Related Experiment Video
Updated: Jun 9, 2026
Ferritinophagy: Assessing the Selective Degradation of Iron by Autophagy in Human Fibroblasts
Published on: February 23, 2024
Identification and Validation of Hub Ferroptosis‑Related Genes in Sepsis: An Integrated Bioinformatics and
1Emergency Department, Wuxi No.2 People's Hospital, Jiangnan University Medical Center, Wuxi, 214002, Jiangsu, China.
Introduction:
Sepsis is a life-threatening condition with heterogeneous pathogenesis. This study aimed to identify ferroptosis-related hub genes, construct their regulatory network, and evaluate their potential as biomarkers and therapeutic targets to elucidate the molecular mechanisms underlying sepsis heterogeneity.
Methods:
Transcriptomic data from 1,042 sepsis patients and 42 controls were integrated from public databases (GEO and ArrayExpress). Based on the expression of 47 ferroptosis-related genes, consensus clustering was performed to identify molecular subtypes. Differentially Expressed Genes (DEGs) between subtypes were identified and analyzed by functional enrichment. Key gene modules were identified using Weighted Gene Co-Expression Network Analysis (WGCNA), and hub genes were screened by intersecting WGCNA results with a Protein-Protein Interaction (PPI) network. An in silico-predicted multi-factor (TF-miRNA-mRNA) regulatory network was constructed using the starBase and Harmonizome databases. Key findings were preliminarily validated in a mouse model of E. coli-induced sepsis using quantitative real-time PCR (qRT-PCR).
Results:
Sepsis patients were stratified into two distinct ferroptosis-based subtypes (Cluster 1, n=702; Cluster 2, n=340). A total of 3,608 DEGs were identified, which were enriched in neutrophil activation, ubiquitination, and bacterial infection. WGCNA identified a key sepsis-associated module, leading to the selection of 21 hub genes from the co-expression network. In the septic mouse model, 9 of these genes (including ANK1, HMBS, and SIAH2) were significantly upregulated, and 2 genes (CA1, HBD) were downregulated in septic mice, providing preliminary experimental support for the bioinformatic findings. A comprehensive in silico-predicted regulatory network involving these 21 hub genes, 146 transcription factors, and 70 miRNAs was established as a resource for hypothesis generation.
Discussion:
The identified hub genes and their regulatory network shed light on the role of ferroptosis in sepsis heterogeneity and pathogenesis. Genes such as SIAH2 and HMBS were differentially expressed between surviving and non-surviving patients, warranting further investigation as potential prognostic biomarkers in future studies. The constructed network identifies actionable targets (e.g., the E3 ligase SIAH2 and specific miRNAs) for therapeutic intervention. Limitations include the retrospective nature of the bioinformatic analysis and the need for further experimental validation of mechanistic roles.
Conclusion:
This study delineates a ferroptosis-associated gene signature and regulatory network in sepsis, providing a hypothesis-generating foundation for future research. The identified hub genes and their regulatory interactions warrant further investigation to explore their potential as prognostic biomarkers or therapeutic targets. However, given the lack of significant association with survival outcomes in the current dataset, these findings should be considered preliminary and require validation in independent cohorts with comprehensive clinical annotation before any clinical applications can be considered.
