Protosappanin A alleviates atherosclerosis by regulating the MDM2/GPX4 axis-mediated endothelial ferroptosis

Abstract Endothelial ferroptosis is a crucial pathogenic driver of atherosclerosis (AS) progression. Protosappanin A (PTA), a bioactive compound from Caesalpinia sappan L., protects cardiovascular vessels by regulating ferroptosis. However, whether PTA inhibits AS specifically through suppressing endothelial ferroptosis remains unclear. To address this, ApoE⁻/⁻ mice were fed a high‑fat diet for 16 weeks to induce atherosclerosis, and PTA or simvastatin was administered daily from week 4 through week 16 (12 weeks of intervention). PTA markedly reduced aortic plaque area (en face Oil Red O and H&E staining) and improved serum lipid profiles (TG, TC, LDL-C, HDL-C).It also alleviated endothelial injury, evidenced by decreased VCAM‑1 and ICAM‑1 expression. Mechanistically, PTA ameliorated mitochondrial ferroptosis in endothelial cells, lowered intracellular Fe²⁺ and lipid peroxidation (MDA, lipid ROS), and restored the expression of GPX4, xCT, and FTH1. Immunofluorescence confirmed that PTA upregulated GPX4 within the aortic endothelium. To explore the underlying mechanism, we performed RNA‑seq on ox‑LDL‑treated HUVECs and integrated the data with ferroptosis‑related databases. MDM2 was identified as a key target, which was validated by qPCR and Western blotting. Overexpression of MDM2 activated endothelial ferroptosis and reversed the protective effects of PTA both in vitro and in vivo. Collectively, these findings demonstrate that PTA alleviates AS by suppressing endothelial ferroptosis via the MDM2/GPX4 axis. Transcriptome Sequencing HUVECs were assigned to three groups with three biological replicates per group: control, ox-LDL model (80 μg/mL, 24 h), and ox-LDL + PTA treatment. Total RNA was extracted, quality-validated, and processed for strand-specific cDNA library construction, followed by 150-bp paired-end high-throughput sequencing on the Illumina NovaSeq 6000 platform. Raw reads underwent quality control and trimming, then were aligned to the human reference genome (GRCh38/hg38) via HISAT2, with gene-level read counts quantified by featureCounts. Differential expression analysis was performed with DESeq2 through two sequential pairwise comparisons: ox-LDL vs. control, and ox-LDL + PTA vs. ox-LDL. Genes overlapping between the two differentially expressed gene sets were defined as PTA-responsive therapeutic DEGs, with the significance threshold set at |log₂(fold change)| ≥ 1.5 and P < 0.05. Identification of Ferroptosis-Related DEGs and Hub Gene Screening To identify key ferroptosis-associated genes, ferroptosis drivers and suppressors were retrieved from the FerrDb (https://www.zhounan.org/ferrdb/v3/pages/index.html).The intersection of differentially expressed genes (DEGs) and ferroptosis targets was analyzed using Venn diagrams. Specifically, the upregulated DEGs from the Model vs. Control comparison were intersected with ferroptosis drivers, while the downregulated DEGs from the PTA vs. Model comparison were intersected with ferroptosis suppressors. These two intersected gene sets were then merged to form a comprehensive candidate pool of ferroptosis-related DEGs for downstream analysis. The protein-protein interaction (PPI) network of the merged candidate genes was constructed using the STRING database (https://string-db.org/) with a minimum required interaction score of 0.400. The resulting network was visualized and further analyzed using Cytoscape (v3.10.1). To identify the core regulatory nodes (hub genes), the cytoHubba plugin was employed. Four topological analysis algorithms—MNC (Maximum Neighborhood Component), MCC (Maximal Clique Centrality), Degree, and EPC (Edge Percolated Component)—were utilized to calculate the importance of nodes. Finally, the hub genes were determined by taking the intersection of the top-ranking genes identified by all four algorithms.