Development of an oxidative phosphorylation signature in high-grade glioma and screening of potential inhibitors

Abstract Background Although there are various ways for cells to supply energy, for mammals, most of the energy of the organism comes from the oxidative phosphorylation pathway. The introduction of the Warburg effect makes it generally believed that the glycolysis process represented by the production of lactic acid is the main way for many tumors to obtain energy. However, recent studies have shown that lactate-based glycolysis is not the primary energy source for all tumors. Especially in glioma, the experimental data show that the energy provided by the oxidative phosphorylation pathway accounts for about 80% of the total tumor consumption. However, at present, there is no gene signature constructed from oxidative phosphorylation-related genes for glioma prognosis assessment, clinical diagnosis, and correlation analysis between oxidative phosphorylation and the immunotherapy response. Materials and Methods In this study, samples from the glioma cohort in the Cancer Genome Atlas (TCGA) database were used as the training set, and samples from the glioma cohort in the Chinese Glioma Genome Atlas (CGGA) database were used as the validation set; univariate Cox regression analysis was used for constructing oxidative phosphorylation signatures. At the same time, the multi-Cox method is used to further optimize this feature to build a risk model. The high-grade glioma samples were further divided into high-risk groups and low-risk groups according to the expression levels of the constituent genes of the oxidative phosphorylation risk model, and the prognosis, proto-oncogenes, and tumor suppressor genes of the high and low-risk groups were analyzed by statistical methods. Associations between mutations, immune checkpoints, and immunotherapy-related factors. Results In this study, a glioma consisting of 17 genes (MPC1, CYB5R3, MRPL3, HCCS, MRPS30, TCIRG1, MRPL12, ATP5F1C, ECH1, CYB5A, ACADM, COX7B, NDUFS5, GRPEL1, CLPP, ATP5PB, and ACADSB) was constructed in glioma prognostic characteristics. High-grade glioma samples were further divided into high-risk groups and low-risk groups according to the expression levels of these 17 genes for prognosis analysis, mutation analysis, immune cell infiltration, immune microenvironment, immune checkpoints, and immunotherapy analysis. The results showed that the model showed good prognosis and diagnostic ability in the glioma cohort; it is worth noting that in the samples from the high-risk group of oxidative phosphorylation, proto-oncogenes and tumor suppressor genes have higher mutation frequencies, stronger immune rejection, and stronger immune disorders and a microsatellite instability environment. Conclusions This study establishes and validates the first OXIP-related signature model in glioma, which has the good predictive ability and diagnostic effects, which may be helpful for the diagnosis, prognostic evaluation, and treatment plan of glioma patients optimization. In addition, the high-risk and low-risk groups differentiated by prognostic characteristics also showed significant differences in the dimensions of proto-oncology and tumor suppressor gene mutations, immune microenvironment, and immunotherapy. Collectively, our findings provide a rationale for the development of novel oxidative phosphorylation phase inhibitors and immunotherapy for glioma.