ATRXLoss to impair glioma stem cell viability and self-renewal via proton radiotherapy induced necrosis.

Abstract

2046 Background: Glioblastoma Multiforme (GBM-Grade IV) is the most aggressive form of glioma in the USA carrying only a 15 month median survival time. Although IDH1R132H is traditionally associated with improved glioma patient survival and clinical response to chemotherapy, astrocytoma (LGG) patients also harbor ATRXLoss, an inactivation of the ATRX chromatin remodeling protein shown to impair DNA damage repair via non-homologous end joining. The recent discovery of glioma stem cells (GSCs), a subpopulation of chemoradioresistant, self-renewable, tumor initiating cells known to promote tumor recurrence, presents a new target for therapeutic intervention. However, the mechanisms by which ATRXLoss contributes to GSC proton radiotherapy (PRT) sensitivity, a more tumor specific and cytotoxic form of radiation therapy than conventional X-rays (XRT), are not well understood. We hypothesize that ATRXLoss impairs GSC post-radiation viability and self-renewal capacity by promoting necrosis to a higher extent after PRT than XRT. Methods: Isogenic TS543-ATRXWT and TS543-ATRXLoss patient derived glioma cells were cultured as neurospheres and treated with 2-8 Gy XRT or PRT. GSC viability was quantified using the CellTiterGlo 3D assay 12 days after irradiation. The Annexin V apoptosis and necrosis assay was conducted every 8 hours up to 72 hours after irradiation. GSC self-renewal, based on neurosphere formation frequency, was quantified using extreme limiting dilution analysis (ELDA). Results: ATRXLoss significantly decreased TS543 GSC viability after 2-8 Gy radiation therapy compared to ATRXWT (Unpaired t-test with Welsh’s correction and Holm-Sidak’s Multiple Comparisons Test p<0.0001). TS543-ATRXLoss also significantly inhibited GSC self-renewal over TS543-ATRXWT when subject to either 6 Gy PRT or XRT (Pairwise Pr(>χ2)<0.01). Interestingly, PRT significantly reduced GSC viability to a higher extent than XRT in both TS543-ATRXWT and TS543-ATRXLoss, which demonstrates the effectiveness of PRT to treat ATRXLoss gliomas. Follow up studies confirmed that ATRXLoss sensitizes TS543 GSCs more extensively to the cytotoxic effects of PRT compared to XRT via the upregulation of necrosis at 72 hours post-irradiation (Two-way ANOVA with Tukey’s post-hoc Multiple Comparisons Test p<0.0001). Conclusions: We conclude that ATRXLoss substantially contributes to PRT sensitivity in GSCs, thereby supporting the benefits of PRT for astrocytoma patients. Our future experiments will further elucidate the biological mechanisms by which ATRXLoss confers PRT sensitivity, especially via dsDNA damage.