Epigenetic alterations are crucial in determining the classification and driving the progression of gliomas. The genetic alteration status of isocitrate dehydrogenase (IDH) is a critical biomarker that distinguishes different glioma subtypes. As a crucial metabolic enzyme, IDH is mainly involved in the cellular energy metabolism process, especially the tricarboxylic acid cycle (TCA cycle). It plays an important role in maintaining normal physiological functions. In glioma cells, IDH mutation is a common molecular alteration, particularly the IDH1-R132H mutation. This mutation leads to changes in enzyme activity, converting α-ketoglutarate (α-KG) into (R)-2-hydroxyglutarate [(R)-2HG]. This metabolite can interfere with normal cellular metabolism and epigenetic regulation, promoting the occurrence and development of tumors (20) ( Figure 1 ).

Interestingly, IDH-mutant gliomas display a CpG island methylator phenotype (G-CIMP), which is characterized by hypermethylation in CpG-rich promoter regions. This epigenetic alteration is associated with distinct clinical outcomes, with the “high” G-CIMP subgroup generally having a better prognosis (21). The dynamic interplay between IDH mutations and epigenetic modifications is further evidenced by the role of G-CIMP in driving the differentiation and aggressiveness of glioma subtypes. For instance, oligodendrogliomas harboring IDH mutation alongside 1p/19q codeletion show a more promising prognosis, whereas glioblastomas with wild-type IDH are associated with a poorer outcome. Compared with untreated patients, patients who experience relapse and exhibit cases with the absence of O(6)-methylguanine-DNA methyltransferase (MGMT) promoter methylation often exhibit a greater load of genetic mutations, indicating that this methylation phenotype may lead to the chemoresistance of tumors (22, 23). These observations underscore the clinical relevance of epigenetic profiling in glioma subtype classification and prognostication.

Even though extensive genomic hypermethylation has been demonstrated to impede the process of cellular differentiation, the role of specific methylation events in tumor progression has been emphasized. This phenomenon involves the transcriptional silencing of tumor suppressor genes (including BRCA1/BRCA2, TP53BP1, SMAD4) and the abnormal activation of oncogenes (such as PDGFRA, PIK3CA, and BRAF), thereby facilitating tumor initiation and progression. These genes are collectively referred to as “methylation-dependent oncogenic drivers” (24). IDH-mutated low-grade gliomas (LGGs) represent a subgroup of brain tumors exhibiting unique molecular features, typically characterized by slow growth and relatively favorable prognosis. However, they carry the potential risk of malignant progression. The progression from low-grade to high-grade malignancy is propelled by the accretion of genetic mutations, epigenetic modifications, and shifts in the tumor microenvironment (TME). Therefore, early identification of the IDH molecular subtypes in glioma patients and targeted treatment have significant clinical implications for patient outcomes. In a recent investigation, Wu and his team constructed a dual-phase framework by probing single-cell RNA sequencing and chromatin accessibility profiles across various glioma grades. During the initial phase, gliomagenesis is driven by oligodendrocyte precursor-like cells in conjunction with epigenetic reprogramming, which deactivate tumor suppressor genes such as CDKN2A and trigger oncogenes like PDGFRA (25). As the disease progresses, tumor expansion is fueled by an increased population of proliferative neural precursor-like cells. Genetic aberrations, including PDGFRA, MYCN, and CDK4 amplifications alongside CDKN2A/B deletions, facilitate tumor advancement. This research underscores the fluctuating impact of IFN signaling during disease progression. In IDH-mutated low-grade gliomas, IFN signaling undergoes epigenetic hypermethylation-mediated suppression, a process reversible by MGMT inhibitors or IDH antagonists, reactivating the IFN pathway. In contrast, high-grade gliomas evade IFN signaling through genetic deletions of IFN genes. These findings indicate a shift from epigenetic to genetic control in glioma progression. Lately, the US Food and Drug Administration (FDA) greenlit vorasidenib, a brain-permeable small molecule targeting mutant IDH1/2 proteins, marking a pivotal change in the therapeutic approach for IDH-mutant gliomas (26).

Epigenetic alterations significantly impact the transcriptional landscape of glioma cells, contributing to their malignant behavior. Research has shown that mIDH1 expression induces transcriptional silencing of the PD-L1-encoding gene CD274 via DNA methylation mechanisms (27, 28). Upon treatment with mIDH1 inhibitors, mIDH1 glioma cells exhibit reduced DNA methylation in the CD274 regulatory region. Concurrently, tumor cell-mediated high PD-L1 expression contributes to T cell exhaustion. Therefore, in gliomas with mIDH1 mutations, the concentration of PD-L1, which restricts T-cell functionality, is notably diminished. A separate study revealed that the absence of ATRX-a key epigenetic regulator linked to adult and pediatric gliomagenesis-epigenetically drives the production of PD-L1 alongside multiple immunosuppressive cytokines. This cascade of events ultimately initiates an immune-tolerance-promoting process in ATRX-mutated gliomas (29–31).

Explorations into the influence of histone modifications on TME modulation are insufficiently documented. Emerging findings highlight that RACK7 (encoded by ZMYND8) displays intensified interaction with the H3.3-G34R histone variant, a mutation linked to pHGG pathogenesis. This augmented binding event results in silencing of specific genomic loci, including CIITA, which plays a pivotal role in governing MHC class II molecule expression (32). The PRC2 complex, a group of histone methyltransferases, plays an essential part in regulating epigenetic silencing, with its activity being modifiable in gliomas. The expression levels of PRC2 proteins are associated with a poor prognosis (33). Recent investigations demonstrate that PRC2-dependent chromatin silencing contributes to immunosuppressive TMEs by inhibiting the transcription of immunostimulatory cytokines in neoplastic cells. Therefore, tumors with PRC2 mutations respond better to ICIs, providing the possibility of enhancing immunotherapy response by epigenetically and pharmacologically restoring PRC2 methyltransferase activity (34). Besides, in mIDH1-mutant gliomas, epigenetic modulation of the Notch signaling cascade is achieved through DNA methylation at CpG dinucleotides within the delta-like ligand 3 (DLL3) genomic locus. As an inhibitory Notch ligand, DLL3 expression exhibits a positive correlation with survival outcomes in mIDH1 gliomas (35), Glioma cases with elevated DLL3 levels demonstrate enhanced immune cell infiltration, suggesting a functional link between Notch pathway activation and immune responsiveness in these neoplasms.

Over the past few decades, it has been revealed that noncoding RNAs, such as microRNAs(miRNAs) and long noncoding RNAs, are crucial in gliomas, where they function as epigenetic regulators among other roles (36), Genomic profiling reveals 27 long non-coding RNAs (lncRNAs) are upregulated and 198 are downregulated in glioblastoma (GBM), underscoring their critical involvement in tumor pathogenesis. Recent findings highlight the role of lncRNAs in modulating GBM cell metabolism. For example, the lncRNA TP53TG1 promotes cell migration and proliferation under glucose-deprived conditions by regulating gene expression of PKM2 and IDH1 in glioma cell cultures (37). Additional research demonstrates that lncRNAs influence glioma biology, progression, and therapeutic responsiveness. For example, HOTAIRM1, an epigenetic modulator, associates with transcriptional start sites to regulate gene expression. LncSNHG6 and lncRNA ZFAT-AS1 facilitate Histone H3 Lysine 27 trimethylation (H3K27me3) deposition, leading to transcriptional silencing. MiRNAs as endogenous regulatory RNAs, function by inhibiting mRNA translation. Depletion of miRNAs that govern key mRNAs- including miR-31 (which targets CDKN2A/B) and miR-34a (involved in EGFR level regulation) (38). T-can drive oncogenic pathways. Variations in miRNA expression across molecular subtypes suggest their role in shaping glioma heterogeneity through transcriptional subtype transitions. MiRNA dysregulates major pathways (TGF-β, PI3K/AKT, EGFR, Notch) in glioma progression likewise. Exosomal miR-148a is related to down-regulated CADM1 expression in GBM samples (39). The antagonistic effect of miR-148a reduces the level of p-STAT3 and subsequently up-regulates the level of the tumor suppressor gene CADM1. In addition, NcRNAs act as intercellular signals; glioma cell-secreted miRNA and lncRNA affect TME behavior. For instance, glioma cell-secreted lncRNA-ATB suppresses miR-204-3p in astrocytes, promoting cell migration. Exosome-secreted lncRNAs have a paracrine effect aiding adaptation and resistance. The ncRNA-mediated communication between glioma cells and TME has therapeutic potential, but its role in TME epigenetic changes, especially in immune cells, requires further study. Furthermore, findings from current research indicate that dysregulated expression of circular RNAs (circRNAs) correlates with glioma initiation and progression (40). Transcriptomic profiling via microarray identified 91 circRNAs with altered expression patterns in glioma tissues compared to matched non-tumor adjacent samples. Notably, elevated expression of circ_0037655 has been shown to promote tumor cell viability and invasive capacity (41, 42). The involvement of noncoding RNAs-including lncRNAs, miRNAs, and circRNAs—in shaping the TME and modulating immune cell functionality broadens therapeutic possibilities. Therapeutic approaches targeting these ncRNAs or their associated regulatory pathways may offer innovative strategies to enhance immunotherapeutic efficacy or circumvent treatment resistance. Nevertheless, the intricate interplay among ncRNAs, TME components, and immune cells requires additional investigation to fully harness their potential as therapeutic targets.

Interestingly, our recent work has revealed a class of miRNAs localized in the nucleus and whose genomic locations highly overlap with enhancers (named NamiRNA, Nuclear Activating miRNA) play a role in enhancing the sensitivity of GBM to temozolomide by activating enhancer activity to positively regulate target genes and induce nucleolar stress ( Figure 2 ). This further suggests the clinical application potential of noncoding RNAs in the treatment of gliomas.

Glioma stem cells (GSCs) are pivotal in driving glioma initiation, progression, and relapse. Their intrinsic resistance to standard therapies and capacity for tumor re-formation pose substantial challenges in glioma management. Emerging research has underscored the complex interplay between epigenetic mechanisms and the preservation of GSC characteristics (43, 44). Epigenetic alterations, such as DNA methylation, histone alterations, and ncRNA-dependent regulation, are central to sustaining the stem-like properties of GSCs (45). For instance, DNA hypermethylation at specific loci can silence tumor suppressor genes, allowing GSCs to evade cellular senescence and apoptosis. Conversely, hypomethylation can activate oncogenes, promoting proliferation and self-renewal (46). Modifications to histones, with a focus on acetylation and methylation, are vital for the upkeep of GSCs. Histone-modifying enzymes, including histone acetyltransferases (HATs) and histone deacetylases (HDACs) manage how accessible chromatin is to transcription factors, which in turn shapes gene expression. High levels of HDAC activity have been observed in GSCs, promoting a repressive chromatin state that favors the expression of stemness-related genes. Inhibitors of HDACs have shown promise in inducing differentiation and reducing the stemness of GSCs, making them more susceptible to conventional therapies. Besides, enhancers are capable of recruiting a large number of transcription factors and coactivators, Histone H3 lysine 27 acetylation (H3K27ac) functions as an indicator of active enhancers and significantly impacts the regulation of genes associated with cell identity and disease. Gimple etc. analyzed 10 glioblastoma specimens obtained via surgical resection and 15 non-neoplastic brain tissue samples through chromatin immunoprecipitation sequencing (ChIP-seq), and screened out the genes WSCD1, ELOVL2 and KLHDC8A that are regulated by glioma stem cell (GSC)-related enhancers, providing new targets for the treatment of gliomas (47) ( Figure 3 ). In addition, non-coding RNAs, such as miRNAs and lncRNAs, are increasingly recognized as pivotal modulators of GSC phenotypes. MiRNAs may act as oncogenic or tumor suppressive regulators by targeting genetic loci associated with stem cell maintenance, differentiation, and self-renewal pathways. Analogously, lncRNAs have been demonstrated to modulate gene expression through epigenetic mechanisms, influencing chromatin structure and transcription factor activity. A recent study has indicated that LncRNA INHEG regulates the 2’-O-methylation of rRNA and promotes the translation of mRNA to facilitate the renewal and growth of GSCs (48).

The interplay between epigenetic modifications and GSCs underscores the importance of developing targeted therapies that can disrupt these regulatory networks. By targeting epigenetic modifiers, it may be possible to induce differentiation of GSCs, reducing their tumor-initiating potential and enhancing effectiveness of traditional treatments. Additional investigations into the epigenetic control of GSCs may facilitate the creation of innovative therapeutic approaches to enhance glioma patient outcomes.

Epigenetic control influences the dysfunction of antigen-presenting mechanisms in neoplastic cells, allowing tumors to avoid T-cell recognition and fostering immune evasion (51). For the presentation of self- and tumor-specific peptides to CD8 T cells, tumor cell proteins must undergo proteasomal digestion to produce short oligopeptides. The antigen processing-associated transporters 1 and 2 (TAP1 and TAP2) assemble into a heterodimer, facilitating the translocation of these peptides from the cytosol to the endoplasmic reticulum (ER) (52, 53). Within the ER, antigen peptides are incorporated into nascent major histocompatibility complex class I (MHC-I) molecules with the aid of chaperone proteins (54, 55). The MHC-I complex responsible for tumor antigen presentation consists of two polypeptide subunits-human leukocyte antigens (HLA) and β2-microglobulin (B2M)-and is translocated to the cell surface for antigen presentation. Neoplastic cell MHC-I expression can be suppressed by both DNMT and HDAC enzymes. Administration of DNA methyltransferase inhibitors (DNMTi) or histone deacetylase inhibitors (HDACi) to tumor cells can restore MHC-I surface expression (56). In certain instances, MHC-I deficiency stems from the epigenetic repression of auxiliary genes in the antigen-presentation pathway, including B2M, TAP-1, and TAP-2. Intervention with DNMTi in tumor cell models or clinical settings can upregulate gene expression essential for antigen presentation (57).

Currently, in the context of developing therapies against tumor antigens in gliomas, cancer-specific somatic mutation-derived antigens serve as optimal immunotherapeutic targets from multiple perspectives. Their expression is predominantly restricted to neoplastic cells, thereby significantly mitigating the risk of off-tumor toxicity. For example, in children with diffuse midline glioma (DMG), the H3F3A locus, which encodes histone H3.3, harbors a frequently recurring point mutation (58). The K27M substitution at amino acid position 27 (lysine to methionine) reduces H3K27 trimethylation, causing global demethylation and abnormal gene expression by inhibiting PRC2 activity. This mutation is an attractive immunotherapy target. Analogous to the IDH1 R132H alteration in adult gliomas, the H3.3 K27M mutation exhibits uniform distribution across tumor tissues. Recently, Pavlina Chuntova et al. documented that blocking D-2HG results in the increased expression of proinflammatory genes in a new HLA-A2/HLA-DR1 transgenic mouse model expressing IDH1R132H glioma (59). Meanwhile, Gerard L Brien et al. identified the oncogenic activity in human hindbrain neural stem cells, the histone H3.3-K27M modification arises from the simultaneous impairment of both PRC2 activity and enhancer functionality (60).

Furthermore, in addition to affecting transcription, epigenetic disorders exacerbate genetic instability in gliomas by altering chromatin structure: mutations in histones and ATRX lead to genetic instability due to abnormal deposition of histone marks, while mutant IDH epigenetically enhances DNA damage response signaling. This genetic instability is a key driver of neoantigen production-it not only induces cytoplasmic extrachromosomal DNA to activate the cGAS/STING axis, but also directly causes chromosomal aberrations, prompting mutated proteins to express neoantigens; the immune activation or tolerance triggered by neoantigens is determined by the tumor microenvironment (TME). Epigenetic regulation mediated by STING promoter methylation can modulate glioma immune responses, and this process is reversible by MGMT inhibitors. Mutations in the H3-G34 locus in pediatric high-grade gliomas (pHGG) also induce genomic instability, activate the cGAS/STING axis, enhance immune activation, and improve the efficacy of DNA damage-based therapies due to abnormal neoantigen expression (61, 62).

Cancer cells can emit signaling molecules that are detected by immune cells at a distance, thereby triggering or inhibiting immune reactions. Glioma cells produce various immunomodulatory factors, such as IL-1β, IL-6, TGF-β, and IL-8. In glioma stem/progenitor-like cells, mIDH1 mediates epigenetic upregulation of G-CSF expression, thereby promoting the remodeling of myeloid cells in the mIDH1-associated TME. Elevated levels of G-CSF stimulate the proliferation of neutrophil precursors and neutrophils, concurrently attenuating the immunosuppressive traits of polymorphonuclear myeloid-derived suppressor cells within the mIDH1-driven tumor microenvironment (22). The epigenetic regulation of tumor cell differentiation impacts the TME. Epigenetic alterations in gliomas, such as H3 mutations in pediatric cases and mIDH1 in adult patients, lead to a halt in cellular differentiation, maintaining cells in a stem-like state. Cells resembling cancer stem cells exhibit reduced immunogenicity and evade immune detection through mechanisms like the downregulation of MHC transcription and the induction of quiescence. A recent investigation revealed that GSCs cultivated in immunocompetent hosts adapt epigenetically and secrete immunosuppressive cytokines (43, 63).

Furthermore, Glioma cells secrete immunomodulatory cytokines like IL-1β, IL-6, TGF-β, and IL-8 to activate or suppress immune response. Studies show GSCs undergo epigenetic adaptation, secreting immunosuppressive cytokines. Additionally, adapted GSCs upregulate interferon regulatory factor 8 (IRF8) secretion. IRF8, a cytokine usually in myeloid cells, may govern myeloid cell lineages and macrophage polarization (64). This implies that epigenetic mechanisms coordinate the functional repatterning of glioma cell populations, which in turn leads to alterations in immune cells and fosters the development of a TME favorable for tumor growth. Independent research has demonstrated that PRC2 constitutes a complex of histone methyltransferases responsible for epigenetic repression. It is hypothesized that PRC2-dependent chromatin repression facilitates immune evasion by suppressing the production of immunostimulatory cytokines in neoplastic cells (65) ( Figure 4 ). Additionally, in gliomas, epigenetic reprogramming of the PI3K/AKT/mTOR signaling axis drives a metabolic shift from oxidative phosphorylation to aerobic glycolysis, and this metabolic reprogramming serves as a core hub for regulating the secretory factor network in the immune microenvironment (66). It exerts its effects through multiple interconnected mechanisms: first, it promotes the massive release of metabolites such as lactate, which not only directly impairs immune cell function and shapes a hypoxia-associated immunosuppressive tumor microenvironment (TME) but also acts as signaling molecules to regulate the expression and secretion of secretory factors; second, it reshapes immune homeostasis via secretory factors through pathways including upregulating TGF-β secretion, inhibiting monocyte differentiation into dendritic cells, and inducing the production of pro-oncogenic cytokines; additionally, metabolic changes in the TME specifically modulate the secretory function of astrocytes, prompting them to release cholesterol to facilitate glioma growth and recruit immunosuppressive macrophages (67). Collectively, these processes reveal that metabolic reprogramming, by systematically regulating the production, release, and function of secretory factors, emerges as a key driver influencing tumor progression and the immune microenvironment.

The interaction between PD-L1 and the immune checkpoint receptor programmed death-1 (PD-1) on T cells conveys inhibitory signals to T cells. This results in the inhibition of T cell proliferation, cytokine generation, and cytotoxic function (68). This phenotype is referred to as “T cell exhaustion”. Consequently, T cells fail to efficiently identify and eliminate tumor cells. This enables tumor cells to dodge the immune system’s monitoring and elimination processes, thereby attaining immune evasion. Epigenetic modulation incontrovertibly assumes a pivotal role in PD-L1 expression. PD-L1 promoter methylation demonstrates an inverse association with PD-L1 transcriptional activity and clinical outcomes in patients (69). Furthermore, HDAC6, which acts as a histone acetylation “eraser,” lysine methyltransferase 2A (KMT2A), a methylation “writer,” and the bromodomain-containing BET family protein BRD4, an acetylation “reader,” each elevate PD-L1 immunoreactivity across melanomas, pancreatic carcinomas, and ovarian neoplasms. As such, targeting these proteins with small-molecule inhibitors can effectively reduce PD-L1 levels and enhance the body’s anti-tumor immune response (70). Additionally, nivolumab is a fully human immunoglobulin G4 monoclonal antibody targeting the programmed death-1 (PD-1) immune checkpoint receptor. Results from a phase III clinical trial (NCT02017717) comparing the efficacy of Nivolumab monotherapy versus bevacizumab in improving survival rates of patients with recurrent glioblastoma showed that the subgroup of glioblastoma patients with methylated MGMT promoters and no baseline corticosteroid dependence may be most likely to benefit from immune checkpoint inhibition (71, 72).

The immune landscape of gliomas is modulated by IDH1/2 genetic alterations, notably, PD-L1 expression is substantially decreased in mutated tumors, providing a mechanistic basis for immune checkpoint inhibitors (ICIs) in IDH1/2 wild-type (wt) patients. Beyond DNA methylation, additional epigenetic regulators may also modulate immune checkpoint (IC) expression in gliomas (73). MiRNAs directly modulate the expression of immune checkpoint regulators, including CTLA-4, PD-1, and PD-L1, via distinct miRNA species. For instance, miR-155 specifically targets CTLA-4, miR-138 fine-tunes both CTLA-4 and PD-1, miR-424 governs PD-L1 and CD80, miR-28 impacts PD-1, while miR-34a, miR-200, miR-513, and miR-138-5p collectively regulate PD-L1 expression (74, 75). Furthermore, overlapping or distinct miRNAs modulate the production of cytokines such as IFN-γ or transcription factors functioning as either activators or repressors of immune checkpoint pathways thereby establishing a redundant and highly intricate regulatory network.

Tumor-infiltrating lymphocytes (TILs) exhibit robust antitumor immune capabilities and represent critical components of immunotherapeutic strategies. In the glioma TME, regulatory T cells (Tregs), CD4⁺ helper T cells (Th cells), and CD8⁺ cytotoxic T cells undergo infiltration. Tregs potently inhibit the functions of antitumor immune cells while elevating the abundance of other immunosuppressive cell types. Helper T cells-particularly Th1 subsets—and CD8⁺ T cells mediate antitumor immunity by respectively enhancing pro-inflammatory responses and executing tumor cell cytotoxicity (76). The efficacy of their antitumor actions is impeded due to limited tumor infiltration and the immunosuppressive milieu of glioma-derived TME. The primary driver of immune suppression is the glioma cells themselves, MDSCs, Tregs, and Tumor-Associated Microglia and Macrophages (TAMMs) through PD-L1 expression and TGF-β/IL-10 secretion. Immune-modulatory molecules trigger T cell functional impairment, encompassing anergy and exhaustion phenotypes. Within exhausted Th and CD8⁺ T cell subsets, proliferative capacity is diminished, and secretion of effector cytokines (e.g., IL-2, TNF-α, IFN-γ) is significantly reduced. Mechanistically, T cell exhaustion is typified by enhanced chromatin openness at immunosuppressive gene loci (PDCD1, CTLA4, LAG3, ENTPD1) and reduced accessibility at cell differentiation gene loci (IL7R, TCF7, LEF1), accompanied by corresponding upregulation and downregulation of transcriptional activity (77). Epigenetic mechanisms have been shown to influence cell-fate determinations in lymphocyte ontogeny (78, 79). Chromatin organization and histone-modifying enzymes represent critical regulators of dendritic cell (DC) functions. For instance, the histone H3K4 demethylase KDM5B exerts negative regulation on the activation of bone marrow-derived dendritic cells, leading to suboptimal T cell responses. Conversely, the methylated DNA “reader” methyl-CpG-binding domain protein 2 (MBD2) is indispensable for the phenotypic activation of dendritic cells and their capacity to prime T cell responses. Moreover, in-depth analysis of enhancer signatures like H3K4me1 and H3K27Ac has demonstrated that enhancers undergo substantial dynamics during T cell activation (80).

RNA modifications are chemical marks on the bases or riboses of RNA molecules. These modifications are widely present in various types of RNA (such as mRNA, tRNA, rRNA, lncRNA, etc.). To date, more than 150 different modifications have been identified, including methylation (e.g., m⁶A, m⁵C, m¹A), pseudouridylation (Ψ), acetylation, and so on. RNA modifications do not alter the sequence of RNA, but can dynamically regulate gene expression by affecting the structure, stability, translation efficiency of RNA and its interaction with other molecules. RNA modifications play a crucial role in the process of tumor immune evasion, reshaping the tumor microenvironment and interfering with immune surveillance mechanisms through multi-dimensional regulation (81, 82). For instance, in GBM, the long non-coding RNA CASC9 is involved in tumor progression by promoting glycolytic metabolism and is significantly associated with poor prognosis in patients. The underlying molecular mechanism involves regulation by N6-methyladenosine (m⁶A) modification: the m⁶A “reader” IGF2BP2 can recognize and bind to methylated CASC9, maintaining its function by enhancing the stability of its transcript. The molecular complex formed by these two further acts on the methylated region of hexokinase 2 (HK2) mRNA, directly driving glycolysis in GBM cells by improving the stability of this mRNA. The accumulation of large amounts of lactic acid resulting from enhanced glycolysis acidifies the tumor microenvironment (TME), thereby inducing immunosuppressive effects: the acidic environment not only impairs the proliferation, activation, and cytotoxic functions of immune effector cells such as T lymphocytes and natural killer (NK) cells, but also promotes the differentiation and expansion of regulatory T cells (Tregs). Ultimately, this inhibits systemic immune responses and facilitates tumor immune evasion (83, 84). A separate investigation revealed that targeting the m6A demethylase ALKBH5 disrupts YTHDF2-mediated stability of ZDHHC3 mRNA. This mechanism inhibits PD-L1 expression in gliomas by promoting PD-L1 degradation (85). Another study on immune evasion in gliomas demonstrated that the 28S rRNA methyltransferase NSUN5, via its cysteine 359 (C359)-dependent methyltransferase activity, directly binds to the chromatin-associated RNA (caRNA) of CTNNB1 and deposits 5-methylcytosine (m5C). This modification is oxidized to 5-hydroxymethylcytosine (5hmC) by TET2, which relies on TET2’s binding affinity for Fe²⁺ and α-KG. Subsequently, 5hmC is recognized by RBFOX2, a 5hmC-specific reader, which promotes the degradation of caRNA. This process downregulates β-catenin, enhances the phagocytic activity of tumor-associated macrophages (TAMs), and impairs the tumor’s ability to evade immunity (86).

The aim of mutant IDH inhibitors is to block the synthesis of the oncometabolite 2-hydroxyglutarate (2HG). As a rival inhibitor of α-ketoglutarate-dependent dioxygenases, 2HG regulation can be achieved by suppressing its production, thereby reversing DNA hypermethylation and promoting the differentiation of mIDH1 glioma cells. Several mIDH1 inhibitors have shown effectiveness in both laboratory and animal models. Combining current first-line therapies (radiotherapy and temozolomide) with mIDH1 inhibitors and PD-L1-targeted immune checkpoint inhibitors (ICI) has enhanced tumor shrinkage in mIDH1 glioma-bearing mice, alleviated T cell exhaustion, and supported the formation of memory CD8+T cells. Presently, multiple clinical trials are assessing mIDH1 inhibitors for glioma treatment, although these remain in initial stages primarily focused on evaluating safety and the ability to reduce 2HG accumulation (87).

Enhancer of zeste homolog 2 (EZH2), a pivotal histone methyltransferase in the PRC2, undergoes alterations in gliomas that lead to both gain- and loss-of-function mutations. These aberrations in EZH2 expression impact gene regulation by interacting with promoters and modulating methylation patterns, thereby acting as an oncogenic driver. EZH2-dependent silencing of TSGs drives oncogenic proliferation, aggressive invasion, and refractory resistance in cancer. The use of EZH2 inhibitors (EZH2is) has been shown to upregulate p16 expression and effectively curb the progression of gliomas. Research indicates that EZH2 serves as a pivotal regulator in regulating cancer cell immune response and mediates escape by downregulating immune activation genes, upregulating checkpoints, and creating an immunosuppressive TME. Tazemetostat, an EZH2i, is being tested in pediatric gliomas. However, A recent investigation indicates that it might influence the proliferation of primary Histone H3 Lysine 27 to Methionine Mutation (H3K27M)-mutant glioma cells occurs when functional p16INK4A is expressed in GBM cells harboring wild-type H3 and IDH (88). Short-term reduction of EZH2 correlates with decreased cell proliferation. However, recent findings show that extended EZH2 inhibition might induce a cell fate transition, promoting both proliferation and DNA damage repair mechanisms, which ultimately drives tumor progression (89).

Epigenetic silencing by DNMTs suppresses gene expression via CpG island methylation. DNMTis reverse this, reactivating tumor suppressors and inducing anti-tumor effects. In gliomas, DNMTis upregulate MHC class I presentation, enhancing CD8+ T cell recognition. By preventing DNMT-driven epigenetic reprogramming, DNMTis also counteract T cell exhaustion. Genetic knockout of DNMT3A in CAR T cells mirrors this effect, bolstering anti-tumor immunity (50, 90). Preclinical research has shown DNMTi exhibit efficacy using in vitro and in vivo model systems of IDH-mutant (IDHmt) gliomas. Nevertheless, this preclinical promise has not yet led to successful clinical outcomes for glioma patients treated with DNMTi, potentially due to the S-phase specificity and relatively brief half-life of agents like 5-azacytidine and decitabine. Nevertheless, whether gliomas require general demethylation remains controversial, activation of proto-oncogenes coupled with hypomethylation-mediated reactivation of the DNA repair gene O⁶- methylguanine-DNA MGMT might confer resistance of glioblastoma to alkylating agents. Notably, low-dose demethylating agents can modulate immunity and potentially trigger an innate immune response through reactivation of retroviruses (91).

Certain frequent epigenetic alterations in neoplastic cells involve dysregulation of histone modifications within oncogene/tumor suppressor gene regulatory domains. Aberrant HDAC expression in cancer cells modifies the cell cycle and initiates tumorigenesis. HDAC expression patterns correlate with glioma malignancy grading; class II and IV HDAC isoforms exhibit reduced expression in GBM compared to low-grade astrocytomas. HDAC1 is excessively expressed in multiple glioma subtypes and associated with diminished overall survival. HDACis modulate antitumor immune responses by facilitating T cell chemokine production, augmenting PD-1-targeted immunotherapy effectiveness, and increasing PD-L1 and HLA-DR surface expression on tumor cells, suggesting a potential cooperative effect between HDACis and ICIs for glioma therapy. Vorinostat, evaluated in a phase II clinical trial for recurrent GBM, showed favorable tolerability as monotherapy and influenced GBM-relevant signaling pathways. Nevertheless, its clinical application is constrained by toxicity and suboptimal efficacy, potentially attributed to inadequate BBB penetration. Integrating HDACis with other therapeutic approaches or enhancing BBB permeability could improve treatment outcomes (92–94).

BRD4 function can be effectively suppressed by inhibitors or degraders. As small-molecule compounds, BRD4 inhibitors have the potential to augment cancer therapy by mimicking acetyl-lysine residues, thereby binding to BET proteins and ultimately interfering with the interaction between BET proteins, acetylated histones, and transcription factors (95). BET proteins play a critical role in driving high-level oncogene expression. BETi can attenuate oncogene transcription by diminishing super-enhancer activity, which consists of large aggregations of transcriptional enhancers that govern gene expression in cell identity and pathological conditions like cancer. Using in vivo RNA interference screening to target chromatin regulators critical for GBM cell viability, BRD4 emerged as a prominent target. Preclinical investigations using orthotopic mouse glioblastoma xenograft models have demonstrated the efficacy of multiple BETi compounds, such as JQ1, I-BET151, and OTX015 (96). Primary glioma cells with IDH mutation (IDHmt) demonstrate significant sensitivity to BET inhibitors JQ1 and GS-626510, exhibiting half-maximal inhibitory concentrations (IC50) 1,000 times lower than that of Temozolomide. A 2015 phase IIa clinical trial evaluating OTX015 for dose optimization in recurrent glioblastoma patients was halted due to insufficient therapeutic response. Alternative BET inhibitors featuring distinct pharmacokinetic properties are currently under preclinical assessment (97, 98).

Chimeric antigen receptor (CAR)-modified T lymphocyte adoptive transfer emerges as a promising immunotherapeutic approach for central nervous system malignancies. Using genetic modification methods, autologous T cells are reprogrammed to display synthetic antigen receptors on their cellular membranes, enabling targeted recognition of tumor-associated antigens (61, 99). These genetically engineered CAR-T cells exhibit the ability to selectively identify and attach to neoplastic cells, subsequently triggering the immune-effector function of T lymphocytes to efficiently eradicate neoplastic cells (100). Actually, the success of this strategy relies on the recognition of relatively specific cancer targets on the cell surface, effectively overcoming the common immune evasion mechanisms adopted by tumor cells. Clinical investigations into CAR-T cell therapy for glioblastoma have devoted to five distinct antigens: EphA2, EGFRvIII, HER2, IL13Rα2, and GD2. The early results of these targets have been published, revealing some remarkable achievements (101, 102). Although CAR T cell therapy didn’t improve OS significantly, one glioblastoma patient survived 59 months after EGFRvIII-CAR T cell therapy without post-CAR treatment, and some on HER2-CAR T cell therapy had stable disease up to 29 months. A patient treated with anti-IL13Rα2 CAR T cells achieved a full remission within 7.5 months, showing potential.

However, tumor cell immune evasion against CAR T cells and inadequate CAR T cell trafficking to tumor loci represent primary challenges for CAR T cell therapy in solid tumor treatment, with patients frequently experiencing relapse following initial tumor regression. Thus, enhancing CAR T cell infiltration into tumor sites emerges as a critical strategy to circumvent this immune evasion ( Figure 5 ). Studies have shown that upregulated chemokine receptor expression in CAR-T cells can boost their ability to infiltrate tumor microenvironments (103). For example, increased chromatin availability at the CD56 genomic region, together with chemokine upregulation including CXCL9, CXCL10, and CXCL12, enhances tumor infiltration by PD-1/TIM3/LAG3-modified CAR-T cells. Moreover, expression of CASTAT5 (a constantly active STAT5 isoform) fosters CD4+ T cell infiltration and migration to tumor microenvironments through epigenetic reprogramming. In addition, T cell dysfunction and impaired trafficking arise when miR-155 directly inhibits suppressor of cytokine signaling 1, disrupting normal regulatory mechanisms (104); The upregulation of miR-155 in CAR T cells appears to be an interesting strategy to promote the transportation of CAR T cells and an effective anti-tumor response (105); Targeting let-7 miRNA, which suppresses CCR2/CCR5 expression in T cells, might improve CAR-T cell infiltration by restoring chemokine receptor activity (106). Latest studies also show CAR T cell activation, proliferation, and survival are regulated epigenetically (88). For example, DNMT3A (a DNA methyltransferase) regulates epigenetic changes during chronic LCMV infection’s effector-to-exhaustion transition. In vivo, DNMT3A-deficient CAR-T cells outperform wild-type ones, enhanced by IL-10; double KO CAR-T cells confirm this. SUV39H1 (a histone methyltransferase) modulates chromatin and gene expression, and disrupting it in CAR-T cell therapy boosts cell expansion, persistence, and efficacy. TET2 (a methylcytosine dioxygenase) promotes DNA demethylation. BATF3, a key regulator of memory T cell proliferation and differentiation, plays a critical role in T cell exhaustion—particularly in TET2-knockout CAR-T cells—underscoring the therapeutic possibility of epigenetic modulation in CD8+ T lymphocytes (107). Leveraging epigenetic modulation to boosting CAR-T cell effectiveness presents a promising clinical approach to address the current limitations of solid tumor treatment, where conventional CAR-T approaches have shown limited success. Combining epigenetic agents with adoptive cell therapy may overcome barriers by reprogramming the TME and enhancing T cell persistence.