient data for LF were obtained from the NCBI Gene Exp- ression Omnibus (GEO, accession number: GSE139602), utilising the GPL13667 platform (Affymetrix Human Genome U219 Array). The original liver biopsy samples were collected and profiled by investigators at an external centre in Barcelona, Spain. This dataset included five LF patients, eight compensated cirrhosis patients, twelve decompensated cirrhosis patients, and six healthy controls. For this analysis, data from the five fibrosis patients and six healthy controls were specifically selected.⁷ Additionally, 1366 ferroptosis-related genes were found from GeneCards by using ferroptosis as the search keyword.⁸

The DESeq2 package was used to perform DEG analysis of the count data derived from GSE139602.⁹ Patient data for liver fibrosis were obtained from the NCBI Gene Expression Omnibus (GEO, accession number: GSE139602), based on the GPL13667 platform. This dataset includes five liver fibrosis patients, eight compensated cirrhosis patients, twelve decompensated cirrhosis patients, and six healthy controls. For the present study, only the five liver fibrosis samples and six healthy control samples were included in the main comparative analysis. Samples from the compensated cirrhosis and decompensated cirrhosis groups were excluded because they were not aligned with the predefined com- parison of liver fibrosis versus healthy liver tissue. The selection criteria for DEGs were a logFC >1 and an adjusted p-value (P.adj) <0.05. Specifically, genes with a logFC >1 and P.adj <0.05 were classified as upregulated DEGs, whereas genes with a logFC <-1 and P.adj <0.05 were classified as downregulated DEGs.

All DEGs were initially filtered for the criteria of logFC >1 and P.adj <0.05. To identify ferroptosis-associated genes implicated in LF, differential analyses comparing data from GeneCards and GSE139602 were used.⁸ These DEGs were intersected with ferroptosis-related genes, producing another Venn diagram to highlight the common genes. Volcano plots of the DEGs were generated using the ggplot2 package, and heatmaps were generated using the pheatmap package in R.

Functional annotation analysis was performed using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways. The GO analysis contained three parts: cellular components (CC), biological processes (BP), and molecular functions (MF). ClusterProfiler R package was used to perform those analyses, applying significance thresholds of p <0.05 and an FDR (q-value) <0.05.

Gene set enrichment analysis (GSEA) was used to assess the distribution of genes.¹⁰ Ferroptosis-related genes were intersected with DEGs from the GSE139602 and divided into two groups. The GSEA parameters were set as shown below: seed number 2020, 1000 permutations, with each gene set containing a minimum of 10 genes and a maximum of 500 genes.

The single-sample Gene Set Enrichment Analysis (ssGSEA) algorithm was used to assess immune cell infiltrates, inclu-ding dendritic cells, CD8+ T cells, macrophages, and T cell sub-types.¹¹ The CIBERSORT algorithm was used for deeper-level immune infiltration analysis. The intersected matrix of ferroptosis-related genes and DEGs from GSE139602 was input into CIBERSORT, utilising the LM22 signature gene matrix. An immune cell enrichment score exceeding zero was selected for further analysis.

A PPI network of ferroptosis-related DEGs was built by using the STRING database (version 12.0).¹² The network was visualised by Cytoscape (version 3.10.2), applying a confidence score threshold of >0.9.¹³ The StarBase database was used to identify miRNA targets.¹⁴ This resource provides extensive data on RNA-binding proteins (RBPs), including interactions such as RBP-ncRNA and RBP-mRNA. Furthermore, the miRNA target prediction database, miRDB, was used to predict RBP target genes.¹⁵ Finally, the Drug-Gene Interaction Database (DGIdb) was used to find potential drugs or small molecules which may interact with ferroptosis-associated DEGs.¹⁶

ROC curve analysis was performed to evaluate the diagnostic performance of candidate ferroptosis-related DEGs in distinguishing liver fibrosis samples from healthy controls. The optimal cut-off value for each gene was determined based on the maximum Youden index. The four genes with the highest AUROC values were selected as representative diagnostic genes, namely CEBPA, MYC, SCD, and SREBF1. The Human Protein Atlas (HPA) database was used for single-cell analysis of the genes in relation to ferroptosis- related genes and DEGs of the GSE139602 dataset.¹⁷ The results were then visualised accordingly.

RESULTS

A volcano plot was generated to see the variance results from the GSE139602 dataset (Figure 1A). To identify ferroptosis-related DEGs, all DEGs from the GSE139602 dataset were intersected, reaching the standard of |logFC| >1 and P.adj <0.05. An R package was used to make a heat map displaying the differential expression of these DEGs (Figure 1B). By intersecting the DEGs and PRGs, a total of 41 ferroptosis-related DEGs were identified, and this intersection was depicted using a Venn diagram (Figure 1C). It showed that the 41 intersecting genes were enriched in BPs such as response to peptides, extracellular stimuli, and nutrient levels (Figure 1D), CCs such as cell leading edge, actin-based cell projections, and filopodia (Figure 1E), and MFs related to DNA-binding transcription activator activity, particularly RNA polymerase II-specific functions (Figure 1F). The KEGG pathway analysis suggested that these genes may be involved in similar transcriptional misregulation in cancer, alcoholic liver disease, and the AMPK signalling pathway (Figure 1G).

Figure 1: Analysis of differential genes, GO, KEGG and GSEA analysis in LF. (A) Volcano plot of differential genes in LF tissue and the healthy liver tissue of GSE139602. (B) Heatmaps of ferroptosis-related DEGs in GSE139602. (C) Common DEGs Venn diagram and ferroptosis-related genes in GSE139602. (D) Biological process analysis bubble chart (E), Cellular component analysis bubble chart. (F) Molecular function analysis bubble chart. (G) KEGG pathway analysis bubble chart. (H) The GSEA analysis of the intersection of genes related to ferroptosis and DEGs from dataset GSE139602 mainly includes 5 main biological characteristics. (I) Reactome cytokine signalling in the immune system. (J) Reactome nuclear receptors metapathway. (K) The Reactome antiviral mechanism by IFN-stimulated genes. (L) Reactome interferon signalling. (M) Reactome interferon alpha/beta signalling. Figure 2: PPI, mRNA-RBP, and mRNA-drugs interaction network, tumour microenvironment assessment of the crossing of genes and ROC curves of the intersection. (A) PPI of the intersection of genes related to ferroptosis and DEGs from dataset GSE139602. (B) The mRNA-RBP of the intersection of genes related to ferroptosis and DEGs from dataset GSE139602. The red box represents mRNA; the blue box represents RBP. (C) mRNA-drugs of the intersection of genes related to ferroptosis and DEGs from dataset GSE139602. (D) The immune infiltration showed intersecting genes linked to ferroptosis and DEGs from dataset GSE139602 using the ssGSEA algorithm. (E) The immune infiltration showed the intersection of genes related to ferroptosis and DEGs from dataset GSE139602 using the CIBERSORT algorithm. (F) Differential expression analysis of the intersecting genes linked to ferroptosis and DEGs from dataset GSE139602. (G–J) ROC curves of the ferroptosis-related intersecting DEGs identified from the GSE139602 dataset: CEBPA (G), MYC (H), SCD (I), and SREBF1 (J). The red box represents mRNA; the blue box represents drugs or molecular compounds. The symbol * represents p ≤0.05; the symbol ** represents p ≤0.01. An AUC >0.90 indicates excellent accuracy. Figure 3: Analysis of expression distribution and single-gene analysis of the intersection of genes related to ferroptosis and DEGs from dataset GSE139602. (A) Protein and CEBPA protein and mRNA expression in human body tissues, and CEBPA single-gene analysis in liver tissues. (B) MYC protein and mRNA expression in human body tissues, and single-gene analysis in liver tissues. (C) SCD protein and mRNA expression in human body tissues, and single-gene analysis in liver tissues. (D) SREBF1 protein and mRNA expression in human body tissues, and single-gene analysis in liver tissues.

The x-axis represents -log (P.adj), the y-axis represents GO and KEGG pathway terms, and the bubble chart colour indicates the elevation or downgrade of these terms. The GSEA analysis was performed to find the association between the intersecting genes and DEGs from the GSE139602 dataset (Figure 1H). The DEGs showed pathways such as Reactome cytokine signalling in the immune system (Figure 1I), Reactome nuclear receptor metapathway (Figure 1J), Reactome antiviral mechanism by IFN-stimulated genes (Figure 1K), Reactome interferon signalling (Figure 1L), and Reactome interferon alpha/beta signalling (Figure 1M).

The protein–protein interaction (PPI) network of ferroptosis- related DEGs identified from the GSE139602 dataset was constructed and visualised using Cytoscape (Figure 2A). Out of these, only 10 DEGs were connected to other genes. Interactions of these 10 DEGs with RBPs using mRNA-RBP data were predicted, and these interactions were visualised using Cytoscape (Figure 2B). Additionally, small-molecule com- pounds and potential drugs targeting these 10 mRNAs were identified through the DGIdb database, resulting in an mRNA-drug interaction network (Figure 2C). To investigate immune cell infiltration, the ssGSEA and CIBERSORT algorithms were applied to calculate differences across groups. The ssGSEA revealed significant enrichment of four immune cell types—dendritic cells, neutrophils, CD8⁺ T cells, and central memory T (Tcm) cells—among the ferroptosis-related intersecting DEGs identified from the GSE139602 dataset (Figure 2D). Meanwhile, CIBERSORT analysis highlighted significant neutrophil infiltration (Figure 2E). The DEGs in the intersection of ferroptosis-related genes and GSE139602 revealed significant differences in the expression of 21 diagnostic DEGs (Figure 2F). Among them, CEBPA, MYC, SCD, and SREBF1 showed the highest AUROC values and were therefore selected for presentation. ROC curves were generated for four genes—CEBPA (AUC = 1.000, Figure 2G), MYC (AUC = 1.000, Figure 2H), SCD (AUC = 0.900, Figure 2I), and SREBF1 (AUC = 0.933, Figure 2J)—demonstrating their strong correlation with LF occurrence.

Finally, the RNA and protein expression profiles of the diagnostic DEGs—CEBPA, MYC, SCD, and SREBF1—were evaluated using the HPA database, with a focus on their expression in human liver tissues. The results indicated significant upregulation of CEBPA and SCD, with high distribution in liver tissue. Furthermore, the expression of these diagnostic DEGs was found to be most significantly correlated with c-12 hepatocytes in liver tissue (Figure 3A-D).

DISCUSSION

Bioinformatics analysis was performed to find the importance of ferroptosis-related genes in LF. One of the innovative findings is the possible molecular mechanism of ferroptosis in LF. Furthermore, the study found the potential of targeting specific ferroptosis-related genes to predict and treat LF. The analysis identified 41 ferroptosis-related DEGs that were significantly associated with LF. Among these, CEBPA, MYC, SCD, and SREBF1 demonstrated strong diagnostic potential, as ROC analysis confirmed their high predictive performance, suggesting that these genes are closely associated with LF.

CEBPA, MYC, SCD, and SREBF1 are key genes in LF with their important roles in ferroptosis. CEBPA, a transcription factor for hepatic differentiation, has been associated with LF due to its regulation of lipid metabolism and oxidative stress, which are crucial processes in ferroptosis.^18,19 Regulation of CEBPA activity may provide a novel therapeutic strategy for the management of ferroptosis-related LF. SCD, an enzyme for fatty acid metabolism, promotes ferroptosis-related lipid peroxidation in LF and exacerbates tissue damage and progression of LF.²⁰ SCD can control the balance of saturated and monounsaturated fatty acids to regulate lipid peroxidation for the progression of LF. It makes SCD an important therapeutic gene for LF. MYC has been found in various liver diseases and is also involved in ferroptosis. It influences the progression of LF through cellular metabolism and the regulation of stress response.²¹ It can also regulate iron metabolism to promote ferroptosis, increasing the susceptibility of iron- induced cell death. SREBF1 has been found to be important in both ferroptosis and LF, indicating that SREBF1 could be a valuable therapeutic target by modulating lipid metabolism and ferroptotic processes. A recent study suggests that SREBF1 can promote ferroptosis by influencing lipid meta-bolism.²² These findings confirm the potential advantages of these genes in the diagnosis and treatment of LF, such as through the relevant molecular pathways.

However, this study has some limitations. First, this research is based on bioinformatics analyses, which require experimental validation of the identified genes and pathways through molecular cytology experiments. Future studies should include both in vitro and in vivo experiments. Second, the patient data used in this study were relatively small. Future research should incorporate larger, more diverse datasets to validate these results across more populations.

CONCLUSION

This study offers an innovative vision into the molecular mechanisms linking ferroptosis and LF, especially CEBPA, MYC, SCD, and SREBF1, which are considered diagnostic markers and therapeutic targets. These findings give new opportunities for further research into ferroptosis with LF.

FUNDING:
This research was supported by the Zhejiang Medical and Health Science and Technology Project (No.2022RC058) and the Basic Public Welfare Research Plan of Zhejiang Province (No. LQ24H030009).

COMPETING INTEREST:
The author declared no conflict of interest.

AUTHOR’S CONTRIBUTIONS:
CW: Conceived the study, performed the bioinformatics analysis, explained the results, drafted the manuscript and approved the final version of the manuscript to be published.

REFERENCES

1. Friedman SL. Liver fibrosis--from bench to bedside. J Hepatol 2003; 38 (Suppl 1):S38-53. doi: 10.1016/S0168-8278 (02)00429-4.
2. Bataller R, Brenner DA. Liver fibrosis. J Clin Invest 2005; 115 (2):209-18. doi: 10.1172/JCI24282.
3. Dixon SJ, Lemberg KM, Lamprecht MR, Skouta R, Zaitsev EM, Gleason CE, et al. Ferroptosis: An iron-dependent form of nonapoptotic cell death. Cell 2012; 149(5):1060-72. doi: 10. 1016/j.cell.2012.03.042.
4. Tang D, Kang R, Vanden Berghe T, Vandenabeele P, Kroemer G. The molecular machinery of regulated cell death. Cell Res 2019; 29(5):347-64. doi: 10.1038/s41422- 019-0164-5.
5. Yuan X, Li L, Zhang Y, Ai R, Li D, Dou Y, et al. Heme oxygenase 1 alleviates nonalcoholic steatohepatitis by sup- pressing hepatic ferroptosis. Lipids Health Dis 2023; 22(1):99. doi: 10.1186/s12944-023-01855-7.
6. Chen J, Li X, Ge C, Min J, Wang F. The multifaceted role of ferroptosis in liver disease. Cell Death Differ 2022; 29(3): 467-80. doi: 10.1038/s41418-022-00941-0.
7. Graupera I, Isus L, Coll M, Pose E, Diaz A, Vallverdu J, et al. Molecular characterization of chronic liver disease dynamics: From liver fibrosis to acute-on-chronic liver failure. JHEP Rep 2022; 4(6):100482. doi: 10.1016/j.jhepr.2022.100482.
8. Stelzer G, Rosen N, Plaschkes I, Zimmerman S, Twik M, Fishilevich S, et al. The genecards suite: From gene data mining to disease genome sequence analyses. Curr Protoc Bioinformatics 2016; 54:1.30.1-33. doi: 10.1002/cpbi.5.
9. Love MI, Huber W, Anders S. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol 2014; 15(12):550. doi: 10.1186/s13059-014- 0550-8.
10. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA, et al. Gene set enrichment analysis: A knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci U S A 2005; 102(43): 15545-50. doi: 10.1073/pnas.0506580102.
    
     
11. Charoentong P, Finotello F, Angelova M, Mayer C, Efremova M, Rieder D, et al. Pan-cancer immunogenomic analyses reveal genotype-immunophenotype relationships and predictors of response to checkpoint blockade. Cell Rep 2017; 18(1):248-62. doi: 10.1016/j.celrep.2016.12.019.
12. Szklarczyk D, Gable AL, Nastou KC, Lyon D, Kirsch R, Pyysalo S, et al. The STRING database in 2021: Customizable protein-protein networks, and functional characterisation of user-uploaded gene/measurement sets. Nucleic Acids Res 2021; 49(D1):D605-12. doi: 10.1093/nar/gkaa1074.
13. Shannon P, Markiel A, Ozier O, Baliga NS, Wang JT, Ramage D, et al. Cytoscape: A software environment for integrated models of biomolecular interaction networks. Genome Res 2003; 13(11):2498-504. doi: 10.1101/gr.1239303.
14. Li JH, Liu S, Zhou H, Qu LH, Yang JH. starBase v2.0: Decoding miRNA-ceRNA, miRNA-ncRNA and protein-RNA interaction networks from large-scale CLIP-Seq data. Nucleic Acids Res 2014; 42(Database issue):D92-7. doi: 10.1093/nar/gkt 1248.
15. Chen Y, Wang X. miRDB: An online database for prediction of functional microRNA targets. Nucleic Acids Res 2020; 48(D1):D127-31. doi: 10.1093/nar/gkz757.
16. Freshour SL, Kiwala S, Cotto KC, Coffman AC, McMichael JF, Song JJ, et al. Integration of the Drug-Gene Interaction Database (DGIdb 4.0) with open crowdsource efforts. Nucleic Acids Res 2021; 49(D1):D1144-51. doi: 10.1093/ nar/gkaa1084.
17. Thul PJ, Lindskog C. The human protein atlas: A spatial map of the human proteome. Protein Sci 2018; 27(1):233-44. doi: 10.1002/pro.3307.
18. Yan T, Yan N, Xia Y, Sawaswong V, Zhu X, Dias HB, et al. Hepatocyte-specific CCAAT/enhancer binding protein α restricts liver fibrosis progression. J Clin Invest 2024; 134(7): e166731. doi: 10.1172/JCI166731.
19. Jiang X, Stockwell BR, Conrad M. Ferroptosis: Mechanisms, biology and role in disease. Nat Rev Mol Cell Biol 2021; 22(4):266-82. doi: 10.1038/s41580-020-00324-8.
20. Tracz-Gaszewska Z, Dobrzyn P. Stearoyl-CoA desaturase 1 as a therapeutic target for the treatment of cancer. Cancers (Basel) 2019; 11(7):948. doi: 10.3390/cancers11070948.
21. Dang CV. MYC on the path to cancer. Cell 2012; 149(1): 22-35. doi: 10.1016/j.cell.2012.03.003.
22. Eberle D, Hegarty B, Bossard P, Ferre P, Foufelle F. SREBP transcription factors: Master regulators of lipid homeostasis. Biochimie 2004; 86(11):839-48. doi: 10.1016/j.biochi.2004. 09.018.