Here, we systematically for the first time analyzed the role of E3-ubiquitin ligase-related LncRNAs in LUAD prognostic prediction by bioinformatics and built a reliable prognostic signature based on seven specific lncRNAs. Further analyses validated the predictive efficacy in the TCGA and GEO datasets. This signature stratified the patients into high- and low-risk subgroups, characterized by distinct tumor immune infiltration level and tumor-related regulatory phenotypes such as EMT and hypoxia via ssGSEA analyses. Through multivariate Cox regression analysis, this signature was included in a fusion nomogram combined with the pathological staging to predict the outcome of LUAD with the I-III stage. Taken together, the newly identified risk model may serve as a promising tool for prognosis estimation of LUAD patients.
Mounting researchers have discovered that E3-ubiquitin ligase-related lncRNAs could mediate immunity and be involved in multiple tumor-related pathways. Our study evaluated 28 immune cell infiltrations and found that risk score was negatively associated with activated CD4+ T cells and effector memory CD8+ T cell that exerted an antineoplastic effect by releasing TNF-α and IFN-γ (Gabrilovich et al., 2012; Apetoh et al., 2011). Thus, low-risk samples tend to have a favorable prognosis. Meanwhile, the high-risk group presented a tight correlation with the pathways/phenotypes including epithelial-mesenchymal transition (EMT), hypoxia, Notch signaling, and PI3K/AKT/mTOR signaling pathways. E3-ubiquitin ligase-related lncRNAs had been described to regulate tumor progression via EMT. Pan et al. demonstrated that lnc-CTSLP4 could reverse EMT of GC cells by recruiting E3 ubiquitin ligase ZFP91 to decrease NRNPAB-mediated Snail transcription, thereby inhibiting GC progression (Pan et al., 2021). LINC00858 could sponge miR-134-5p to promote E3 ubiquitin ligase RAD18 elevation, resulting in enhancing EMT of cancer cells (Xue et al., 2020). Besides, the overexpression of lnc-DARS-AS1 was able to interfere with the interaction between RBM39 and E3 ubiquitin ligase RNF147 via hypoxia regulation, leading to myeloma genesis (Tong et al., 2020). In short, E3-ubiquitin ligase-related lncRNAs are inversely associated with the migration and invasion of LUAD.
We identified 7 E3-ubiquitin ligase-related lncRNAs included in the risk signature and further developed a nomogram. GABPB1-AS1 was highly expressed in normal tissue and acted as a favorable gene in renal carcinoma (Gao et al., 2020a), and its overexpression could inhibit proliferation, migration, and invasion. A previous study reported that LINC02178 may be a promising prognostic biomarker of cancers like LUAD (Li et al., 2018b). Other lncRNAs were reported for the first time and are worthy of further investigation. Nomogram, as a graphical representation tool, has been widely used to integrate a variety of variables especially clinical information and genomics in the evaluation of prognosis (Balachandran et al., 2015). According to the hazard ratio of the various factors, each variable can get a corresponding score, and clinicians will easily calculate the total score and offer the corresponding survival probability to individuals. Currently, researchers have performed nomograms to provide individualized prognosis predictions in multiple cancers like breast cancer, gastric cancer, lymphoma, and lung cancer (Wang et al., 2021; Geng et al., 2021; Gao et al., 2020b; Wu et al., 2021). Herein, we constructed a prognostic nomogram by combining pathological staging with E3-ubiquitin ligase-related lncRNA signature. The result revealed that the accuracy and clinical usefulness of our nomogram for predicting prognosis were excellent compared to the TNM staging system.
There were several limitations in this study. Our study included the TCGA and GEO datasets and all results were derived and validated using them. Given the limited number of patients with an advanced stage in the two datasets, we only established and validated a nomogram for predicting survival probability in stage I–III LUAD patients. Additionally, additional LUAD samples should be employed to estimate the predictive value of lncRNAs. Likewise, the roles of the lncRNAs and their interactions with E3-ubiquitin ligase genes should be determined in future researches in vitro and in vivo.