A risk score model based on glycosylation-related genes for predicting radioresistance and prognosis of lung adenocarcinoma.

Radiotherapy resistance (RR) is the main cause of radiotherapy failure in lung cancer patients, and its mechanisms are still unrevealed. Glycosylation, as a type of post-translational modification of proteins, plays a key role in tumor progression. Some studies have shown a strong link between glycosylation and RR. However, the absence of a systematic glycosylation-related genes (GRGs) model to predict radiotherapy efficacy in lung adenocarcinoma (LUAD) patients highlights a significant clinical and research gap. The aim of the research was to investigate the prognostic characteristics of GRGs in LUAD treated with radiotherapy. RNA sequencing data of LUAD were obtained from The Cancer Genome Atlas (TCGA) database. The expression and prognostic significance of GRGs in patients who underwent radiotherapy were analyzed with bioinformatics tools, and the Gene Expression Omnibus (GEO) database was used for verification. Gene set enrichment analysis (GSEA), Gene Ontology (GO), Kyoto Encyclopedia of Genes and Genomes (KEGG), etc. were used to analyze the potential mechanism of risk model constructed by GRGs in LUAD. The predictive significance of risk model was investigated by immune infiltration analysis, somatic mutations, and drug susceptibility analysis, etc. Single-cell sequencing and molecular docking were used to find new potential targets for LUAD patients. Finally, our bioinformatics analysis results were verified by wet experiments. GO and KEGG analyses found that glycosylation played a pivotal role in LUAD RR. Forty-four differentially expressed radiotherapy-related glycosylation genes (DERRGGs) were identified in LUAD. KREMEN2, NRARP, QSOX2, GOLGA3, CELSR2, and SRI were screened out by least absolute shrinkage and selection operator (LASSO) analysis. A new risk model was constructed by these six DERRGGs, which showed good predictive power. Multivariate regression found that RiskScore was an independent prognostic factor. Immune infiltration analysis suggested that patients in the high-risk group were more susceptible to suffer from immunosuppression. Single-cell sequencing analysis showed the six genes were mainly distributed in malignant tumors. Drug sensitivity analysis found that the patients in the high-risk group were more sensitive to the clinical drugs, such as afatinib, cytarabine, gemcitabine and so on. Molecular docking demonstrated that tretinoin showed good binding affinity with NRARP, KREMEN2 and QSOX2. Our wet experiment results not only demonstrated that NRARP, KREMEN2 and QSOX2 were more abundant in LUAD irradiation-resistance cells and NRARP protein was significantly up-regulated in radiation-resistant samples, but also showed that tretinoin inhibited the survival of the irradiation-resistance cell obviously. This study constructed glycosylation related RiskScore to predict the prognosis of LUAD patients, specifically in the context of radiotherapy. We explored the relationship between radiotherapy efficacy, glycosylation and prognosis of LUAD patients, which provides new ideas for personalized treatment of LUAD patients. And we suggested that tretinoin may be a potential radiotherapy sensitizer for LUAD, providing a foundation for future investigations.