Glycogen metabolism genes as a molecular signature for subtyping, prognostic prediction, and immunotherapy selection in clear cell renal cell carcinoma.

Abstract

Glycogen accumulation is a typical feature in clear cell renal cell carcinoma (ccRCC). It has been reported that glycogen metabolism-related genes can promote the progression of ccRCC, but its role in molecular typing, prognosis, immune infiltration, and immunotherapy response has rarely been reported. We applied an unsupervised clustering approach for molecular typing of ccRCC. The least absolute shrinkage and selection operator regression (LASSO) was used for prognostic model construction. The robustness of the model is evaluated by multicenter mutual verification. Weighted gene co-expression network analysis (WGCNA) was used to explore potential biological mechanisms. RT-qPCR was used to identify mRNA relative expression. We found ccRCC can be divided into two subtypes based on glycogen metabolism-related genes, and the prognosis of patients between the two subtypes is significantly different. Furthermore, we constructed a prognostic model for ccRCC patients based on glycogen metabolism-related genes using LASSO algorithm. We found that the model has a strong prognostic effect. Subsequently, we explored the underlying mechanisms through WGCNA and found that the model is associated with immune-related signaling pathways. Finally, we also found that this prognostic model can be used as a marker of response to immunotherapy in patients with advanced ccRCC. In conclusion, glycogen metabolism-related genes have critical value in molecular typing and prognosis evaluation of ccRCC.