Uncover DNA damage and repair-related gene signature and risk score model for glioma

AbstractBackground Glioma is a common primary central nervous system tumor with complex pathogenesis. DNA damage and repair (DDR) is widely involved in regulating cell proliferation and tumorigenesis by correcting and repairing DNA damage mechanisms. Recent studies have reported the following properties in cancer cells in glioma, increased DNA damage and reduced DNA repair capacity. However, the relationship between glioma and DDR-related genes was unclear.Methods DDR-related risk score model was built. The validity of this model was validated in detail through the Kaplan-Meier survival analysis, tumor mutational burden (TMB) analysis, immune cell infiltration, sensitivity to treatment regimens. Moreover, the model’s adaptability was validated in different glioma data cohorts and different glioma subgroups. To further investigate the molecular mechanism of one of DDR-related gene (NUDT1) in glioma, U251 cell was used for the knockdown experiment, followed by MTT, wound healing and transwell analysis.Results Ten prognostic-related DDR-related signature genes were obtained, including EID3, MGMT, YWHAG, PMS1, SHPRH, HUS1, NUDT1, GADD45G, APEX1 and FAM175A. The RT-qPCR results suggested that the latter five genes were highly expressed in glioma patients. Interestingly, high TMB score had longer survival. In high-risk score groups, reduced immune cell infiltration in the tumor microenvironment lead to poorer patient outcomes. Sensitivity to treatment regimens analysis indicated that low-risk score groups were more sensitive to chemotherapeutics. Moreover, the risk score model had a good prediction effect on different glioma datasets and different glioma subgroups. In vitro mechanism study showed that knockdown of NUDT1 reduced tumorigenesis. Furthermore, knockdown of NUDT1 remarkably reduced the expression level of HIF-1α.Conclusion DDR-related risk score model built-in this work has good predictive performance for glioma.Key messagesTen prognostic-related DDR-related signature genes were obtained, including EID3, MGMT, YWHAG, PMS1, SHPRH, HUS1, NUDT1, GADD45G, APEX1 and FAM175A.In high-risk score groups, reduced immune cell infiltration in the tumor microenvironment leads to poorer patient outcomes.The risk score model had a good prediction effect on different glioma datasets and different glioma subgroups.Knockdown of NUDT1 reduced tumorigenesis of glioma and remarkably reduced the expression level of HIF-1α.