Background
Extensive research has revealed that tumor stemness plays a central role in promoting tumor progression. However, the underlying involvement of stemness-related genes in renal clear cell carcinoma (ccRCC) remains controversial.
Methods
The data used for bioinformatics analysis were downloaded from The Cancer Genome Atlas database. The R software, SPSS and GraphPad Prism 8 were used for mapping and statistical analysis.
Results
We first quantified the stemness index of each patient through a machine learning algorithm. Then, we identified the differentially expressed genes between high and low stemness index as stemness-related genes. Based on these genes, we finally established a stable and effective prognosis model to predict patients' overall survival using a random forest algorithm (Training cohort; 1-year AUC: 0.67; 3-year AUC: 0.79; 5-year AUC: 0.73; Validation cohort; 1-year AUC: 0.66; 3-year AUC: 0.71; 5-year AUC: 0.7). The model genes include AC010973.2, RNU6-125P, AP001209.2, Z98885.1, KDM5C-IT1 and AL021368.3. The gene AC010973.2 was selected for further research for its highest importance. In vitro experiments demonstrated that AC010973.2 is highly expressed in ccRCC tissue and cell lines. Meanwhile, knockdown of AC010973.2 could significantly hamper the proliferation of ccRCC cells according to the colony formation and CCK8 assays.
Conclusion
In summary, our finding indicated that the stemness-related gene AC01097.3 is closely associated with patients' survival and could remarkably facilitate cell proliferation in ccRCC, making it potential to be a novel therapeutic target.