Abstract Glioma is the most common and aggressive primary adult brain tumor. Radiation therapy (RT) is a critical part of glioma treatment, but radioresistance severely limits its therapeutic efficacy. To better understand how radioresistance develops, we analyzed 212 paired primary and recurrent gliomas from the Glioma Longitudinal Analysis Consortium. We found a significant increase in mutational signatures associated with the activity of the Apolipoprotein B mRNA-editing enzyme catalytic polypeptide-like (APOBEC) family of cytidine deaminases in RT-treated recurrent gliomas (p<0.001, paired Wilcoxon signed-rank test). Further analysis of post-treatment mutations showed that RT increased APOBEC mutational signatures independent of other factors (p<0.001, multivariable log-linear regression), suggesting that APOBEC proteins become activated by radiation. However, their role in promoting tumor evolution and treatment resistance is currently unknown. Further analyses of extended signature contexts and expression data nominated APOBEC3B (A3B) and APOBEC3G (A3G) as candidates for creating these mutational signatures. To investigate their roles in RT response, we knocked out (KO) A3B and A3G individually in patient-derived glioma cell lines before RT. Compared to Control cells, irradiated A3B-KO and A3G-KO cells acquired significantly fewer APOBEC mutational signatures by whole genome sequencing (p<0.05, Kruskal–Wallis test) and exhibited enhanced radiosensitivity in vitro. Interestingly, the KO cells displayed a significant decrease in mutational signatures associated with Non-Homologous End-Joining (NHEJ), one of the main pathways for repairing the RT-induced DNA damage (p<0.05, Kruskal–Wallis test), suggesting that APOBEC deletion impaired DNA repair. This was further supported by persistent gH2AX DNA damage foci 48h post-RT and reduced autophosphorylation of DNA-dependent protein kinase (DNA-PK), a vital component of the NHEJ pathway, particularly in A3G KO cells. Overall, our results identify APOBEC proteins as potential targets for radiosensitizing gliomas and highlight the clinical utility of using mutational signatures to predict and monitor response to treatment.
Read full abstract