Identification and validation of STEAP3 as a ferroptosis-related biomarker in heart failure.

Abstract

Heart failure (HF) is a major health threat, with aging and programmed cell death (PCD) playing key roles. However, the link between aging-related PCD (aging-PCD) genes and HF remains unclear. We used ssGSEA and random forest to analyze PCD types in HF, identified aging-related PCDs via correlation analysis, and applied eight machine learning algorithms to develop a diagnostic model. SHAP and LIME were used to explain key features, and the relationship between the aging-PCD index and immune microenvironment was explored. Gene expression was verified by qRT-PCR, and the function of STEAP3 was further investigated in an H 2 O 2 -induced AC16 cell model. Ferroptosis-related changes were assessed by measuring reactive oxygen species (ROS), malondialdehyde (MDA), glutathione peroxidase (GSH-Px), and ferrous ion (Fe 2+ ) levels. Ferroptosis, autophagy, and necroptosis were strongly correlated with aging in HF. Eighteen differentially expressed aging-related PCD genes were identified. The LASSO model showed the best diagnostic performance. Significant differences in the immune microenvironment were observed between the aging-PCD index-high and index-low groups. A regulatory network of 15 key genes and 19 transcription factors was constructed. The qRT-PCR results validated the bioinformatics analysis. Functional experiments further suggested that STEAP3 may participate in ferroptosis-related injury in cardiomyocytes, potentially through glutathione metabolism (GPX4/SLC7A11 axis) and iron homeostasis. We identified aging-related PCD signatures in HF that may provide candidate biomarkers for further clinical validation.