A cuproptosis-related lncRNAs risk model to predict prognosis and guide immunotherapy for lung adenocarcinoma
Qixuan Li1,2,3, Tianyi Wang1,2,3, Jiaqi Zhu1,2,3
1Nantong Key Laboratory of Translational Medicine in Cardiothoracic Diseases, and Research Institution of Translational Medicine in Cardiothoracic Diseases, Affiliated Hospital of Nantong University, Nantong, China.
Background:
Cuproptosis, one of the newest forms of cell death induction, is attracting mounting attention. However, the role of cuproptosis in lung cancer is currently unclear. In this study, we constructed a prognostic signature utilizing cuproptosis-related long noncoding RNAs (CRL) in lung adenocarcinoma (LUAD) and researched its clinical and molecular function.
Methods:
RNA-related and clinical data were downloaded from The Cancer Genome Atlas (TCGA) database. Differentially expressed CRLs were screened using the 'limma' package of R software. We used coexpression analysis and univariate Cox analysis to further identify prognostic CRLs. Applying least absolute shrinkage and selection operator (LASSO) regression and Cox regression models, a prognostic risk model based on 16 prognostic CRLs was constructed. To validate prognostic CRL function in LUAD, vitro experiments were conducted to explore the expression of GLIS2-AS1, LINC01230, and LINC00592 in LUAD. Subsequently, according to a formula, patients in the training, test, and overall groups were split into high- and low-risk groups. Kaplan-Meier and receiver operating characteristic (ROC) analyses were applied to assess the predictability of the risk model. Finally, the associations between risk signature and immunity-related analysis, somatic mutation, principal component analysis (PCA), enriched molecular pathways, and drug sensitivity was investigated.
Results:
A cuproptosis-related long noncoding RNAs (lncRNAs) signature was constructed. Using quantitative polymerase chain reaction (qPCR) trial, we verified that the expressions of GLIS2-AS1, LINC01230, and LINC00592 in LUAD cell lines and tissues were consistent with the above screening results. Based on this signature, a total of 471 LUAD samples from TCGA data set were split into two risk groups based on the computed risk score. The risk model showed a better capacity in predicting prognosis than traditional clinicopathological features. Moreover, significant differences were found in immune cell infiltration, drug sensitivity, and immune checkpoint expression between the two risk groups.
Conclusions:
The CRLs signature was shown to be a prospective biomarker to forecast prognosis in patients with LUAD and presents new insights for personalized treatment of LUAD.
