Epigenomic Reprogramming and Therapeutics Development in High-Grade Gliomas

Abstract

Tumors originating in the central nervous system comprise the most common class of childhood malignancy. Approximately 20% of pediatric gliomas are high-grade with an unsatisfactory long-term survival (>5 years) of 10-15%. Trials to develop and refine chemotherapeutic and radiation regimens for these fatal malignancies have seen little progress. Oncogenes classically have been traditionally considered genes, such as MYC, MET, and RAS, that induce malignant transformation via cellular reprogramming. These genes are often mutated, leading to hyperactivity or resistance to degradation, and drive tumorigenesis. Recent work has demonstrated that histones, the building blocks for chromatin, are recurrently mutated in bone cancers and gliomas. Although our knowledge on the role of histone mutations in perturbing epigenetic modifications has expanded from a biochemical perspective, fewer comprehensive studies identifying therapeutic vulnerabilities and oncogenic signaling mechanisms engaged by histone mutants have been completed. This dissertation aims to characterize specific epigenetic mechanisms in H3.3K27M and H3.3G34R/V gliomas and identify therapeutics for each subtype. In Chapter 1, the introductory chapter, we discuss important aspects of histone biology, seminal discoveries and biochemical studies for histone mutations, and key pathways implicated in disease pathogenesis for histone mutant gliomas. In Chapter 2, this thesis describes the intersection of H3K27M tumor epigenetics with metabolism and how the epigenetic-metabolic axis in these tumors can be targeted for therapeutic benefit. An integrated transcriptomic, metabolomic, proteomic, and epigenetic analysis of tumor cells and tissues revealed enhanced expression of enzymes involved in glycolysis, glutaminolysis, and the tricarboxylic acid cycle. These upregulated metabolic processes were associated with elevated alpha-ketoglutarate levels in H3.3K27M cells which heterogeneously utilize glutamine and glucose to produce alpha-ketoglutarate. Importantly, we found that depleting sources of alpha-ketoglutarate could raise H3K27me3 and slow tumor cell growth in vitro. Data presented in this chapter provides evidence that this metabolic targeting approach is efficacious in slowing in vivo tumor growth. In Chapter 3, this thesis focuses on understanding therapeutic dependencies arising from epigenetic reprogramming by H3.3G34R/V mutations. Comprehensive omics analysis revealed concomitant and cooperative changes in DNA and histone methylation that led to epigenetic activation of the Leukemia Inhibitory Factor/Signal Transducer and Activator of Transcription (LIF/STAT3) signaling pathway. We found that LIF activated STAT3 in an autocrine/paracrine manner to promote H3.3G34R/V glioma cell survival in vitro. Further immunohistochemical and single-cell sequencing studies on patient samples confirmed pervasive STAT3 activation and expression, suggesting LIF/STAT3 as a potential therapeutic axis for these tumors. Furthermore, the data presented in this chapter demonstrates that STAT3 may regulate other master forebrain transcription factors that are critical for H3.3G34R/V gliomagenesis. Additional experiments established multiple preclinical animal models of H3.3G34R/V tumors using mouse and patient-derived tumor cells. Genetic and pharmacological targeting of STAT3 using these models provide the first evidence that LIF/STAT3 is a potent target for therapeutic intervention. Finally, Chapter 4 summarizes and synthesizes the findings of each body of work presented in this dissertation. Critical remaining questions and future directions of study are discussed. Collectively, the work presented sought to uncover epigenetically-driven mechanisms of gliomagenesis for distinct tumor subtypes and to provide much needed therapeutic leads