Introduction

High-grade gliomas (HGGs) are central nervous system (CNS) tumors that occur in both children and adults, although bear distinct molecular features and neuroanatomy in younger compared with older patients [1,2,3,4]. While these types of cancers are relatively rare, patient prognosis is quite poor, with an average 2-year overall survival of only 20% [5], despite multimodal treatment regimens consisting of surgery, radiation therapy, and chemotherapy [6, 7]. Tumor location or disease progression can further complicate therapeutic interventions; therefore, novel treatment modalities such as targeted therapies, including epigenetically directed therapies, are critical to improve patient outcomes.

The identification of cancer-specific targets supporting tumor growth is a major research objective to develop therapeutic options in HGG. Advances in sequencing technology and single-cell analyses of HGGs and subsets of medulloblastoma have revealed frequent alterations in kinase signaling proteins and proteins which regulate their activity (i.e., tumor suppressors) [8,9,10,11]. For example, several receptor tyrosine kinases (RTKs) are frequently overexpressed or mutated in HGG [8, 11], causing hyperactivity of downstream signaling cascades leading to increased cell proliferation, growth, and survival of cancer cells. To date, there has been limited clinical efficacy from single-agent inhibition of dysregulated signaling pathways in HGG [12].

In addition to aberrant proliferative signaling pathways, disruption of the epigenome has been identified as a contributor to tumorigenesis, cancer progression, and chemotherapy resistance [13, 14]. Epigenetic regulators, termed “writers” and “readers”, catalyze the reversible chemical modifications of histones and DNA [15]. The most predominant epigenetic alterations are post-translational modifications to histones, involving the addition and removal of methyl and acetyl marks, and DNA methylation [16]. The epigenetic “readers” recognize specific modifications and translate their effects on gene expression and other cellular processes. Pharmacological inhibitors have been developed to target various chromatin regulators, such as those directed against the histone deacetylase (HDAC) family, bromodomain and extraterminal (BET)-containing proteins, protein arginine methyltransferase (PRMT) family, Enhancer of zeste homolog 2 (EZH2), and lysine-specific demethylase 1 (LSD1).

Understanding the complex interactions between the cancer genome and epigenome is paramount when designing novel therapeutic strategies. Targeted therapies directed solely at epigenetic regulators or dysregulated kinases have shown limited success in sustaining clinical responses in HGG [17,18,19,20,21,22]. Evidence increasingly shows that epigenetic modulators cooperate with several relevant kinase signaling pathways in gliomas to promote cancer progression and contribute to therapeutic resistance. In this review, we systematically explore the epigenetic regulators and their interactions with kinase signaling networks in HGG and how combination strategies have developed and could be envisioned via existing small molecule inhibitors.

Molecular dysfunction of the genome and epigenome

Kinase signaling pathway alterations

Over the past two decades, substantial effort has been made to sequence and molecularly characterize primary cancer samples across all cancer types, including adult and pediatric HGG. This initiative has identified frequent alterations in several kinase signaling pathways and their regulators (Table 1.) One of the more common alterations detected involve the RTKs, including epidermal growth factor receptor (EGFR), platelet-derived growth factor receptor (PDGFRA), and KIT, also known as mast/stem cell growth factor receptor. These alterations are composed of gene mutations and/or copy number amplification, which hyperactivate downstream signaling networks involved in cell proliferation, differentiation, cell growth, metabolism, and survival. One convergent activating pathway downstream of these RTKs is phosphatidylinositol-3 kinase (PI3K)/AKT/mammalian target of rapamycin (mTOR) signaling. Alterations are found in the catalytic subunit of PI3K, PIK3CA, and the regulatory subunit, PIK3R1, which can activate downstream signaling at AKT and mTOR. Additionally, copy number deletions are found within Phosphatase and Tensin Homolog (PTEN), a tumor suppressor that negatively regulates PI3K/AKT signaling. A common deletion in PTEN includes homozygous deletion, which contributes to hyperactivation of the PI3K/AKT/mTOR pathway.

In addition to the PI3K/AKT pathway, the mitogen-activated protein kinase (MAPK) cascade is downstream of RTKs and involved in HGGs. In this pathway, loss of function mutation in neurofibromin 1 (NF1), a small GTPase activating protein that regulates signal transduction through RAS (Rat sarcoma virus), affects the downstream MAPK signaling leading to activation. Lastly, the cell cycle control gene, cyclin-dependent kinase inhibitor 2 A (CDKN2A), is frequently found to have homozygous deletion. Beyond the aforementioned kinase alterations, several other gene alterations form the HGG genomic landscape. When comparing adult and pediatric gene alterations, there are many overlap** changes within the kinase signaling pathways; however, the frequency at which they occur differs with age.

Table 1 Alterations within kinase signaling cascades and their regulators across adult and pediatric high-grade gliomas [23, 24]. The frequency of gene mutations and copy number alterations were obtained from the cBioPortal for Cancer Genomics using two data sets, Glioblastoma Multiforme (TCGA, Firehose Legacy) and Pediatric Brain Cancer (CPTAC/CHOP, Cell 2020). AMP – amplification; CNA – copy number alterations; HOMDEL – homozygous deletion; pHGG – pediatric high-grade glioma
Full size table

Epigenetic abnormalities from histone modifiers

Unlike genetic changes, epigenetic dysregulation is not typically the result of mutations and instead occurs through alterations in chromatin accessibility and gene expression [25]. Transcriptional dysregulation can result from overexpression of chromatin modulators and their subsequent hyperactivity. The absence or presence of specific histone modifications, including acetylation and methylation of critical amino acids, governs chromatin structure leading to changes in gene transcription.

Numerous enzymes and protein complexes have been identified to be responsible for regulating the expression of various genes. These epigenetic proteins are over- or under-expressed in tumors, including HGG. One of the most well-studied epigenetic modulators is the HDAC family of enzymes. In gliomas, HDAC activity generally suppresses the expression of regulatory proteins and DNA repair genes as a component of repressive transcriptional complexes. Several HDAC family members have been identified as having altered gene expression in HGG. For example, HDAC1, 2, 3, and 7 have been found to be overexpressed in grade IV gliomas compared to normal brain tissue and low-grade gliomas [26]. Meanwhile, other enzymes that act as readers of these acetylated amino acids are often dysregulated in gliomas. These sets of proteins include the BET proteins. Two BET proteins, BRD2 and BRD4, are significantly overexpressed in gliomas, and the knockout of BRD4 diminishes glioma proliferation [27]. Similarly, PRMT enzymes, such as PRMT1, 2, and 5, function to methylate arginine residues and can promote dysregulation in brain tumors arising from the aberrant expression or activity [28,29,30,31]. Another epigenetic regulator relevant to HGG is the methyltransferase, EZH2, which is overexpressed in gliomas and correlated with high-grade gliomas [32, 33]. EZH2’s activity contributes to glioma progression by silencing tumor-suppressor genes [34]. An additional epigenetic modulator that is commonly dysregulated in gliomas is LSD1, a histone demethylase that is a component of several repressive complexes and is associated with reduced gene transcription. LSD1 is found to be overexpressed in several cancer types, including glioblastoma [35,36,37], and has been associated with poor patient prognosis in certain types of cancer [53]. For example, methylation profiling can be used as a surrogate to identify the mutation status of isocitrate dehydrogenase (IDH) [53]. IDH mutations, associated with the production of an oncometabolite (2-hydroxyglutarate), leads to global hypermethylation of the CpG islands, thereby causing gene silencing [54, 55]. The presence of this methylation pattern in gliomas is called the glioma – CpG island methylator phenotype (G-CIMP) and occurs frequently in low-grade gliomas [54, 55]. Large cohort studies have shown that G-CIMP is highly associated with the presence of an IDH mutation and correlated with a favorable prognosis [56]. Similarly, histone mutations common to pediatric HGGs lead to global changes in DNA methylation that can be detected through DNA methylation profiling [55, 57]. Methylation profiling can be used to derive copy number profiles inclusive of gene amplifications/deletions and chromosome alterations (7+/10- and 1p/19q codeletion) associated with different glioma subtypes [53]. Finally, the DNA methylation status of O6-methylguanine-DNA methyltransferase (MGMT) is widely used as a predictive biomarker for therapeutic response to the alkylating agent, temozolomide, and as a prognostic marker in glioblastoma patients [55, 58]. Thus, the value of methylation profiling in gliomas is expanding beyond its impact on gene expression, to help determine tumor phenotype and prognosis.

The World Health Organization (WHO) is beginning to adopt DNA methylation profiling in their classification of CNS tumors to provide a robust classification method and identify new tumor types and subtypes [53]. Notably, the emergence of DNA methylation profiling reveals an intriguing overlap with distinct kinase mutations traditionally associated with gliomas. For example, through methylation profiling, HGG is grouped into eight classes, including three classes characterized by RTK alterations, such as PDGFRA and EGFR amplification [59]. Furthermore, in pediatric HGGs with histone alterations, methylation profiling revealed several alterations in kinase signaling pathways, including PDGFRA, EGFR, KIT, MET, KRAS, PTEN, PIK3CA, and CDK4/6 [57]. Additional studies of pediatric HGGs used DNA methylation alongside whole genome sequencing and RNA sequencing in a clinical trial to molecularly characterize tumors and determine a treatment approach based on the identified alterations [60]. Overall, these studies highlight the potential of methylation profiling and its utility in understanding glioma subtypes, their associated kinase alterations, and appropriate therapeutic selection.

Single-agent targeted therapies in clinical development

Kinase inhibitors

Clinical trials have been implemented and are currently underway to assess the safety and efficacy of small molecule inhibitors directed against protein or lipid kinases in HGG patients with kinase dysregulation and aberrant signaling activation (Table 3). Several clinical studies have tested the effects of inhibition of RTKs, including small molecule inhibition of EGFR, PDGFR, FGFR (fibroblast growth factor receptor), c-MET (mesenchymal-epithelial transition factor), KIT (also referred to as CD117), AXL, and VEGFR (vascular endothelial growth factor receptor). Numerous inhibitors have assessed inhibition of EGFR in HGG as single agents and predominantly show a tolerable safety profile without improvements in survival. Likewise, the results from available clinical trials targeting other RTKs, PDGFR and MET, show tolerable safety profiles but limited efficacy. Currently, other RTK inhibitors targeting FGFR and KIT are being investigated in early-phase trials to determine the safety and efficacy of these small molecule inhibitors for adult (FGFR and AXL) and pediatric HGGs (KIT). Studies are also investigating the inhibition of intracellular signaling kinases in HGG downstream of RTKs. For example, clinical studies are ongoing in pediatric HGG patients receiving small molecule inhibitors against MEK and PI3K. Overall, many completed clinical trials conclude that single-agent inhibitors have tolerable toxicities in phase I but limited efficacy when assessed in phase II trials. This lack of clinical efficacy has been extensively reviewed elsewhere [61,126, 153]. Thus, this idea of identifying predictive biomarkers, or precision medicine, by analyzing a patient’s genomic landscape will improve patient selection and spare non-responders from toxicity. Inhibitors need to be more effective at lower doses and given to patients with a high likelihood of a response to improve the toxicity profile of targeted therapy.

Beyond safety, resistance is another obstacle to targeting chromatin modifiers and kinase signaling proteins. Due to the redundancy in kinase signaling pathways, inhibiting a single kinase is often unsuccessful, as compensatory pathways mitigate the effects of single-agent kinase inhibition. An additional mechanism of resistance to kinase inhibition is through epigenetic changes. Furthermore, epigenetic regulation is highly interdependent between the writers, readers, and erasers and can produce unexpected effects that may limit therapeutic efficacy. For example, inhibition of HDAC via vorinostat can increase H3K4 methylation, making it vulnerable to LSD1 demethylation [35]. One solution to overcome resistance and improve drug toxicity profile is to design combination treatment strategies inclusive of kinase inhibition and epigenetically directed inhibition.

Several in vitro and in vivo models have identified synergistic relationships between kinase and epigenetic inhibition. However, there is still a need to explore the interplay between epigenetic regulators and kinase signaling pathways and understand their specific mechanisms. Identifying safe drug combinations with synergistic or additive effects may afford improvements in efficacy and create an opportunity in the clinical trial setting for HGGs. In the clinic, treatment combinations could allow for dose reductions that exert a clinical effect and decrease unwanted side effects. Certain drug combinations may also circumvent resistance mechanisms associated with single-agent inhibition to extend a drug response. Additionally, other combination treatment strategies to overcome the lack of single-agent success may include introducing immunotherapies to either kinase inhibition or epigenetically directed agents. Previous studies have shown that epigenetic alterations change the tumor microenvironment to contribute to the immune suppressive niche for tumor cells. More recently, studies in several cancer types have shown inhibition of epigenetic regulators (LSD1, EZH2, BET proteins) can enhance the anti-tumor immune response of anti-PD1 therapy [13, 154,155,156]. Interestingly, studies in other cancer models highlight that kinase inhibition can enhance anti-tumor effects with immunotherapies [157,158,159]. Investigating triple therapy targeting chromatin modulators and kinase pathways and using immunotherapy in the ongoing efforts to attenuate tumor cell proliferation to improve patient outcomes and increase overall survivability would be worthwhile. In conclusion, enhancing our understanding of the cooperation across the HGG epigenome and genome will guide the development of new therapeutic strategies.