Integrated single-cell RNA-seq analysis reveals that EZH2 regulates the MIF-CD74 axis to modulate T cell activation and exhaustion in hepatocellular carcinoma

Hepatocellular carcinoma (HCC) is a highly lethal malignancy characterized by a complex pathological mechanism involving multiple genes and progressive stages. The efficacy of targeted and immunotherapy remains limited, highlighting the urgent need for a reliable model to predict prognosis and response to immune checkpoint inhibitors (ICIs). We developed an integrated model based on genes related to autophagy, senescence, dormancy, mitochondrial function, and tumor stemness. The predictive capability of this model for HCC prognosis and ICI response was evaluated. Single-cell transcriptomic analysis and immunocompetent mouse models were further utilized to elucidate the role of model-associated genes in regulating the tumor immune microenvironment. A 16-gene integrated model was constructed using genes associated with mitochondrial function, autophagy, dormancy, stemness, and senescence. This model demonstrated robust predictive power for HCC prognosis and ICI responsiveness. Single-cell trajectory analysis revealed that EZH2 plays a crucial role in immune cell infiltration, activation, and HCC progression. Additionally, in vivo mouse models further indicated that EZH2 may regulate CD8+ T cell activation and exhaustion through the MIF-CD74 signaling pathway. The integrated model holds potential as a prognostic and predictive tool for HCC immunotherapy. EZH2 may influence CD8+ T cell activation and exhaustion via the MIF-CD74 axis, providing insights for patient stratification and potential therapeutic strategies to enhance immunotherapy efficacy.