Key points are not available for this paper at this time.
Background Immune checkpoint blockade targeting programmed death-1/programmed death-ligand 1 (PD-1/PD-L1) has revolutionized cancer therapy. However, its efficacy is frequently limited by primary and acquired resistance. While inflammatory signals transiently upregulate PD-L1 transcription, post-translational regulation is crucial for its sustained expression in chronically stressed tumors. Whether and how tumor-intrinsic stress-response pathways control PD-L1 stability to promote immune evasion remains incompletely understood. Methods Using human oral squamous cell carcinoma (OSCC) specimens, syngeneic mouse models, single-cell RNA sequencing, and genetic and pharmacological perturbations, we dissected the role of the unfolded protein response effector ATF4 in regulating PD-L1 stability and antitumor immunity. Its therapeutic potential was further evaluated in immunocompetent mice treated with anti-PD-1 therapy. Results We identified ATF4 as a tumor-intrinsic driver of immune evasion. In malignant cells, ATF4 induced reactive oxygen species (ROS), which activated the AKT-mTOR pathway and suppressed autophagy, thereby stabilizing the PD-L1 protein independently of inflammatory cues. Genetic ablation of ATF4 or pharmacological inhibition of ROS-AKT-mTOR signaling restored autophagic flux, reduced PD-L1 levels, and enhanced CD8 + T-cell infiltration and function. When combined with PD-1 blockade, ATF4 targeting further suppressed tumor growth. Clinically, ATF4 expression was inversely correlated with CD8 + T-cell infiltration and autophagy markers, positively correlated with PD-L1 levels, and predicted a poor response to immunotherapy in patients with OSCC. Conclusions Our findings establish ATF4 as a stress-responsive regulator of PD-L1 proteostasis, directly linking tumor-intrinsic stress adaptation to immune checkpoint stabilization and therapy resistance. Targeting the ATF4-ROS-AKT-mTOR axis may represent a promising strategy to overcome resistance to PD-1/PD-L1 blockade.
Wang et al. (Fri,) studied this question.