Exploration of telomere-related biomarkers for lung adenocarcinoma and targeted drug prediction

AimBioinformatics analyses were performed to identify telomere biomarkers to develop a diagnostic model for lung adenocarcinoma (LUAD) and to predict potential target drugs for patients with LUAD.BackgroundTelomeres function crucially in maintaining genome stability and chromosome integrity, and telomere-related genes (TRGs) serve as potential prognostic markers in a variety of cancers. However, studies focusing on TRGs in LUAD are limited.ObjectiveTo screen key telomere-related markers for LUAD and to evaluate their potential impact on the occurrence and development of LUAD.MethodsLUAD samples were collected from University of California Santa Cruz (UCSC) Xena and 2093 telomere-related genes (TRGs) were obtained from TelNet database. Hub genes were screened using "WGCNA" package. Differentially expressed genes (DEGs) between tumor and control samples were filtered using "DESeq" package. Protein-protein interaction (PPI) network analysis was performed to select candidate genes, from which telomere-related biomarkers were identified by machine learning and used to develop a nomogram. Functional enrichment pathways of the biomarkers were analyzed using "clusterProfiler" package. Correlation between immune cell infiltration and the biomarkers was examined by Spearman method. Targeted drugs were predicted and molecular docking models were developed using AutoDockTools. Finally, the screened biomarkers were validated by performing in vitro cellular assays.ResultsA total of 259 hub genes, 2848 DEGs, and 48 differentially expressed TRGs in LUAD were screened. Subsequently, 13 candidate genes were obtained by PPI network analysis. LASSO and support vector machine-recursive feature elimination (SVM-RFE) algorithms further reduced the number of telomere-related biomarkers to four (CCNB1, CDC20, PLK1, and TOP2A). A nomogram with a strong predictive performance was created. These four biomarkers were mainly enriched in the mitogenic pathways and exhibited a strong correlation with immune cell infiltration. Three drugs (Lucanthone, Fulvestrant, and Myricetin) targeting the four biomarkers were predicted to be able to treat LUAD. Finally, in vitro cellular experiments demonstrated that CCNB1 and PLK1 have potential effects on proliferation, migration, invasion and AKT/mTOR signaling pathway in LUAD cells.ConclusionThis study provided novel diagnostic biomarkers, therapeutic targets, and potential drugs for LUAD.