Identification and experimental verification of an anoikis and immune related signature in prognosis for lung adenocarcinoma

BACKGROUND: Both metastasis and immune resistance are huge obstacle in lung adenocarcinoma (LUAD) treatment. Multiple studies have shown that the ability of tumor cells to resist anoikis is closely related to the metastasis of tumor cells. METHODS: In this study, the risk prognosis signature related to anoikis and immune related genes (AIRGs) was constructed by cluster analysis and the least absolute shrinkage and selection operator (LASSO) regression by using The Cancer Genome Atlas (TCGA) Program and the Gene Expression Omnibus (GEO) database. Kaplan-Meier (K-M) curve described the prognosis in the different groups. Receiver operating characteristic (ROC) was applied to evaluate the sensitivity of this signature. Principal component analysis (PCA), t-distributed stochastic neighbor embedding (t-SNE), independent prognostic analysis, and nomogram were utilized to assess the validity of the signature. In addition, we used multiple bioinformatic tools to analyze the function between different groups. Finally, mRNA levels were analyzed by quantitative real-time PCR (qRT-PCR). RESULTS: The K-M curve showed a worse prognosis for the high-risk group compared to that for the low-risk group. ROC, PCA, t-SNE, independent prognostic analysis and nomogram showed well predictive capabilities. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis revealed that differential genes were mainly enriched in immunity, metabolism, and cell cycle. In addition, multiple immune cells and targeted drugs differed in the two risk groups. Finally, we found that the mRNA levels of AIRGs were remarkably different in normal versus cancer cells. CONCLUSIONS: In short, we established a new model about anoikis and immune, which can well predict prognosis and immune response.