Targeting the PHB2–ACSL3 Lipid-Remodeling Axis Overcomes Cisplatin Resistance by Restoring Ferroptosis in Gastric Cancer

Platinum-based chemotherapy resistance remains a major obstacle in gastric cancer (GC) treatment. Through integrated transcriptomic profiling of cisplatin-resistant xenografts and pharmacogenomic interrogation of the NCI-60 dataset, we identified Prohibitin-2 (PHB2) as a previously unrecognized determinant of chemoresistance. PHB2 was consistently upregulated in resistant tumors and associated with poor clinical outcomes. Mechanistically, PHB2 binds the lipid-metabolic enzyme ACSL3 through a defined AMP-binding–domain interface (residues W244/H254/E260), enhancing ACSL3 activity to promote monounsaturated fatty acid incorporation into phospholipids. This phospholipid remodeling suppresses lipid peroxidation and establishes a ferroptosis-resistant state. Structure-based virtual screening of an FDA-approved drug library nominated the CXCR4 antagonist Mavorixafor as a potent inhibitor of the PHB2–ACSL3 interaction. Mavorixafor disrupted lipid remodeling, restored ferroptosis susceptibility, and resensitized cisplatin-resistant GC in cell line–derived xenografts (CDOs) and patient-derived organoids (PDOs). The therapeutic effect was abrogated by the ferroptosis inhibitor Liproxstatin-1, confirming ferroptosis dependence. Collectively, our findings define a PHB2–ACSL3 lipid-metabolic axis that drives ferroptosis escape and identify Mavorixafor repurposing as an immediately translatable strategy to overcome cisplatin resistance in treatment-refractory GC.