Development of a ferroptosis-related gene prognostic model and molecular subgroups characterization in sepsis

Abstract

Sepsis, a common and life-threatening condition often leading to multiple organ dysfunction, currently lacks a prognostic model based on ferroptosis-related genes (FRGs) for predicting clinical outcomes. In this study, we utilized the FerrDb database and GSE65682 dataset to evaluate the prognostic significance of FRGs in sepsis. Differential expression analysis identified 27 DE-FRGs, and Consensus clustering revealed three distinct FRG molecular subtypes in sepsis with notable differences in immune infiltration landscapes. Univariate and multivariate Cox regression, along with LASSO analysis, were employed to construct an FRG-based prognostic model, which indicated significantly better clinical outcomes for the low FRG score subgroup compared to the high FRG score subgroup. Validation through nomogram prediction models and independent prognostic analysis confirmed the accuracy of FRGs in assessing sepsis prognosis. Single-cell sequencing further demonstrated the distribution of the FRG prognostic signature across cellular subpopulations in sepsis samples. Functional experiments, including siRNA transfection, malondialdehyde (MDA) assays, Western blot, and reactive oxygen species (ROS) assays, revealed that TFRC plays a critical role in sepsis by inhibiting ferroptosis. These findings suggest that the FRG prognostic scoring model is a reliable predictor of sepsis prognosis, with TFRC identified as a key regulatory factor inhibiting ferroptosis in sepsis.

Keywords: Ferroptosis; Immune infiltration landscape; Molecular subgroup; Sepsis; TFRC.