Combating Treatment Resistance in H3K27M-Mutant Diffuse Midline Glioma

Abstract

Brain tumors are the most common solid tumor in children and account for the highest incidence of cancer-related death. Of these, diffuse midline gliomas (DMGs), including all diffuse intrinsic pontine gliomas (DIPGs), are particularly lethal. DMG tumors are often unresectable given the locations in which they form, and they are often resistant to chemotherapy. This leaves radiation therapy (RT) as the only standard of care that provides meaningful, albeit limited, benefit to patients as tumors typically recur within the high-dose radiation field. Therefore, it is imperative to discover new therapeutic strategies to overcome treatment resistance in H3K27M-DMG. DMG tumors are often characterized by the presence of a driver mutation in the tail domain of Histone H3 that converts the 27th residue from a lysine (K) to a methionine (M), termed H3K27M. This mutation leads to global epigenetic disruption characterized by global depletion of the inhibitory H3K27 trimethylation mark. This loss of H3K27 trimethylation induces tumor-specific aberrant transcriptional programs, metabolic rewiring, and DNA damage repair dynamics. In this dissertation, I highlight two avenues of overcoming treatment resistance in H3K27M-DMG: targeting metabolism and targeting the DNA damage response (DDR). Using steady-state metabolomics in H3K27M-isogenic models, I determined that H3K27M-expressing cells and tissue have a distinct metabolome compared with H3-WT counterparts. Following radiation therapy, the standard of care for patients with this disease, I observed that the most differentially abundant metabolites between H3K27M and H3K27M-KO cells were enriched for purine metabolic pathways. Stable isotope tracing revealed that at steady-state, H3K27M cells rely on de novo synthesis (DNS) to produce purines, especially guanylates. Inhibiting guanylate DNS radiosensitized H3K27M cells well in vitro but only modestly in vivo. I found that H3K27M cells can bypass DNS inhibition of guanylate production through high, constitutively active guanine salvage. Genetically inhibiting the rate-limiting enzyme of DNS guanylate production, HPRT1, blocked this compensatory mechanism and greatly radiosensitized tumors in vivo. I then explored the metabolic effect of the novel anti-H3K27M-DMG therapeutic ONC201 that has shown promising preclinical and clinical success in targeting this disease. Using steady state metabolomics, I found that ONC201-treated H3K27M-DMG cells showed altered glucose metabolism that suggested a break between the flow of glucose-derived carbons from glycolysis into the tricarboxylic acid (TCA) cycle. Further, I performed in vivo stable isotope tracing to determine if this same phenomenon could be observed in live tumor tissue. Unfortunately, the data suggest only limited target engagement, requiring further optimization of drug treatment schedule. Lastly, I wanted to determine the effect of the H3K27M mutation on RT-induced DNA damage repair in DMG to understand how DMG cells promote RT resistance. We found that H3K27M tumors and cells had decreased promoter methylation and increased expression the gene Ataxia Telangiectasia Mutated (ATM), one of the three apical serine/threonine PI3K-like kinases that facilitate the DDR. Further, we found that H3K27M cells had high basal levels of DNA damage when compared to H3K27M-KO controls and higher basal phosphorylation and activation of ATM as a result. Inhibiting ATM signaling increased radiosensitivity in vitro and extended the survival of mice bearing H3K27M-DMG tumors. This work highlights the importance of the H3K27M mutation in influencing DMG cellular biology to facilitate treatment resistance. Together, these findings also illustrate how targeting H3K27M-specific biology may prove effective in decreasing H3K27M-DMG treatment resistance and improving patient outcomes.