A novel risk model construction and immune landscape analysis of gastric cancer based on cuproptosis-related long noncoding RNAs.

Abstract

Recent studies have identified cuproptosis, a new mechanism of regulating cell death. Accumulating evidence suggests that copper homeostasis is associated with tumorigenesis and tumor progression, however, the clinical significance of cuproptosis in gastric cancer (GC) is unclear. In this study, we obtained 26 prognostic cuproptosis-related lncRNAs (CRLs) based on 19 cuproptosis-related genes (CRGs) via Pearson correlation analysis, differential expression analysis, and univariate Cox analysis. A risk model based on 10 CRLs was established with the least absolute shrinkage and selection operator (LASSO) Cox regression analysis and multivariate Cox proportional hazards model to predict the prognosis and immune landscape of GC patients from The Cancer Genome Atlas (TCGA). The risk model has excellent accuracy and efficiency in predicting prognosis of GC patients (Area Under Curve (AUC) = 0.742, 0.803, 0.806 at 1,3,5 years, respectively, P < 0.05). In addition, we found that the risk score was negatively correlated with the infiltration of natural killer (NK) cells and helper T cells, while positively correlated with the infiltration of monocytes, macrophages, mast cells, and neutrophils. Moreover, we evaluated the difference in drug sensitivity of patients with different risk patterns. Furthermore, low-risk patients showed higher tumor mutation burden (TMB) and better immunotherapy response than high-risk patients. In the end, we confirmed the oncogenic role of AL121748.1 which exhibited the highest Hazard Ratio (HR) value among 10 CRLs in GC via cellular functional experiments. In conclusion, our risk model shows a significant role in tumor immunity and could be applied to predict the prognosis of GC patients.