Hinchcliffe, Edward H.

Abstract EZHIP (EZH inhibitory protein) is a cancer-testis antigen (CTA) typically restricted to germline cells, where it inhibits Polycomb repressive complex 2 (PRC2)-mediated deposition of the repressive histone mark H3K27me3. Its aberrant expression in H3K27M mutation-negative diffuse midline gliomas (DMGs) and ependymomas leads to PRC2 inhibition and widespread epigenomic dysregulation. H3K27M-mutant gliomas are similarly characterized by global loss of H3K27me3 due to inhibition of PRC2 by the oncohistone H3K27M, positioning EZHIP as a functional epigenetic mimetic of this mutation.To explore its role in tumorigenesis, we overexpressed EZHIP in the H3K27 wild-type DMG cell lines that lack H3K27M mutations. EZHIP overexpression resulted in a significant reduction in H3K27me3 levels, indicating PRC2 inhibition. In a Nestin-TVA mouse model of DMG, the introduction of EZHIP into neural stem cells via the RCAS-EZHIP-P2A-PDGFβ vector led to the development of aggressive, high-grade diffuse gliomas. The tumorigenic potential of EZHIP was further supported by Western blot analysis of EZHIP-expressing neural stem cells, which revealed increased expression of stem cell markers, including Nanog, GFAP, and Nestin, consistent with a stem-like transcriptional state. RNA and ATAC sequencing of EZHIP-overexpressing DMG cell lines, compared to their isogenic wild-type counterparts, revealed a substantial increase in the number of upregulated genes typically repressed by PRC2, indicating widespread transcriptional activation associated with PRC2 inhibition. Comparative transcriptional analysis further showed that EZHIP-induced gene expression changes more closely resembled those in PRC2 knockout cells rather than in cells harboring the H3K27M mutation. Together, these findings highlight the complex role of EZHIP in disrupting PRC2-mediated gene silencing, promoting gliomagenesis, and altering the epigenomic landscape. While EZHIP and H3K27M mutations share some mechanistic similarities in inhibiting PRC2, key differences in their downstream effects suggest distinct contributions to tumor development. These findings advance our understanding of EZHIP-driven gliomas and reveal potential therapeutic vulnerabilities that could inform future treatments.

Abstract BACKGROUND Diffuse midline gliomas (DMGs) and posterior fossa group A (PFA) ependymomas are aggressive pediatric brain tumors. Most DMGs harbor histone H3.3K27M or G34R mutations, or exhibit EZHIP overexpression. With a median survival of just 9–15 months, DMGs are among the most lethal human cancers. PFAs, nearly universally overexpress EZHIP and have a 5-year survival rate of ~50%. Unlike K27M or EZHIP-overexpressing DMGs, G34R-mutant tumors retain H3K27me3. METHODS Using the RCAS/Nestin-TVA system, we introduced H3.3 K27M–, G34R–, or S31A–P2A–PDGFβ or H3.3WT–P2A–PDGFβ into neonatal mice. Controls included, histone mutants EZHIP alone and ΔEZHIP (PRC2-binding-deficient). Tumor bearing mice were assessed for incidence, grade, survival, and H3K27me3 and H3.3S31P status. RESULTS H3.3K27M, G34R, and S31A–P2A–PDGFβ induced high-grade gliomas with significantly reduced survival (p < 0.001). Mechanistically, G34R—like K27M—reduced mitotic Ser31 phosphorylation (S31P), causing chromosomal instability and impaired p53 checkpoint function. S31A directly modeled loss of S31P and similarly drove high-grade tumor formation. H3.3WT–P2A–PDGFβ induced only low-grade lesions. EZHIP–P2A–PDGFβ enhanced tumor grade but not survival; only EZHIP fully abolished H3K27me3. ΔEZHIP and all unlinked constructs were non-tumorigenic. CONCLUSION These models establish a robust platform to dissect glioma drivers and chromatin vulnerabilities. We previously showed that H3.3K27M reduces S31P, promoting chromosomal instability and impaired p53 function. We now demonstrate that H3.3G34R also reduces S31P and drives tumor formation in vivo. Direct loss of S31P via H3.3S31A is likewise sufficient for gliomagenesis. Although EZHIP does not alter S31P, it promotes genome instability through distinct mechanisms—disrupting PALB2–BRCA2-mediated DNA repair and mislocalizing TopoIIα, a process dependent on H3K27me3. H3.3K27M, G34R, and EZHIP thus converge on genome instability through distinct pathways. This platform supports mechanistic and therapeutic studies and is now being expanded using tetracycline-regulated expression to investigate tumor maintenance.

Abstract Diffuse midline and diffuse hemispheric glioma contain the driver mutations H3.3 K27M and G34R/V, respectively. These mutations alter histone methylations, suggesting that dysregulation of gene expression may contribute to tumor formation. We recently reported that these mutations also inhibit the phosphorylation of H3.3 at Serine 31. Importantly, H3.3 S31A mice developed high-grade gliomas – similar to those from G34R and K27M mice – indicating that loss of phosphoS31 promotes gliomagenesis. Chk1, a key kinase in DNA damage response and mitotic fidelity, phosphorylates H3.3 S31. H3.3 S31 is phosphorylated at the pericentromere in metaphase and is typically absent during interphase. Reduction of S31 phosphorylation by H3.3 K27M, G34R, G34V, or S31A was sufficient to induce mitotic errors leading to aneuploidy. Using CRISPR/Cas9, we generated novel H3.3 WT/WT and WT/S31A lines from glioma patient-derived SF-8628 (H3.3 WT/K27M) and KNS-42 (H3.3 WT/G34V) cells. Immunofluorescence of metaphase spreads revealed H3.3 phosphoS31 at the pericentromeres as expected. Interestingly, total Chk1 and active Chk1 (pS296) were significantly enriched at the pericentromeres of H3.3 K27M and G34V cells, as compared to H3.3 WT cells. While total ATR and active ATR (pT1989) were present at the pericentromere, their levels were unaltered by the H3.3 mutations. Surface plasmon resonance of purified proteins revealed that Chk1 has two binding affinities for H3.3 - likely corresponding to the previously identified Chk1 phosphorylation sites (T11 and S31) on H3.3. Chk1 binding at one site was unaffected by H3.3 K27M, while the second site showed an altered dissociation from H3.3 K27M. In summary, our data suggest that the low pericentromeric pS31 levels associated with H3.3 glioma mutations result from Chk1 recognizing the mutant histone, but without completing a phosphorylation cycle, becoming trapped at the pericentromere. This result suggests that H3.3 mutations, by trapping Chk1, may impede Chk1 activities beyond just reducing H3.3 phosphorylation.

Abstract BACKGROUND Posterior fossa group A ependymomas (PFAs) and diffuse midline gliomas (DMGs) are lethal pediatric brain tumors characterized by global loss of H3K27me3. This epigenetic disruption is caused by either EZHIP overexpression or H3.1/H3.3K27M mutations, which are mutually exclusive. EZHIP is aberrantly expressed in nearly all PFAs and approximately 10% of DMGs, while K27M mutations are found in ~80% of DMGs. In normal development, EZHIP expression is restricted to primordial germ cells during meiosis. We used human neural stem cells (NSCs), which give rise to ependymal and oligodendrocyte precursor lineages, to model EZHIP’s developmental effects. These cell types reflect the likely cells of origin for PFAs and DMGs, making them a biologically relevant platform to interrogate early tumorigenic mechanisms. METHODS AND RESULTS EZHIP was expressed in NSCs, astrocytes, and DMG lines. In NSCs, EZHIP induced strong repression of genome stability and mitotic fidelity genes, confirmed by gene set enrichment analysis (NES –2.91, p = 1e–10), and activated stem-like developmental programs. This repression was not observed in differentiated cells or tumor lines, indicating an NSC-specific vulnerability. The repressed targets were highly enriched for E2F-regulated genes (NES –3.61, p = 1e–10). EZHIP-expressing NSCs showed reduced DNA repair capacity, mitotic abnormalities, and continued proliferation. EZHIP is known to impair double-strand break repair by binding PALB2 and blocking BRCA2 recruitment, and to disrupt TopoIIα localization via H3K27me3 depletion, compounding mitotic stress in NSCs. CONCLUSIONS EZHIP disrupts genome maintenance and promotes transformation selectively in neural progenitors during a defined developmental window. Published data show that H3.3K27M induces profound transcriptional reprogramming in week 13 human NSCs, but these effects are diminished or absent by week 19, suggesting a shared, time-limited window of epigenetic vulnerability in both H3K27M- and EZHIP-driven tumors. Understanding this context-dependent repression has important implications for identifying stage-specific therapeutic vulnerabilities in pediatric gliomas.

Abstract Heterozygous H3.3 K27M mutations are central to the pathogenesis of diffuse midline gliomas (DMG), causing a global reduction in H3K27 trimethylation (H3K27me3) and subsequent epigenetic dysregulation. H3.3 contains a unique Serine 31 (S31) phosphorylation site crucial for mitotic checkpoint regulation. In DMG cells, reduced mitotic S31 phosphorylation correlates with chromosomal instability and aneuploidy. Using a mouse model, we demonstrate that both H3.3K27M and H3.3S31A mutations promote high-grade glioma development, whereas wild-type H3.3 does not. Importantly, S31A, which maintains normal K27me, suggests that S31 phosphorylation loss alone is oncogenic. The H3.3K27M mutation disrupts epigenetic regulation by inhibiting the Polycomb Repressive Complex 2 (PRC2), leading to decreased H3K27me3 and unintended gene induction. Concurrently, K27M exerts a compensatory repressive effect on the transcriptome, mediated by reduced S31 phosphorylation, independent of PRC2 activity. Using CRISPR-engineered isogenic DMG cells and analyzing the transcriptome and chromatin accessibility with RNA-seq and ATAC-seq, we found that the K27M mutation produces a balanced gene expression profile, while PRC2 loss primarily leads to gene induction. Notably, cells with S31A mutations also maintain this balanced profile, linking the PRC2-independent effect of K27M to reduced S31 phosphorylation. Conversely, cells with wild-type K27 but S31A mutation show a repressive genomic state, evidenced by an up/down gene ratio of 0.75, owing to functional PRC2. We further show that the genes uniquely dysregulated by K27M or S31A, independent of PRC2, overlap significantly, with dysregulated genes displaying altered chromatin accessibility. This suggests that the effect of K27M and S31A mutations on gene regulation is primarily due to changes in chromatin state. This complex interplay points to targeting K27M-associated pathways, particularly S31 phosphorylation, as a promising therapeutic avenue in DMG, a malignancy with fewer effective treatments and dismal prognosis.

Abstract Diffuse gliomas are pediatric high-grade gliomas that frequently harbor heterozygous histone H3.3 K27M, G34R, or G34V mutations. These mutations alter histone methylations, although the precise mechanism(s) of gliomagenesis remains uncertain. H3.3 contains a unique Serine at position 31 that is phosphorylated in prophase and dephosphorylated in anaphase during normal cell division. Previously we have reported that chromosome missegregation triggers H3.3 S31 phosphorylation along chromosome arms in ana/telophase leading to p53-induced cell cycle arrest. Here we show, in vitro and in vivo, that H3.3 K27M, G34R, and G34V suppress S31 phosphorylation by its mitotic kinase, Chk1. In diploid cells, overexpression of H3.3 K27M, G34R, G34V, or non-phosphorylatable S31A significantly increases the frequency of chromosome missegregation, while H3.3 K36M did not. This missegregation phenotype could be rescued by overexpression of K27M/S31E or G34R/S31E double mutants. Next, we used CRISPR/Cas9 gene editing to revert K27M or G34V patient-derived DMG cells to WT or replace the mutant allele with S31A. Reversion to WT increased metaphase S31 phosphorylation and reduced chromosome missegregations, but inserting an S31A mutation caused more mitotic defects. Following a mitotic error, newly formed aneuploid cells overexpressing K27M, G34R, or G34V suppress p53 levels and continue to divide, while WT overexpressing cells always arrest. To test the effects of diminished S31 phosphorylation on gliomagenesis, we utilized a novel mouse model of glioma. Expression of PDGFß linked to H3.3K27M, H3.3G34R, and H3.3S31A in neural stem cells in vivo promoted the development of high-grade gliomas in under 100 days, whereas PDFß-H3.3WT did not. Notably, the S31A tumors had histologically normal levels of K27me3 and K36me3. This work shows that a single glioma-associated point mutation in H3.3 can generate chromosomal instability, and which is a major contributing factor to H3.3 K27M and G34-altered glioma formation. Citation Format: Charles A. Day, Florina Grigore, Alyssa Langfald, Faruck Hakkim, David Daniels, Kevin Vaughan, James Robinson, Edward Hinchcliffe. Histone H3.3 K27M and G34R/V glioma driver mutations inhibit mitotic phosphorylation of H3.3 S31, driving tumor formation by inducing chromosomal instability [abstract]. In: Proceedings of the AACR Special Conference on Brain Cancer; 2023 Oct 19-22; Minneapolis, Minnesota. Philadelphia (PA): AACR; Cancer Res 2024;84(5 Suppl_1):Abstract nr B033.

Abstract Background H3K27-altered diffuse midline glioma (DMG) is the deadliest pediatric brain tumor; despite intensive research efforts, every clinical trial to date has failed. Is this because we are choosing the wrong drugs? Or are drug delivery and other pharmacokinetic variables at play? We hypothesize that the answer is likely a combination, where optimization may result in a much needed novel therapeutic approach. Methods We used in vitro drug screening, patient samples, and shRNA knockdown models to identify an upregulated target in DMG. A single small molecule protein kinase inhibitor with translational potential was selected for systemic and direct, loco-regional delivery to patient-derived xenografts (PDX) and genetically engineered mouse models (GEMM). Pharmacokinetic studies were conducted in non-tumor bearing rats. Results Aurora kinase (AK) inhibitors demonstrated strong antitumor effects in DMG drug screens. Additional in vitro studies corroborated the importance of AK to DMG survival. Systemic delivery of alisertib showed promise in subcutaneous PDX but not intracranial GEMM and PDX models. Repeated loco-regional drug administration into the tumor through convection-enhanced delivery (CED) was equally inefficacious, and pharmacokinetic studies revealed rapid clearance of alisertib from the brain. In an effort to increase the drug to tumor residence time, continuous CED over 7 days improved drug retention in the rodent brainstem and significantly extended survival in both orthotopic PDXs and GEMMs. Conclusions These studies provide evidence for increasing drug-tumor residence time of promising targeted therapies via extended CED as a valuable treatment strategy for DMG.

Histone post-translational modifications modulate gene expression through epigenetic gene regulation. The core histone H3 family members, H3.1, H3.2, and H3.3, play a central role in epigenetics. H3 histones can acquire many post-translational modifications, including the trimethylation of H3K27 (H3K27me3), which represses transcription. Triple methylation of H3K27 is performed by the histone methyltransferase Enhancer of Zeste Homologue 2 (EZH2), a component of the Polycomb Repressive Complex 2. Both global increases and decreases in H3K27me3 have been implicated in a wide range of cancer types. Here, we explore how opposing changes in H3K27me3 contribute to cancer by highlighting its role in two vastly different cancer types; (1) a form of glioma known as diffuse midline glioma H3K27-altered and (2) epithelial ovarian cancer. These two cancers vary widely in the age of onset, sex, associated mutations, and cell and organ type. However, both diffuse midline glioma and ovarian cancer have dysregulation of H3K27 methylation, triggering changes to the cancer cell transcriptome. In diffuse midline glioma, the loss of H3K27 methylation is a primary driving factor in tumorigenesis that promotes glial cell stemness and silences tumor suppressor genes. Conversely, hypermethylation of H3K27 occurs in late-stage epithelial ovarian cancer, which promotes tumor vascularization and tumor cell migration. By using each cancer type as a case study, this review emphasizes the importance of H3K27me3 in cancer while demonstrating that the mechanisms of histone H3 modification and subsequent gene expression changes are not a one-size-fits-all across cancer types.

Abstract The heterozygous H3.3 K27M mutation is found in the majority of pediatric High-Grade Glioma’s (pHGG). H3.3 K27M is thought to promote tumor formation through altered epigenetic gene regulation. However, it is unclear if epigenetic reprogramming alone is sufficient for glioma formation. H3.3 contains a unique Serine at position 31 which is phosphorylated in early mitosis (where it functions in chromosome segregation) and again in late M/early G1 when a mitotic error has occurred (triggering p53 expression). We found that pHGG cell lines harboring H3.3 K27M have significantly reduced S31 phosphorylation as compared to both pHGG and non-transformed H3.3 WT cells. When compared to WT H3.3 in an in vitro kinase assay, H3.3 K27M exhibits an ~60% reduction in S31 phosphorylation by Chk1, the mitotic S31 kinase. In normal diploid cells, expression of K27M or non-phosphorylatable S31A mutant significantly increased chromosome missegregation. Yet expressing a phosphomimetic double mutant (K27M/S31E) did not significantly alter the rate of chromosome mis-segregation compared to WT. Furthermore, patient-derived pHGG lines harboring K27M have significantly higher rates of chromosome mis-segregation compared to an H3.3 WT pHGG line or H3.3 WT non-transformed human cells. We used CRISPR gene editing to remove the H3.3 K27M allele or replace it with WT H3.3. In both cases, loss of K27M elevated pS31 levels and reduced chromosome mis-segregations. Yet, when K27M was replaced with S31A, the mis-segregation rate remained similar to the parental K27M cells. In normal cells, chromosome missegregation stimulates p53-dependent cell cycle arrest in G1 to prevent the proliferation of aneuploid daughters. However, cells expressing H3.3 K27M or S31A failed to arrest following missegregation - despite having WT p53. To determine if the induction of chromosomal instability would alter tumor formation, we expressed H3.3 WT, K27M or S31A in combination with PDGFβ in a glioma mouse model. Expression of WT H3.3 failed to generate high-grade tumors. But 66% of mice expressing K27M and 93% of those expressing H3.3 S31A developed diffuse high-grade brain tumors by 100 days. Our results suggest that loss of phospho-S31 alone is oncogenic because H3.3 S31A-expressing cells are WT for K27me3. Our results demonstrate that H3.3 K27M inhibits Ser31 phosphorylation both in vitro and in vivo, leading to both chromosome missegregation and loss of subsequent G1 arrest - thus creating diffuse midline gliomas with both dynamic, complex karyotypes and epigenetic reprogramming. Citation Format: Charles Day, Florina Grigore, Alyssa Langfald, David J. Daniels, James Robinson, Edward Hinchcliffe. Inhibition of H3.3 S31 phosphorylation by the pediatric glioma driver mutation, H3.3 K27M, results in chromosomal instability, loss of p53 regulation, and tumorigenesis [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 705.

Abstract Pediatric high-grade gliomas (pHGG) are among the most lethal of all human cancers. Histone H3.3 G34R/V mutations are an early event in these tumors and show reduced H3.3 K36 trimethylation; implicating epigenetic dysregulation in tumorigenesis. Here we present evidence that H3.3 G34R/V mutations promote tumor formation via the induction of chromosomal instability (CIN). Pericentromeric H3.3 is phosphorylated at S31 by Chk1 during mitosis. We observed that the H3.3 G34R mutation reduced Chk1 phosphorylation of H3.3 S31 by >90% in vitro. Furthermore, H3.3 G34 mutant cells have reduced pericentromeric H3.3 S31 phosphorylation in mitosis compared to WT H3.3 cell lines. H3.3 G34 mutant pHGG cells also have significantly elevated rates of CIN as compared to H3.3 WT cells. Overexpression of H3.3 G34R, G34V or non-phosphorylatable S31A in H3.3 WT, diploid cells caused a significant increase in CIN, but H3.3 K36M overexpression had no effect on chromosome segregation. These studies demonstrate that H3.3 G34R/V mutations are sufficient to induce CIN in normal, diploid cells. To determine if this process contributes to tumorigenesis, we used RCAS Nestin-TVA mice to overexpress H3.3 WT, G34R, or S31A – P2A-linked to PDGFB in glial precursor cells of newborn mice. Over 100 days, S31A and G34R mice had drastically reduced survival (averaging 77, 81, and 100 days for S31A, G34R, and WT). Furthermore, most G34R and S31A mice developed HGG, while H3.3 WT mice remained tumor-free. Our work implicates CIN as a driver of H3.3 G34 mutant pHGG formation.