Preview

PDF, English Download (137MB) | Terms of use

Abstract

Pediatric high-grade gliomas (pHGGs) are the leading cause of cancer related death in children and among thevmost aggressive and devastating brain tumors, characterized by rapid progression and poor prognosis. Withinvthis group, H3K27M diffuse midline gliomas (DMGs) stand out as particularly aggressive, due to a definingvmutation in the histone H3 genes. This mutation results in the substitution of lysine 27 with methionine (K27M),vwhich impairs the function of the polycomb repressive complex 2 (PRC2), leading to a global reduction in histone H3 lysine 27 di- and trimethylation (H3K27me2/3) and subsequent increase in H3K27 acetylation (H3K27ac). These epigenetic changes affect gene regulation, promoting oncogenic processes that drive tumor growth and resistance to conventional therapies. Despite extensive research, there are currently no effective treatment strategies that specifically target the unique pathology of H3K27M DMGs, underscoring a critical need for the development of novel therapeutic approaches. In this thesis, I focused on developing a targeted combination therapy for H3K27M DMGs. To contextualize the unique therapeutic responses associated with H3K27M DMGs, this work also compared the H3K27M DMGs with other pHGGs harboring different histone H3 statuses, such as H3 wild-type (H3 WT) and H3 glycine 34 mutants (H3G34R/V). Each of these tumor types exhibits distinct molecular signatures and clinical outcomes, highlighting the necessity of tailored therapeutic approaches. Epigenetic drug library screens of nine H3K27M, two H3G34R/V and three H3 WT pHGG cell models along with three non-malignant cell lines identified atuveciclib, mivebresib and alisertib as hits with H3K27M DMG-selective potency. These hits targeted cyclin-dependent kinase 9 (CDK9), bromodomain and extra-terminal domain motif containing protein (BET) and aurora kinase (AURK), respectively. These top hits were further subjected to combination screens to discover optimal compound combinations. The most potent combination partners were identified to be CDK9 inhibitor (CDK9i) atuveciclib and BET inhibitor (BETi) BMS-986158, the interaction of which was confirmed to be synergistic for H3K27M DMG models. Two different compound combinations were investigated: combo high (one to one combination of the two drugs at their 50% inhibitory concentration) and combo low (combination of the two drugs reaching to 50% inhibition), the latter emphasized the synergistic interaction between the compounds. High content microscopy (HCM) revealed that single-agent treatments with atuveciclib (CDK9i) and BMS-986158 (BETi) demonstrated cytostatic effect. The outcome of the combo low treatment was similar to monotherapies stressing that same effect size could be achieved with the combination of the two compounds at much lower doses compared to using higher doses of the individual compounds, increasing the therapeutic window and potentially reducing the risk of side effects. On the other hand, combo high elicited apoptotic cell death, which was specifically and significantly more pronounced for H3K27M DMG models. This was further validated in vivo using a zebrafish xenograft model. In line with these results, combo high prevented colony formation of an H3K27M DMG model in soft-agar colony formation assay. The effect of the combination treatment was also examined in conjunction with radiotherapy (RT), which is the standard of care for H3K27M DMGs. Observed synergism or additivity for H3K27M DMG models in the RT screens with compound treatments suggested that this combination therapy is compatible with RT. Multi-omics analyses revealed that DNA replication and cell cycle-related genes were downregulated by atuveciclib (CDK9i), BMS-986158 (BETi) and combo low. Different DNA damage response (DDR) pathways were suppressed by these treatments, indicating a synergistic effect on inhibiting DNA repair. Moreover, DNA damage accumulation was observed under combo high treatment, leading to the hypothesis that both CDK9 and BET inhibitors induce DNA damage primarily through R-loop formation, caused by collisions of the replication machinery with stalled RNA polymerase II (RNAPII), which remains unresolved due to downregulation of spliceosome by the presence of both inhibitors. This damage accumulates over time due to reduced DNA damage repair activity. Cells are unable to halt progression through the cell cycle due to underlying mutations in TP53 or PPM1D or due to the impact of inhibitors on cell cycle checkpoints, proceed through mitosis with accumulating DNA damage, eventually leading to apoptosis via mitotic catastrophe. In summary, compound and synergy screens discovered the synergistic combination of CDK9 and BET inhibitors, which was validated to be specific and more cytotoxic for H3K27M DMG models in vitro and in vivo in zebrafish embryo xenografts. The combinatorial inhibition of CDK9 and BET capitalizes on the transcriptional vulnerabilities of H3K27M DMGs, characterized by decreased H3K27me2/3 and increased H3K27ac, a result of PRC2 dysfunction. These inhibitors together trigger apoptosis, by most likely synergistically inducing DNA damage, which remains unrepaired due to the downregulation of DDR pathways. Future in vivo studies in mice are planned to further assess the clinical viability of the CDK9 and BET inhibitor combination.