A novel Golgi related genes based correlation prognostic index can better predict the prognosis of glioma and responses to immunotherapy
Beichuan Zhao,Ruoheng Xuan,Guitao Yang,Tianyu Hu,Yihong Chen,Lingshan Cai,Bin Hu,Gengqiang Ling,Zhibo Xia
DOI: https://doi.org/10.21203/rs.3.rs-3961611/v1
2024-01-01
Abstract:Abstract Background The Golgi apparatus (GA) serves as the center of protein and lipid synthesis and modification within cells, playing a crucial role in regulating diverse cellular processes as a signaling hub. Dysregulation of GA function can give rise to a range of pathological conditions, including tumors. Notably, mutations in Golgi-associated genes (GARGs) are frequently observed in various tumors, and these mutations have been implicated in promoting tumor metastasis. However, the precise relationship between GARGs and glioma, a type of brain tumor, remains poorly understood. Therefore, the objective of this investigation was to assess the prognostic significance of GARGs in glioma and evaluate their impact on the immune microenvironment. Methods The expression of GARGs was obtained from the TCGA and CGGA databases, encompassing a total of 1564 glioma samples (598 from TCGA and 966 from CGGA). Subsequently, a risk prediction model was constructed using LASSO regression and Cox analysis, and its efficacy was assessed. Additionally, qRT-PCR was employed to validate the expression of GARGs in relation to glioma prognosis. Furthermore, the association between GARGs and immunity, mutation, and drug resistance was investigated. Results A selection of GARGs (SPRY1, CHST6, B4GALNT1, CTSL, ADCY3, GNL1, KIF20A, CHP1, RPS6, CLEC18C) were selected through differential expression analysis and Cox analysis, which were subsequently incorporated into the risk model. This model demonstrated favorable predictive efficiency, as evidenced by the area under the curve (AUC) values of 0.877, 0.943, and 0.900 for 1, 3, and 5-year predictions, respectively. Furthermore, the risk model exhibited a significant association with the tumor immune microenvironment and mutation status, as well as a diminished sensitivity to chemotherapy drugs. qRT-PCR analysis confirmed the up-regulation or down-regulation of the aforementioned genes in glioma. Conclusion The utilization of GARGs in our constructed model exhibits a high level of accuracy in prognosticating glioma and offers promising avenues for the development of therapeutic interventions targeting glioma.