Glucose 6 phosphatase dehydrogenase (G6PD): a novel diagnosis marker related to gastrointestinal cancers.

BACKGROUND: Glucose 6 phosphatase dehydrogenase (G6PD) is a key regulator of the pentose phosphate pathway (PPP). However, the exact role of G6PD in gastrointestinal cancers remains unclear. The purpose of this study is to explore the correlation of G6PD with clinical features, pathological stages, diagnosis and prognosis of gastrointestinal cancers, as well as uncover possible mechanisms of G6PD on mutations, immunity and signaling pathways.

METHODS: G6PD mRNA expression data were downloaded from TCGA and GEO databases. Protein expression was examined by the HPA database. The correlation of G6PD expression with clinical and pathological characteristics was explored. The pROC package in R language was used to evaluate the diagnostic value of G6PD expression in gastrointestinal cancers. We accessed the correlation of disease-free survival (DFS) with G6PD online by Kaplan-Meier plotter. Univariate Cox regression and stepwise multiple Cox regression analysis were performed to determine the association between G6PD and patient's overall survival. In addition, genomic alterations, mutation profiles, immune infiltration, drug sensitivity and enrichment analysis related with G6PD were visualized.

RESULTS: After a pan-cancerous genomic analysis, we found that G6PD expression was the highest in African American esophageal carcinoma (ESCA) patients ( P <0.05). G6PD was correlated with age, weight, disease stage, lymph node metastasis and pathological grade. Notably, G6PD showed an excellent predictive diagnosis ability for liver hepatocellular carcinoma (LIHC) (AUC=0.949, 95% CI=0.925-0.973, P <0.001). G6PD can improve the DFS of esophageal adenocarcinoma (EAC) and pancreatic adenocarcinoma (PAAD) patients ( P <0.05). Both Univariate Cox regression and stepwise multiple Cox regression analysis in R language determined that G6PD expression was closely related with LIHC ( P <0.001). G6PD was found to have a high mutation rate in colon adenocarcinoma and ESCA and gene amplification in ESCA, Cholangiocarcinoma, PAAD and LIHC. Copy number of G6PD was missing in LIHC. G6PD was also related to mutation of TP53 ( P <0.05). Particularly, it was positively correlated with CD276 in all gastrointestinal cancers and negatively with HERV-H LTR-associating 2 in ESCA and stomach adenocarcinoma. The abnormal expression of G6PD was related to the increase of CD4+ Th2 subsets and the decrease of CD4+ (non-regulatory) of T cells. G6PD was sensitive to FK866, Phenformin, AICAR etc., while resistant to RO-3306, CGP-082996, TGX221 etc. G6PD was found to closely interact with TALDO1, GAPDH and TP53. G6PD related biological processes included aging, nutritional response and daunorubicin metabolism, and related pathways included PPP, cytochrome P450 metabolism of exogenous substances and glutathione metabolism.

CONCLUSION: G6PD is highly expressed in gastrointestinal cancers. It is a carcinogenic indicator related to prognosis and can be used as a potential diagnostic marker of gastrointestinal cancers, so as to provide new strategy for cancer treatment.