A lactylation-ferroptosis cross-talk gene signature predicts hepatocellular carcinoma prognosis and reveals STMN1/PRDX1 as therapeutic targets

Introduction Hepatocellular carcinoma (HCC) presents great difficulties for diagnosis and prognosis. Metabolic reprogramming and ferroptosis resistance play critical roles in HCC development and progression. Methods We combined transcriptome data from TCGA-LIHC to screen out lactylation-ferroptosis-related genes (LFRGs). Based on LASSO-Cox regression and multivariate analysis, a 4-gene-prognostic signature (STMN1, PRDX1, TP53, G6PD) was constructed. Results The signature stratified HCC patients into two risk subgroups with obvious survival differences (P < 0.001). Functional validation demonstrated that STMN1/PRDX1 knockdown in MHCC-97H/SNU-449 cells could decrease the expression of LDHA (lactylation enzyme) and GPX4 (ferroptosis inhibitor), and further inhibited cell proliferation and migration (P < 0.01). The risk model was positively correlated with tumor stage and TMB, TP53 mutation, and immunosuppressive microenvironment. Drug sensitivity profile analysis suggested that high-risk patients may be sensitive to Dactolisib/Trametinib, while low-risk patients showed resistance to Axitinib/Ibrutinib. Mechanistically, STMN1/PRDX1 act as dual hubs in lactate metabolism (stabilizing LDHA) and ferroptosis resistance (modulating GPX4). Discussion We identified the first lactylation-ferroptosis cross-talk signature to predict HCC prognosis and identified STMN1/PRDX1 as potential targets for treating HCC by stratifying therapies.