Targeting NOTCH-immune crosstalk in gliomas: from mechanisms to therapeutics

Abstract

Glioblastoma (GBM) is a highly aggressive and treatment-resistant malignancy characterized by an immunosuppressive tumor microenvironment (TME) that promotes disease progression and poor clinical outcomes. Increasing evidence suggests that NOTCH signaling is a key regulator of GBM progression, governing both tumor cell plasticity and immune evasion. In addition to its roles in development and stem cell maintenance, NOTCH signaling modulates interactions among glioma stem-like cells, immune cells, and stromal components. NOTCH signaling in both tumor and stromal compartments modulates immune cell fate decisions, antigen presentation, and inflammatory cytokine expression. This suggests its potential role as a key modulator of immune crosstalk within the TME of GBM. Moreover, dysregulation of the NOTCH signaling pathway may lead to an immunosuppressive TME, enhance tumor heterogeneity, and restrict responses to conventional therapies. Recent investigations also reveal the epigenetic and non-coding RNA patterns that reinforce aberrant NOTCH activity in GBM. Understanding the multifaceted interactions among NOTCH signaling, oncogenic pathways, and immunomodulatory mechanisms could provide insights into the molecular pathogenesis of GBM progression. Here, we discuss the pleiotropic and context-dependent functions of NOTCH signaling in GBM, with particular emphasis on its implications for tumor–immune interactions and its potential as a therapeutic target to address immune escape and treatment resistance.

Introduction

Gliomas represent a diverse category of primary brain tumors defined according to the classification system of the World Health Organization into various grades and histological subtypes, including astrocytoma (isocitrate dehydrogenase (IDH)-mutant, grades 2–4), oligodendroglioma (IDH-mutant and 1p/19q-codeleted), and glioblastoma (GBM, IDH-wildtype, grade 4) (1–3). GBM is the most aggressive and lethal variant, constituting over 50% of diffuse gliomas and demonstrating a median survival time of only 12 to 15 months, even with intensive multimodal therapies (4, 5). The dismal prognosis of GBM is attributed not only to the intrinsic properties of tumor cells, such as cellular diversity, rapid growth, widespread infiltration, genomic instability, and therapy resistance, but also to the intricate and immunosuppressive tumor microenvironment (TME) (4–6). TME is an intricate and interconnected network system of cancer cells, cancer stem-like cells, reactive astrocytes, immune cell infiltrates (including myeloid-derived suppressor cells (MDSCs), microglia and tumor-associated macrophages (TAMs), and T cells), fibroblasts, and blood vessels, along with an extracellular matrix rich in growth factors, chemokines, and cytokines that all of which collectively influence tumor progression, immune evasion, and resistance to therapy (7–9). Within this micromilieu, complex intercellular signaling networks regulate immune evasion, angiogenesis, and tumor invasion. Among the key signaling pathways mediating communication within the TME, NOTCH signaling serves a dual function, acting as both an oncogenic driver and a tumor suppressor, depending on cellular context (10, 11). The NOTCH signaling pathway is a conserved evolutionary pathway that mediates intercellular communication and plays pivotal roles in stem cell preservation, cell fate determination, and tissue equilibrium. In GBM, aberrant activation of NOTCH receptors (NOTCH 1–4) and their ligands (Jagged (JAG) 1 and 2, as well as Delta-like ligands (DLL) 1, 3, and 4) has been implicated in maintaining the cancer stem cell (CSC) phenotype, promoting resistance to therapy, and driving tumor invasion. The four mammalian NOTCH receptors (NOTCH 1–4) and their canonical ligands (DLL1, DLL3, DLL4, JAG1, and JAG2) are often overexpressed in glioma tissues, and their activation leads to the nuclear translocation of the NOTCH intracellular domain (NICD) that triggers the transcription of oncogenic target genes, like hairy and enhancer of split (Hes) 1 and Hey1. The NOTCH pathway is particularly crucial for the cellular maintenance and viability of glioma stem-like cells (GSCs), a subpopulation that drives tumor recurrence and therapeutic resistance (12–14). Enhanced expression of NOTCH ligands, including DLL4 and JAG1, has been identified in both tumor cells and the GBM vasculature, contributing to angiogenesis and stemness preservation (15–17). More recently, NOTCH3 has been identified as a clinically significant subtype-specific oncogenic factor. Elevated NOTCH3 expression correlates with poorer overall survival, enhanced tumor cell proliferation, and greater immune cell infiltration in IDH1-wildtype, classical, and mesenchymal GBM subtypes (17). Beyond its intrinsic effects on tumor cells, NOTCH signaling regulates the immune signature of the glioma TME, showing dual functions, with either pro-inflammatory or anti-inflammatory effects depending on the specific cellular context and signaling strength (18). Within myeloid cells, the activation of NOTCH signaling can strengthen immunosuppression, whereas in T cells, it has the potential to promote differentiation and effector responses under particular conditions (19–21). This dual and context-sensitive function of NOTCH signaling presents both challenges and opportunities for designing novel therapeutic strategies, making it a potential target to overcome immunosuppression and enhance therapeutic efficacy in glioma. However, the context-sensitive effects of NOTCH activation point out the necessity for precise modulation to limit interference with anti-cancer immunity. Gaining a comprehensive understanding of the multifaceted role of NOTCH in the immune microenvironment of GBM is required for the rational design of NOTCH-targeted therapies and their incorporation into immunotherapeutic approaches. Given these considerations, there is an increasing necessity to more accurately define the function of NOTCH signaling in shaping the immunosuppressive microenvironment of GBM. While existing literature has thoroughly characterized the tumor-intrinsic oncogenic functions of NOTCH, particularly in gliomagenesis, stemness, lineage plasticity, and therapeutic resistance (22–27), the immune-related aspect of NOTCH signaling is still not well understood and has not been clearly brought together across studies. Therefore, in this review paper, we highlight the role of NOTCH signaling in shaping the TBM of GBM, with a particular focus on its context-dependent effects on immune regulation. Specifically, we discuss how NOTCH signaling affects key immune cell populations, including microglia, TAMs, MDSCs, and T-cell subsets, as well as its impact on cytokine networks, metabolic reprogramming, and intercellular communication within the TME. Our objective is to integrate current mechanistic insights across tumor and stromal compartments, highlighting how NOTCH signaling contributes to immune modulation and, in many contexts, immune suppression and tumor immune evasion. Importantly, we critically assess both supportive and contradictory evidence to clarify the context-dependent nature of NOTCH activity in GBM. By synthesizing these findings, this article aims to address a key gap in the field: the lack of a comprehensive framework linking NOTCH signaling to tumor–immune crosstalk in GBM. Finally, we also address the therapeutic implications of targeting NOTCH signaling, emphasizing the need for precise, cell–type–specific, and context-aware strategies to effectively incorporate NOTCH modulation into immunotherapeutic approaches.

Activation and signal transduction mechanisms of NOTCH pathway

NOTCH Signaling is triggered by a precisely controlled proteolytic cascade that begins with the interactions between NOTCH receptors (NOTCH 1-4) and their respective signaling ligands (JAG1, JAG2, as well as Dll1, Dll3, and Dll4) expressed on adjacent cells (28–31). NOTCH receptors are produced as single-pass transmembrane precursors within the endoplasmic reticulum. After being translocated to the Golgi apparatus, they experience furin-mediated cleavage (S1 cleavage), resulting in a heterodimeric receptor that comprises an extracellular subunit (NECD) and a transmembrane subunit (NTM). The fully developed receptor is then transported to the plasma membrane, where it is ready for ligand interaction. The activation of NOTCH signaling occurs when the binding of a ligand to its receptor causes conformational alterations in the NOTCH extracellular domain. This interaction necessitates the mechanical force produced by the endocytosis of the ligand-receptor complex within the signaling cell, leading to sequential proteolytic cleavage events: (i) S2 cleavage by ADAM (a disintegrin and metalloproteinase), which removes the extracellular domain, and (ii) S3 cleavage by γ-secretase, which releases the NOTCH intracellular domain (NICD) from the membrane (28–31).

The released NICD migrates to the nucleus, where it establishes a transcriptional activation complex in conjunction with the recombination signal-binding protein for immunoglobulin kappa J region (RBP-J, alternatively termed CSL/CBF1) and the coactivator Mastermind-like (MAML). This assembly facilitates the expression of primary NOTCH target genes, which encompass members of the Hes (Hes1, Hes5) and Hey (Hey1, Hey2) families, implicated in regulating cell differentiation, proliferation, and viability (29, 32, 33). The core components and activation steps of the NOTCH pathway are illustrated in Figure 1. Beyond the canonical RBP-J-dependent pathway, NOTCH signaling exerts diverse cellular effects through non-canonical pathways that are increasingly recognized for their regulatory roles in lineage-specific differentiation, progenitor cell function, and tumorigenesis (34). In these alternative pathways, the NICD bypasses the classical transcriptional activation via RBP-Jκ and instead interacts directly with a range of oncogenic pathways (35–37). Importantly, NOTCH can regulate Wnt/β-catenin signaling, either by stabilizing β-catenin or controlling Wnt receptor expression (38, 39). Furthermore, crosstalk with the JAK/STAT axis has been observed, where NOTCH activates IL-6/STAT3 signaling through a mechanism regulated by the NF-κB components IKKα/IKKβ, pointing to the role of NOTCH in cancer-associated inflammation (40). The PI3K/AKT/mTOR pathway is another critical effector of non-canonical NOTCH function. For instance, NICD interacts with mTORC2 to activate AKT, which is critical for the maintenance of brain tumor–initiating cells (41–43). A key advancement in understanding NOTCH signaling in GBM requires distinguishing the roles of individual receptors and ligands, as growing evidence shows that this pathway exhibits considerable functional heterogeneity. Quantitative analyses demonstrate that although core components are coordinately upregulated in GBM, the four mammalian receptors (NOTCH1–4) drive divergent or even opposing phenotypes (32). NOTCH1, for example, exhibits a context-dependent and somewhat paradoxical profile. Although higher NOTCH1 expression has been reported in patients with longer survival (>1 year), its overexpression has also been accompanied by reduced overall survival, indicating a dual role in gliomagenesis (44, 45). Furthermore, NOTCH1 is preferentially expressed in peritumoral glioma stem cells, indicating spatial heterogeneity in receptor activity (32, 46). NOTCH4 expression shows a clear association with increasing tumor grade and is preferentially elevated in primary GBM, consistent with a role in tumor progression and aggressive disease phenotypes (47). NOTCH3 acts as a consistent pro-tumorigenic oncogene enriched in aggressive classical and mesenchymal subtypes, correlating with poor survival and high immune infiltration (48, 49). By contrast, NOTCH2 supports neural stem cell maintenance and astroglial lineage entry but negatively correlates with oligodendroglial lineage commitment (47, 50).

Figure 1.

Molecular architecture of canonical and non-canonical NOTCH signaling transduction. Activation is initiated by ligand–receptor interactions (JAG1/2, DLL1/3/4 engaging NOTCH1–4) between adjacent signal-sending and receiving cells. This interaction triggers sequential proteolytic events: S2 cleavage by ADAM metalloproteinases and subsequent S3 cleavage by the γ-secretase complex, releasing the active NOTCH intracellular domain (NICD). Canonical Pathway: NICD translocates to the nucleus, associates with the DNA-binding protein RBP-J (CSL) and co-activator MAML, displacing co-repressors to drive transcription of target genes (Hes and Hey families). Non-canonical Pathway: NICD bypasses RBP-J to crosstalk directly with alternative oncogenic cascades (including Wnt/β-catenin, PI3K/Akt/mTOR, NF-κB, and JAK/STAT), coordinating diverse cellular responses.

Moreover, canonical ligands trigger distinct and often antagonistic programs. In the vascular compartment, DLL4 and JAG1 exert opposing effects on angiogenic sprouting. DLL4, predominantly expressed by endothelial tip cells, inhibits excessive sprouting, whereas JAG1, expressed in stalk cells, counteracts DLL4-mediated signaling and promotes vascular sprouting (32, 51–53). Furthermore, DLL3 acts primarily as a cis-inhibitor; it is epigenetically regulated via promoter methylation and is highly upregulated in IDH1-mutant gliomas, where it correlates with improved overall survival. Conversely, JAG1 serves as a major driver of local and systemic immune evasion, directly upregulating PD-L1 via the PI3K/Akt pathway and promoting regulatory T cell (Treg) recruitment and systemic lymphopenia (23). Table 1 presents the distinct NOTCH receptors and their canonical ligands, detailing their expression patterns, cellular contexts, functional effects, and clinical implications in glioma.

Table 1.

Receptor- and ligand-specific functions of notch signaling in glioblastoma (GBM).

NOTCH receptor/ligand	Cell types	Effects in GBM/Glioma	Clinical implications	Ref.
NOTCH1	Peritumoral GSCs, bulk tumor cells, Thymic/systemic T cells	Sustains GSC self-renewal; Drives radio/chemoresistance; mediates PDL1 expression and immune suppression	Dual prognostic role; target for receptor-specific antibody therapies	(32, 44)
NOTCH2	GSCs, neurospheres, oligodendroglioma cells	Maintains stemness and promotes astrocytic differentiation; limits GBM neurosphere proliferation.	Tumor-promoting in GBM but suppressive in oligodendroglioma/ LGG	(50, 54)
NOTCH3	GBM cells (classical/mesenchymal), CD4+ T cells	Promotes cell proliferation, invasion, and high immunosuppressive immune cell infiltration	Independent prognostic biomarker; candidate therapeutic and diagnostic target	(17, 48, 49)
NOTCH4	Primary GBM tumor cells, Tumor vascular endothelial cells	Cooperates with DLL4 in angiogenesis; wild-type is proangiogenic.	Loss-of-function mutations serve as a predictive biomarker for favorable ICI therapy response	(47, 52, 55)
DLL1	Bulk GBM tumor tissue (co-upregulated with NOTCH1/NOTCH4/JAG1)	Activates PI3K/AKT signaling to promote cell survival and tumor angiogenesis.	Marker for radiotherapy resistance; Component of a pro-tumorigenic signature	(32, 52, 53, 56)
DLL3	IDH1-mutant glioma/LGG tumor cells	Acts as a noncanonical inhibitory ligand; epigenetically regulated via promoter methylation.	Subtype-specific prognostic biomarker in IDH- mutant LGG; candidate for stratification	(57–59)
DLL4	Endothelial tip cells (tumor vasculature); Tumor cells	Restrains excessive angiogenic sprouting (VEGF feedback); drives anti-VEGF resistance.	Poor-prognosis marker; therapeutic target to disrupt vascular-immune niche	(23, 52, 60)
JAG1	Vascular stalk cells, LGG/GBM tumor cells, TAMs, Thymic epithelial cells	Antagonizes DLL4 to promote sprouting; drives TAM-M2- polarization and systemic thymic atrophy.	Key driver of microenvironmental immune escape; candidate combination target with ICIs	(23, 61, 62)
JAG2	Not distinctly characterized for specific glioma/GBM cell populations	One of the canonical NOTCH ligands in general pathway-activation and receptor–ligand interaction contexts, No GBM data were identified for JAG2	Represents a knowledge gap warranting investigation of JAG2-specific effects and clinical correlates in glioma	(10, 11)

GBM, glioblastoma; LGG, low-grade glioma; GSC, glioma stem-like cell; TAM, tumor-associated macrophage; PD-L1, programmed death-ligand 1; ICI, immune checkpoint inhibitor; VEGF, vascular endothelial growth factor.

Physiological and pathological significance of NOTCH signaling pathway in GBM

NOTCH signaling pathway is important for processes such as maintenance of stem cell populations, cell fate specification, and lineage determination during embryogenesis, as well as for regulating tissue homeostasis. In the central nervous system, NOTCH enables neurogenesis and gliogenesis by modulating the balance between progenitor cell proliferation and differentiation (63). It is also central in shaping immune cell development, especially in T cell lineage specification and myeloid differentiation (64). However, dysregulation of NOTCH signaling can contribute to a wide range of pathological conditions. Abnormal NOTCH activity resulting from genetic mutations (65), epigenetic changes (66), or imbalanced ligand-receptor expression (67) has been implicated in congenital disorders (68), fibrotic diseases (69), and multiple cancer types (37). In hematologic malignancies such as T-cell acute lymphoblastic leukemia, activation of NOTCH1 mutations promotes uncontrolled cell proliferation (70), whereas in epithelial cancers such as squamous cell carcinoma, NOTCH typically acts as a tumor suppressor (71). In the context of GBM, NOTCH signaling contributes to different hallmarks of gliomagenesis, many of which are outlined in Figure 2. High-throughput transcriptomic studies and tissue profiling have shown increased expression of NOTCH receptors (NOTCH1–4), canonical ligands (Jagged1, DLL1, DLL4), and downstream effectors (Hes1, Hey1) in GBM tissues compared to control brain tissue (12, 54, 72, 73). NOTCH1, in particular, is often upregulated and is associated with a higher glioma grade, poor prognosis, and greater invasiveness (44, 74). Emerging evidence highlights the crucial role of NOTCH3 in the pathogenesis and clinical behavior of GBM. While all four NOTCH receptors (NOTCH1–4) are upregulated in GBM, NOTCH3, in particular, has been identified as an independent prognostic marker associated with poor survival outcomes (17). Integrated analyses of TCGA and CGGA datasets revealed that high NOTCH3 expression is enriched in aggressive GBM subtypes, including the classical and mesenchymal types, and correlates with IDH1 wild-type and non-G-CIMP tumors, which are known to have a poor prognosis (17). NOTCH3 expression also demonstrated a strong association with tumor cell proliferation and immune cell infiltration, particularly involving CD4⁺ T cells, macrophages, and dendritic cells (75). Immunohistochemical validation in a clinical GBM cohort further confirmed that over 50% of patient samples exhibited high NOTCH3 protein levels, and patients with elevated NOTCH3 expression had significantly shorter survival times. One of the most significant functions of NOTCH is in sustaining the GSC population (76). Activation of NOTCH1 and its downstream targets, such as Hes1 and Hey1, maintains GSC self-renewal, suppresses differentiation, and confers resistance to radiation and chemotherapy (12, 77). Moreover, NOTCH signaling interacts synergistically with key oncogenic pathways, such as PI3K/Akt, STAT3, and NF-κB, to enhance gliomagenesis, particularly by facilitating mesenchymal-like transcriptional transition (MES-like state) and sustaining angiogenesis (73, 78, 79). Under hypoxic stress, hypoxia-inducible factor (HIF)-1α cooperates with NOTCH to preserve the GSC phenotype and promote a vascular, pro-invasive tumor architecture (80). Moreover, NOTCH activity is affected by epigenetic regulators and miRNAs. For instance, miR-34a, miR-181c, and histone methylation patterns affect the expression of NOTCH effectors like Hes1 and Hey1, thereby linking chromatin dynamics to NOTCH-mediated oncogenesis (81–83). The enhanced cell proliferation induced by knockdown of KMT2A, an epigenetic regulator and histone methyltransferase, could be reversed by expressing either constitutively active NOTCH1 or NOTCH3 in GBM (84). The function of the NOTCH signaling pathway in glioma is highly context-dependent. While hyperactivation of NOTCH signaling generally promotes tumorigenesis, glioma subtypes exhibit distinct patterns of pathway alteration. For instance, DLL3 expression is more prevalent in IDH-mutant gliomas, whereas in mesenchymal GBM, NOTCH activation is enriched in patients with high p-STAT3 expression, suggesting a transcriptional convergence between NOTCH and STAT3 signaling (85, 86). These observations highlight the marked heterogeneity of NOTCH pathway regulation across glioma subtypes and suggest that its downstream effects are shaped by the surrounding molecular and transcriptional landscape. Consistent with this subtype-specific complexity, NOTCH signaling is widely recognized as a highly context-dependent pathway, capable of exerting both oncogenic and tumor-suppressive functions depending on the cellular and molecular environment. While this review primarily focuses on its pro-tumorigenic and immunomodulatory roles in GBM, accumulating evidence highlights that NOTCH may also function as a tumor suppressor in specific glioma contexts (11). Indeed, comprehensive genomic analyses of lower-grade gliomas (LGGs) have identified recurrent loss-of-function mutations in key NOTCH pathway components, including NOTCH1, NOTCH2, and RBPJ, which are associated with reduced expression of canonical target genes and increased tumor cell proliferation, suggesting that pathway inactivation contributes to glioma progression (87–91). Consistently, experimental models demonstrate that genetic ablation of NOTCH signaling accelerates tumor growth, whereas enforced activation of NOTCH2 or downstream effectors such as HEY2 can suppress glioma cell proliferation and tumor formation (92, 93). Mechanistically, these tumor-suppressive effects have been linked, at least in part, to NOTCH-mediated repression of pro-neurogenic and proliferative drivers such as ASCL1, thereby maintaining cellular quiescence and blocking malignant transformation (92, 94–96). Importantly, such findings underscore that the functional outcome of NOTCH signaling is shaped by lineage state, genetic background (e.g., IDH mutation status), and tumor subtype, which may explain the apparent discrepancies across studies. Therefore, a critical interpretation of NOTCH biology in GBM requires integrating these context-specific and sometimes opposing roles, particularly when considering therapeutic strategies aimed at modulating this pathway.

Figure 2.

NOTCH signaling–mediated oncogenic mechanisms in GBM. NOTCH signaling acts as a central regulator of gliomagenesis, contributing to multiple malignant processes. This schematic illustrates key oncogenic functions radiating from a central NOTCH signaling hub, including GSC maintenance, enhanced proliferation and survival, MES-like state and invasion, angiogenesis, interaction with other oncogenic pathways (e.g., PI3K/Akt, STAT3), epigenetic and miRNA-mediated regulation, and modulation of the tumor immune microenvironment.

Beyond its tumor-intrinsic functions, the NOTCH signaling pathway also contributes to the regulation of the glioma TME (97). In particular, NOTCH signaling modulates immune cell differentiation, cytokine production, and angiogenic signaling within the TME (11, 98, 99). However, these effects are highly context-dependent and require distinction between tumor cell–intrinsic and microenvironmental mechanisms. Emerging evidence supports a dual mode of action. On one hand, activation of NOTCH signaling within glioma cells and GSCs can regulate the expression of immunomodulatory factors, including cytokines such as IL-10 and TGF-β and immune checkpoint molecules, thereby indirectly shaping immune responses and promoting immunosuppressive phenotypes in TAMs and T cells. On the other hand, NOTCH signaling can also operate in a cell-autonomous manner within immune cells. For example, NOTCH activation in myeloid cells, including macrophages and microglia, influences polarization programs through interactions with NF-κB and STAT signaling pathways, while in T cells it contributes to lineage specification and effector differentiation (64). In GBM, NOTCH receptors are primarily expressed on tumor cells, whereas their ligands are predominantly expressed on endothelial cells (100). Endothelial cells in the perivascular niche express NOTCH ligands that support GSC self-renewal (101), while stromal cells and TAMs engage NOTCH to support immunosuppressive circuits (102). Collectively, the Notch pathway plays a dual role as both a driver of glioma pathogenesis and a key orchestrator of immune regulation within the TME. Importantly, much of the current evidence is derived from preclinical models or correlative analyses, and the extent to which these interactions represent causative, cell–type–specific mechanisms in human GBM remains incompletely defined. Its diverse biological effects, including stemness maintenance, angiogenesis, and immune regulation, highlight its potential as a therapeutic target, as discussed in the following section.

Therapeutic targeting of NOTCH signaling pathway in GBM

Targeting NOTCH signaling is a potential therapeutic strategy for GBM, with several approaches evaluated in preclinical studies and early clinical trials (103). γ-secretase inhibitors (GSIs), such as MRK003, DAPT, and RO4929097, are the most widely studied NOTCH inhibitors. Preclinical studies have shown that these agents suppress GSC proliferation, reduce stemness markers (CD133, Nestin, BMI-1, OLIG2), and sensitize tumors to radiation and temozolomide (TMZ) by downregulating AKT and STAT3 pathways (104–109). MRK003, for instance, effectively reduced tumor size and enhanced radiosensitivity in CD44 high/CD133 low GBM-initiating cells, marking it as a potential candidate for stem-targeted therapy (108). Similarly, RO4929097, another GSI, exhibited anti-proliferative effects in GBM spheroid models and prolonged survival in orthotopic mouse models, especially when combined with radiation and chemotherapy (107, 109). DLL4-induced NOTCH signaling promotes resistance to anti-vascular endothelial growth factor (VEGF) therapy by driving large-vessel formation and activating alternative pathways, and blocking these pathways enhances the efficacy of bevacizumab, an anti-VEGF drug used in GBM, in DLL4-expressing tumors in xenograft GBM mouse models (60). Silencing of EFEMP1, an extracellular matrix protein, reduces GBM cell survival in response to TMZ, whereas its overexpression promotes TMZ resistance through multiple signaling pathways, including γ-secretase–dependent activation of NOTCH signaling (110). Combining NOTCH inhibition with radiation markedly reduced both proliferation and self-renewal in three-dimensional organotypic explant cultures derived from surgical GBM specimens (105, 109). GBM neurosphere cultures with high endogenous NOTCH activity were sensitive to NOTCH inhibition, showing increased G0/G1 arrest and reduced colony formation with limited impact on self-renewal or differentiation, whereas NOTCH activation decreased the G0/G1 population, enhanced colony formation, self-renewal, and differentiation (106). Inhibition of the NOTCH signaling pathway via DN-MAML1 expression or application of MRK003 (a GSI) significantly reduced GBM cell growth both in vitro and in vivo, with NOTCH1 and NOTCH2 contributing differentially. NOTCH2 plays a predominant role and suppresses the proliferation of human GBM-derived neurospheres (104).

Despite these encouraging experimental findings, clinical translation remains limited. A Phase 0/I trial of RO4929097 in newly diagnosed GBM patients showed a reduction in NICD levels and CD133⁺ GSCs, indicating on-target activity, but limited tumor control potentially due to drug metabolism issues (CYP3A4 auto-induction) and NOTCH pathway redundancy (111). Similarly, a Phase I study of combination therapy with RO4929097 and bevacizumab showed tolerability and partial responses in a few patients, but with limited long-term benefit (112). In a phase II study, treatment with the γ-secretase inhibitor RO4929097 in 47 patients with recurrent GBM has shown minimal clinical activity, achieving a 6-month progression-free survival of 4% and inhibiting neurosphere formation in only 1 of 7 resected tumor samples (113). The clinical utility of RO4929097 in GBM is limited by several factors, including poor bioavailability across the blood-brain barrier, variable efficacy between tumor grades due to differential NOTCH pathway activity, and off-target toxicity implicating normal physiological functions (114). To overcome these challenges, CB-103, a first-in-class small molecule that blocks the NOTCH transcriptional activation complex, is under evaluation in an ongoing Phase I/IIa trial, testing its effectiveness both as a monotherapy and in combination with other therapies (115). Unlike GSIs, CB-103 bypasses upstream receptor processing, is well tolerated, and may avoid the toxicity issues associated with γ-secretase inhibitors (116). Alternative approaches involve NOTCH receptor–specific antibodies, which minimize the toxicity associated with pan-NOTCH inhibition. For instance, selective NOTCH1 blockade preserves antitumor efficacy while decreasing gastrointestinal adverse effects (117). Another potential approach targets α-secretase, the enzyme responsible for the initial proteolytic step that precedes γ-secretase–mediated NOTCH activation. In this context, Floyd et al. demonstrated that INCB3619, a dual ADAM10/17 inhibitor, effectively suppresses GSC growth and tumor progression in vivo, significantly prolonging survival in mouse models of GBM (118). The antitumor effect was attributed to diminished NOTCH signaling activity, as evidenced by decreased expression of the NOTCH downstream effectors Hes1 and Hey1. Moreover, combinatorial treatment with DAPT, a γ-secretase inhibitor, further enhanced these effects, not only inhibiting classical NOTCH target genes but also downregulating glioma-associated oncogenic mediators, such as leukemia inhibitory factor and YKL-40 (CHI3L1) (118). While NOTCH inhibitor monotherapy has shown limited clinical efficacy, combination approaches, particularly with radiotherapy, TMZ, or immune-modulating agents, may provide greater translational potential (119). A detailed overview of current and emerging therapeutic strategies targeting the NOTCH pathway in GBM is provided in Table 2. Although Table 2 summarizes current therapeutic strategies targeting the NOTCH signaling pathway in glioma, several limitations should be acknowledged. First, the available clinical evidence remains limited, as most studies are derived from early-phase clinical trials with relatively small patient cohorts, which restricts the generalizability of the reported outcomes (115, 120, 121). In addition, heterogeneity among patient populations, including differences between GBM and anaplastic astrocytoma, as well as variations in molecular characteristics such as IDH1/2 mutation status, may influence treatment response and complicate the interpretation of clinical results (122). Another important limitation relates to challenges in evaluating treatment response, since radiological progression during therapy may represent pseudoprogression caused by therapy–induced blood–brain barrier disruption rather than true tumor progression (121, 123). Furthermore, pharmacokinetic variability and heterogeneous drug penetration across tumor regions due to the blood–brain barrier can affect the extent of NOTCH pathway inhibition achieved in vivo (111). Moreover, the context-dependent nature of NOTCH signaling in glioma, where both activation and suppression can differentially influence tumor and immune compartments, further complicates therapeutic targeting and may contribute to heterogeneous treatment responses. These observations underscore important considerations for the development of future NOTCH-directed therapeutic strategies. In this context, a deeper understanding of the isoform-specific and context-dependent roles of NOTCH signaling, as well as its interactions with immune and stromal components of the TME, will be critical for the development of precision-based therapies for GBM. Notably, NOTCH signaling dynamically interacts with the tumor immune microenvironment to promote immune suppression and therapy resistance, as described in the following section.

Table 2.

Therapeutic targeting of NOTCH signaling in GBM.

Therapeutic agents	Type/mechanism	Preclinical evidence	Clinical findings	Limitations	Ref.
MRK003	γ-secretase inhibitor (GSI)	↓ GSC proliferation; ↓ CD133, Nestin, BMI-1, OLIG2; ↑ radiosensitivity in CD44^high/CD133^low GSCs	Not yet in clinical trials for GBM	Effective in targeting resistant GSC populations	(108)
DAPT	GSI	Enhances TMZ/radiation response; ↓ NOTCH targets (e.g., Hes1); synergy with α-secretase inhibitors	Not clinically tested in GBM	Useful in combination therapies to suppress NOTCH and tumor-promoting factors (LIF, YKL-40)	(107, 118)
RO4929097	GSI	↓ GSC proliferation; ↓ spheroid formation; ↑ survival in mouse GBM models (monotherapy + combo)	Phase 0/I Trial (NCT01119599): ↓ NICD, ↓ CD133+ cells; Phase I combo with bevacizumab (NCT01189240): partial responses, tolerability	Auto-induces CYP3A4 → rapid metabolism; NOTCH redundancy	(109, 111, 112, 118)
CB-103	Small molecule NOTCH transcriptional complex inhibitor (downstream of γ-secretase)	Bypasses γ-secretase, directly blocks NOTCH target gene activation	Ongoing Phase I/IIa trial (NCT03422679)	May reduce toxicity and overcome resistance to GSIs	(116)
NOTCH1-specific antibodies	Monoclonal antibodies targeting NOTCH1	↓ tumor growth with limited GI toxicity vs. pan-NOTCH inhibition	Not yet tested clinically in GBM	Isoform-specific targeting reduces systemic side effects	(117)
INCB3619	Dual ADAM10/17 (α-secretase) inhibitor	↓ NOTCH activity (↓ Hes1, Hey1); ↓ GSC growth; ↑ survival in GBM mouse models; synergy with DAPT	Not in clinical trials yet	Targets upstream of γ-secretase; combinatorial potential with GSIs	(118)
Combination therapies (e.g., GSI + TMZ/radiation)	Multi-targeted approach	↓ stemness, ↓ AKT/STAT3, ↓ LIF/YKL-40; ↑ radiosensitivity and chemosensitivity	Clinical trials (e.g., RO4929097 + bevacizumab) show modest benefit	Combination more effective than monotherapy; need better patient stratification	(108, 109, 111, 112, 118)

Clinical trial identifiers: RO4929097 (NCT01119599, NCT01189240), CB-103 (NCT03422679). GSI, γ-secretase inhibitor; NICD, NOTCH intracellular domain; GSC, glioma stem cell; TMZ, temozolomide; DLL3, Delta-like ligand 3, Ab, antibody; LIF, Leukemia Inhibitory Factor; YKL-40/CHI3L1, GBM-associated glycoprotein; ADAM10/17, α-secretases; GI, Gastrointestinal.

NOTCH as an integrative regulator of the GBM microenvironment

Rather than acting through isolated cell-specific effects, NOTCH signaling in GBM operates across interconnected cellular compartments to establish a coordinated immunosuppressive network. In this framework, tumor-intrinsic NOTCH activation, particularly within GSCs, initiates the production of cytokines, chemokines, and angiogenic factors that reshape the surrounding TME. These signals subsequently drive the recruitment and polarization of myeloid populations, impair antigen presentation, and promote T-cell dysfunction. Thus, the immunosuppressive TME should be viewed not as a collection of independent NOTCH-mediated effects, but as the outcome of a multi-step, hierarchical process linking tumor cell signaling to systemic immune dysregulation. Because these effects are mediated by multiple cellular compartments with distinct biological functions, understanding NOTCH signaling in GBM requires consideration of its cell-type-, receptor-, and ligand-specific activities.

Cellular context-dependence of NOTCH signaling in GBM

An important consideration in interpreting the role of NOTCH signaling in GBM is its strong dependence on cellular context. Although NOTCH activity is frequently associated with tumor progression, stemness, and therapeutic resistance in glioma cells and GSCs, its effects in immune and stromal compartments may differ substantially. In GSCs, NOTCH1–HES1/HEY1 signaling supports self-renewal, inhibits differentiation, and contributes to resistance to radiation and chemotherapy (70–72). NOTCH signaling also interacts with PI3K/AKT, STAT3, NF-κB, and hypoxia-associated HIF-1α pathways, further supporting GSC maintenance and aggressive tumor phenotypes (73, 78–80). In contrast, within the immune compartment, NOTCH signaling can regulate myeloid-cell states and T-cell differentiation, with outcomes that depend on the cellular and molecular context (19–21, 64). In GBM, NOTCH-associated signaling in TAMs and microglia has been linked to immunoregulatory polarization and production of tumor-supportive mediators, including IL-10, TGF-β, CCL2, and matrix metalloproteinases (21, 124, 125). However, NOTCH activity in T cells cannot simply be considered immunosuppressive, because NOTCH also has established roles in T-cell lineage specification and effector differentiation (19–21, 64). This distinction is particularly relevant when considering systemic effects of tumor-associated NOTCH signaling, as glioma-derived JAG1 can alter thymopoiesis through NOTCH1-dependent mechanisms contributes to changes in peripheral T-cell composition (126–128). NOTCH signaling is similarly heterogeneous in endothelial and stromal compartments. In the GBM vasculature, DLL4 and JAG1 participate in regulating vascular organization and the perivascular niche associated with GSC maintenance (17, 23, 100, 101). However, is should be noted that these ligands are not functionally interchangeable. Differential enrichment of DLL4 and JAG1 has been observed across distinct GBM microvascular patterns, supporting ligand-specific regulation of vascular behavior (23). Thus, the biological consequences of NOTCH signaling depend not only on whether the pathway is activated, but also on which receptor, ligand, and cellular compartment are involved. Collectively, these observations indicate that NOTCH should not be viewed as a uniformly oncogenic or immunosuppressive pathway in GBM. NOTCH1–4 and their ligands can generate distinct, and in some settings, opposing biological effects depending on the responding cell and molecular environment. This context dependence provides an important rationale for selective rather than indiscriminate NOTCH targeting in GBM.

NOTCH as a multicellular driver of TME remodeling

Failure to treat patients with metastatic or recurrent GBM is largely attributed to tumor heterogeneity and the difficulty of effectively targeting the TME (129). Within this context, NOTCH signaling functions as a central regulator of the TME, exerting immunomodulatory effects through interconnected transcriptional, epigenetic, and intercellular mechanisms. Importantly, NOTCH activity is not restricted to tumor cells but extends to endothelial cells, astrocytes, microglia, and TAMs, forming a dynamic intercellular signaling network (15, 72, 130). Disruption of NOTCH-mediated, dynamic, and interdependent interactions among different components of the TME can drive tumor progression through various mechanisms: (i) Transcriptional upregulation of immune checkpoint molecules, like PD-L1 and B7-H4 that inhibit cytotoxic T cell activation and drive T cell exhaustion (131); (ii) Suppression of antigen presentation, including the downregulation of major histocompatibility complex (MHC) class I and co-stimulatory molecules, thereby impairing CD8⁺ T cell recognition of tumor antigens and facilitating escape from immune attack (132); and (iii) Polarization of myeloid cells, with TAMs adopting M2-like immunosuppressive states and microglia shifting toward immunoregulatory phenotypes, promotes the release of anti-inflammatory cytokines, matrix-remodeling enzymes, and pro-angiogenic factors, creating a tumor-supportive niche (133, 134). These mechanisms generate a tolerogenic immune milieu, provide an immunosuppressive environment, and enhance GBM tumorigenicity, recurrence, and therapy resistance. These interactions are schematically represented in Figure 3. Upregulation of NOTCH1, Dll4, and downstream transcriptional targets, such as Hey1 and Hes1, has been consistently observed in patient-derived GBM tissues and has been linked to aggressive tumor behavior and poor clinical outcomes (56, 135, 136). Within the TME, astrocytes and microglia are key TME components that support GBM growth and progression (137). The spatial proximity between GBM cells, glial cells, and immune cells facilitates a direct cell-to-cell contact network, enabling localized Dll4-NOTCH signaling activation, thereby amplifying both pro-angiogenic and immunosuppressive pathways in a subset of GBM tissues (26, 136). This endothelial–glial immune crosstalk plays a dual role for NOTCH signaling: driving tumor cell survival and stemness while modulating immune-vascular niches (138, 139). Targeting this axis may thus disrupt this pathological signaling network and enhance the efficacy of immunotherapeutic strategies in GBM. In a comprehensive review, Li et al. provided evidence that aberrant NOTCH signaling modulates diverse immune cell populations, including TAMs, MDSCs, regulatory T cells, and dendritic cells, often promoting immunosuppressive phenotypes in TME (21). This signaling cascade induces the immunosuppressive condition by enhancing VEGF-mediated angiogenesis and supporting the recruitment and reprogramming of innate immune cells. Microglia and macrophages, which infiltrate gliomas extensively, secrete matrix metalloproteinases as well as cytokines under the influence of NOTCH and NF-κB. This further enhances tumor invasion, ECM remodeling, and aberrant vascularization in GBM (124, 125, 140–142). Therapy-induced modulation of this network is also evident. In oncolytic virotherapy models, viral infection triggers JAG1-mediated NOTCH activation in TAMs, initiating a feed-forward loop involving CCL2-driven recruitment of MDSCs, M2 polarization, and suppression of CD8⁺ T-cell responses. Pharmacological inhibition of NOTCH reverses these effects, highlighting its role in adaptive immune resistance and supporting combinatorial therapeutic strategies (61) (73). Collectively, these findings position NOTCH signaling as a central coordinator of multicellular interactions that drive immunosuppressive remodeling of the GBM microenvironment.

Figure 3.

NOTCH signaling-mediated immunosuppression in glioblastoma (GBM). This schematic illustrates how NOTCH signaling modulates both local and systemic immunosuppressive mechanisms in GBM. Locally, NOTCH activation promotes immune checkpoint upregulation (e.g., PD-L1, B7-H4), drives polarization of TAMs and microglia toward an M2-like phenotype, recruits MDSCs and Tregs, and supports the formation of a vascular niche that restricts T-cell infiltration. Systemically, GBM-derived Jagged1 activates thymic NOTCH1 signaling, leading to heme oxygenase-1 (HO-1) downregulation, thymic atrophy, loss of CD4⁺-CD8⁺ thymocytes, and expansion of regulatory T cells, contributing to peripheral lymphopenia and impaired anti-tumor immunity. Note: Abbreviations: GBM, glioblastoma; TME, tumor microenvironment; NICD, NOTCH intracellular domain; TAMs, tumor-associated macrophages; MDSCs, myeloid-derived suppressor cells; Tregs, regulatory T cells; PD-L1, programmed death-ligand 1; B7-H4, B7 homolog 4; HO-1, heme oxygenase 1.

NOTCH and immune checkpoint regulation

One of the most direct mechanisms by which NOTCH signaling mediates immune evasion in GBM is through the regulation of immune checkpoint molecules that suppress T-cell activation and function. Targeting immune checkpoints in GBM, for example, via HIF-2α inhibition, can remodel the immune infiltrate, shifting the TME from an immunosuppressive condition to a state that supports a stronger and sustained therapeutic response (143). Similarly, NOTCH signaling can shape the immune microenvironment of LGG. In an integrative analysis using TCGA datasets, Shi et al. developed a quantitative NOTCHScore, revealing that high NOTCH activity correlates with poorer prognosis and increased immune checkpoint expression, particularly PD-L1, in LGG (62). These findings indicate that NOTCH signaling contributes to immune evasion and tumor progression across gliomas, regardless of their grade or intrinsic aggressiveness. Among NOTCH pathway components, JAG1 emerged as a critical regulator of LGG cell proliferation and migration. Mechanistically, JAG1 promotes PD-L1 expression via the PI3K–Akt axis and the vitamin D receptor, linking NOTCH activation to immunosuppressive signaling. Knockdown of JAG1 not only impaired tumor cell growth but also disrupted immune checkpoint regulation, reinforcing its functional relevance in glioma immune escape (62). These observations are consistent with broader findings that gliomas exploit NOTCH signaling to modulate the TME, promoting regulatory T cell (Treg) recruitment, TAM polarization, and T-cell dysfunction (8, 9, 144). Given the clinical success of immune checkpoint inhibitors (ICIs) in other solid tumors (145), their application in GBM has been explored (146). Although long-term responses to ICIs in GBM remain below 10% (147); monitoring temporal changes in the tumor-specific genetic profile could serve as a predictive biomarker of therapy response in patients with GBM (148). Furthermore, targeting NOTCH components, such as JAG1, in combination with ICIs or alone has been proposed as a potential immunotherapeutic approach for gliomas, including LGG. Beyond canonical NOTCH ligands and receptors, emerging evidence highlights additional upstream regulators that converge on the NOTCH–immune checkpoint axis to reinforce immunosuppressive signaling in glioma. Zinc finger proteins are key transcriptional regulators in GBM that shape the TME by driving angiogenesis, immune suppression, invasion, and cellular plasticity (149, 150). Recent evidence identifies ZNF480, a member of the Zinc finger protein family, as a novel transcriptional regulator contributing to glioma pathogenesis through NOTCH pathway activation and immune modulation. In LGGs, high ZNF480 expression independently predicts poor prognosis and correlates with NOTCH signaling enrichment, as revealed by gene set enrichment analysis (151). ZNF480 is functionally associated with an immunosuppressive TME, characterized by elevated M2 macrophage infiltration and PD-L1 upregulation, aligning with NOTCH-mediated immune evasion mechanisms. Experimental silencing of ZNF480 reduces its oncogenic expression profile, highlighting its potential as a therapeutic target. These findings position ZNF480 at the intersection of NOTCH-driven stemness and immune suppression, offering a dual-target strategy to overcome therapy resistance in LGG (151). It has been suggested that NUP205, a core nucleoporin of the nuclear pore complex (NPC), is a potential contributor to glioma progression and immune evasion through NOTCH pathway regulation. In a comprehensive transcriptomic analysis of LGG, Liang et al. reported that NUP205 is significantly overexpressed in aggressive LGG subtypes (IDH-wildtype, 1p/19q non-codeleted) and correlates with worse prognosis and an immunosuppressive TME (152). NUP205 modulates both NOTCH signaling and cell cycle regulators, implicating NPC dysfunction in the transcriptional control of oncogenic and immune-related pathways. High NUP205 expression was linked to enhanced PD-L1 expression and elevated infiltration of M2-like macrophages, suggesting that NUP205 may facilitate immune escape through the enhancement of NOTCH-mediated immune checkpoint activation and myeloid reprogramming (152). However, it is critical to note that these immunoregulatory links are currently supported primarily by in silico correlations and expression profiling. These findings are consistent with prior studies demonstrating that NOTCH activation via JAG1 or NOTCH1 promotes PD-L1 upregulation and suppresses T-cell function in glioma. As such, NUP205 represents a novel upstream node in the NOTCH–PD-L1 axis and may serve as a dual-purpose therapeutic target, simultaneously attenuating glioma growth and reversing immune evasion. Together, these findings identify ZNF480 and NUP205 as regulatory nodes that connect transcriptional control with immune checkpoint activation and myeloid reprogramming in glioma. Collectively, these findings indicate that NOTCH signaling promotes immune evasion by directly regulating checkpoint molecule expression, thereby limiting effective T-cell–mediated anti-tumor responses.

In addition to PD-1/PD-L1 signaling, increasing evidence indicates that multiple inhibitory checkpoints contribute to the complex and redundant immune-suppressive network of GBM. Among these, TIGIT, LAG-3, and TIM-3 are inhibitory immune checkpoints expressed on both lymphoid and myeloid cells that can suppress antitumor immune responses and contribute to resistance to single-checkpoint blockade. An investigation has identified an unexpected role for TIGIT in GBM-associated myeloid cells. Asad et al. demonstrated broad TIGIT expression in myeloid populations within human GBM and showed that GBM-derived extracellular vesicles can induce TIGIT expression in monocytes. TIGIT knockdown reduced the immunosuppressive polarization of these cells and restored T-cell function in an NLRP3-dependent manner, with TIGIT and NLRP3 showing substantial co-expression. These findings suggest that TIGIT functions not simply as a marker of T-cell exhaustion but as a regulatory node controlling the balance between immunosuppressive myeloid activity and NLRP3-associated pro-inflammatory responses in the TME of GBM (153). This myeloid checkpoint axis is particularly relevant to therapeutic strategies aimed at reprogramming the immunosuppressive TME. LAG-3 (CD223) represents another layer of immune suppression, predominantly through exhausted T cells and regulatory T cells. In human glioma, LAG-3⁺ tumor-infiltrating lymphocytes are enriched in IDH-wildtype GBM and are associated with CD8⁺ T-cell infiltration, PD-1 expression, and PD-L1-positive tumor cells, supporting its relationship with an already activated but functionally suppressed immune microenvironment (154). Mechanistically, LAG-3 can engage several ligands, including MHC-II, FGL1, and galectin-3, and its functional cooperation with PD-1 can reinforce T-cell dysfunction (155–157). Of particular relevance to GBM, a recent study demonstrated that radiotherapy induces galectin-3 expression in GSCs through EGR1, thereby activating the Gal3–LAG-3 axis and promoting exhaustion of both CD4+ and CD8+ T cells in recurrent GBM. The combination of EGR1 inhibition and LAG-3 blockade suppressed tumor growth in orthotopic models, pointing to a therapy-induced mechanism of immune escape (158).

TIM-3 (HAVCR2) provides an additional checkpoint linking adaptive immune exhaustion to myeloid and innate immune regulation. TIM-3 is expressed by T cells, NK cells, macrophages, and dendritic cells, as well as in GBM tumor cells. Its interaction with ligands, such as galectin-9 and CEACAM1, can suppress T-cell and NK-cell activity while promoting immunosuppressive myeloid phenotypes (159, 160). In GBM, TIM-3 expression has been associated with higher tumor grade, an IDH-wildtype/mesenchymal phenotype, M2-like macrophage accumulation, and tumor-promoting signaling through a Gal-9/TIM-3–IL-6/NF-κB axis (161, 162). TIM-3 may also suppress innate immune sensing by limiting HMGB1-mediated uptake of tumor-derived nucleic acids and cGAS–STING activation in dendritic cells, thereby further weakening antitumor immunity (163, 164).

Collectively, these findings support a model in which GBM immune escape is maintained by interconnected rather than independent checkpoint pathways. Consistent with this concept, a recent clinical study of recurrent GBM showed that nivolumab treatment activated intratumoral T cells but simultaneously induced compensatory upregulation of TIGIT, LAG-3, TIM-3, and CTLA-4, including increased populations co-expressing multiple checkpoints. Thus, NOTCH-driven immune suppression may be viewed within a broader network in which developmental signaling, myeloid reprogramming, T-cell exhaustion, and innate immune regulation converge to produce a highly resilient immunosuppressive TME. Therapeutically, these observations provide a rationale for multi-target or context-dependent checkpoint strategies, instead of inhibition of a single checkpoint.

Crosstalk with developmental pathways (Wnt and Hedgehog)

In addition to its direct immunoregulatory functions, NOTCH signaling operates within a broader developmental signaling network that amplifies tumor progression and immune suppression. Crosstalk with pathways such as Wnt and Hedgehog reinforces NOTCH-driven transcriptional programs and stabilizes immunosuppressive phenotypes. Pan-cancer analyses have identified WNT5A as a key immunoregulatory hub that co-activates NOTCH and Hedgehog pathways to facilitate tumor progression (165). The interaction between Wnt and NOTCH signaling pathways controls development and the pathogenesis of various diseases (166). WNT5A, a non-canonical Wnt ligand, is upregulated in GBM and promotes cell infiltration in TME (167). NOTCH signaling is active in spheroid-forming GSCs, and inhibition of NOTCH signaling suppresses the expression of Ror1, a receptor or co‐receptor for Wnt5a, in GBM cells (168, 169). WNT5A is consistently overexpressed and correlates with poor prognosis, increased immune checkpoint expression (such as programmed death-ligand 1 (PD-1) and CTLA-4), and enrichment of NOTCH transcriptional programs in gliomas. WNT5A amplifies NOTCH-driven immune suppression by promoting Treg recruitment, M2 macrophage polarization, and co-regulating pro-inflammatory cytokines, such as IL-12, IL-23, and TNF-α through NOTCH1 (168).

Interestingly, tumors with high WNT5A expression exhibit features of “immune-inflamed” (hot) tumors, characterized by increased T-cell infiltration, elevated interferon-γ signaling, enhanced PD-1 expression, and high tumor mutational burden, which are generally predictive of better responses to ICIs (170). However, WNT5A-driven gliomas often display an “immune-hot but suppressed” TP1 subtype, where despite abundant immune infiltration, clinical outcomes remain poor, supporting the role of WNT5A as a co-driver of NOTCH-mediated immune evasion (168). Moreover, dysregulation of the Wnt/planar cell polarity (PCP) pathway contributes further to GBM aggressiveness. VANGL2, a core PCP gene, has been linked to genes enriched in NOTCH signaling and stem cell pluripotency pathways. Functional studies have demonstrated that VANGL2 enhances glioma cell migration, invasion, proliferation, and tumor sphere formation, highlighting its role as a mediator of glioma progression (171). Thus, WNT5A represents both a functional modulator and a potential biomarker of NOTCH activity, with therapeutic value when co-targeted alongside GSIs or ICIs to disrupt this immunosuppressive circuit. These findings demonstrate that NOTCH signaling functions within an interconnected signaling network that integrates developmental and immune pathways to sustain immunosuppression. Importantly, the evidence supporting these NOTCH pathway interactions is heterogeneous. Among the currently described crosstalk mechanisms, the NOTCH–Hedgehog interaction provides a particularly informative example of functional pathway compensation, as increased Hedgehog activity following NOTCH inhibition and direct interaction between HES1 and GLI1 have been demonstrated in GBM neurosphere and primary GBM cells (172). Consistent with this mechanism, combined inhibition of NOTCH and Hedgehog signaling produces greater effects on apoptosis, cell growth, and colony formation than either treatment alone (172). In contrast, several Wnt–NOTCH associations, including those involving WNT5A, are supported primarily by transcriptomic, observational, and preclinical studies and should therefore be considered hypothesis-generating rather than definitive evidence of clinical efficacy. These observations nevertheless provide a rationale for investigating pathway-selective combination strategies and for exploring molecular features of pathway co-activation as potential biomarkers for patient stratification. Such approaches remain preclinical and require validation before clinical application.

Systemic NOTCH-mediated immunosuppression: thymic atrophy and Treg bias in glioma

While NOTCH signaling has been extensively characterized within the local TME, emerging evidence indicates that its immunoregulatory effects extend beyond the tumor site to influence systemic immune homeostasis. In particular, NOTCH plays a central role in T-cell development, where receptor–ligand interactions within distinct thymic compartments govern lineage commitment, maturation, and functional polarization (173). NOTCH signaling regulates the antitumor immune response, and higher numbers of T cells (particularly activated CD8+ cytotoxic T lymphocytes and Th1 cells) are associated with improved survival in various malignancies (174). In the context of glioma, tumor-derived signals can disrupt this tightly regulated process. El Andaloussi et al. demonstrated that intracranial glioma progression induces profound thymic atrophy through activation of the JAG1–NOTCH1 axis, leading to depletion of CD4⁺CD8⁺ double-positive thymocytes and impaired maturation of single-positive T-cell subsets. Mechanistically, aberrant NOTCH activation suppresses HO-1, a key anti-apoptotic regulator, thereby promoting selective apoptosis of developing thymocytes. This shift in thymic dynamics favors the expansion of CD4⁺CD25⁺Foxp3⁺ regulatory T cells (Tregs), resulting in a systemic immunosuppressive phenotype (126–128). Importantly, bone marrow transplantation experiments confirmed that glioma-derived JAG1 is sufficient to induce these effects, pointing to a tumor-intrinsic mechanism for shaping the peripheral T-cell repertoire. These findings suggest the JAG1–NOTCH1–HO-1 axis as a potentially significant pathway contributing to immune dysregulation in specific preclinical models of glioma. However, further validation in representative models and patient-derived systems is required to determine its broader clinical relevance and status as a primary driver of systemic immunosuppression.

Functionally, these systemic effects are not independent of the TME but instead reinforce local immunosuppressive circuits. The expansion of Tregs and reduction in effective cytotoxic T-cell output from the thymus limit the availability of anti-tumor effector cells, thereby amplifying NOTCH-driven immune evasion within the TME. This bidirectional relationship suggests that glioma leverages NOTCH signaling to coordinate immune suppression across both central and peripheral compartments. Clinically, such systemic immune alterations may contribute to lymphopenia observed in GBM patients and help explain the limited efficacy of immunotherapies that depend on robust T-cell priming. Therapeutically, these findings highlight the need for strategies that selectively modulate NOTCH signaling while preserving essential immune functions, potentially through combinatorial approaches that prevent both local and systemic immunosuppression. Taken together, these findings indicate that NOTCH signaling extends beyond the TME to systemically reshape immune cell development and composition. Importantly, these systemic alterations are not independent of local tumor processes; rather, they reinforce and sustain the immunosuppressive circuitry established within the glioma microenvironment. This bidirectional interaction suggests that NOTCH signaling operates across multiple biological scales, integrating tumor-intrinsic programs with both local and systemic immune regulation.

Contextual complexity of NOTCH signaling

The complexity introduced by pathway crosstalk helps explain the context-dependent characteristics of NOTCH signaling in glioma. While NOTCH activation is predominantly associated with immune suppression, its functional impact varies depending on cellular context and microenvironmental conditions. Experimental evidence indicates that NOTCH inhibition can impair anti-tumor immunity. In murine glioma models, loss of NOTCH activity reduces MHC class I expression and cytokine production, leading to diminished CD8⁺ T-cell recruitment and enhanced immunosuppressive myeloid activation. Furthermore, NOTCH-deficient tumors exhibit resistance to interferon-γ signaling and macrophage-reprogramming therapies, highlighting a role for NOTCH in maintaining immune competence (165). Transcriptomic analyses support these findings, showing that low NOTCH activity is associated with downregulated interferon responses and activation of oncogenic pathways. These findings indicate that excessive NOTCH signaling promotes immune evasion, while complete suppression may disrupt essential immune surveillance mechanisms (165). Taken together, these data point to the context-dependent nature of NOTCH pathway, where both activation and inhibition can differentially shape immune responses and influence therapeutic outcomes.

NOTCH as a multi-level immunoregulatory hub

Collectively, the evidence presented positions NOTCH signaling as a multi-level immunoregulatory hub that integrates tumor-intrinsic programs with microenvironmental and systemic immune dynamics in glioma. Rather than acting through isolated mechanisms, NOTCH coordinates a hierarchy of processes spanning tumor cells, immune populations, and stromal components. At the tumor-intrinsic level, NOTCH activation promotes stemness, metabolic adaptation, and the production of immunomodulatory factors. These signals extend into the microenvironment, where NOTCH drives immune checkpoint expression, suppresses antigen presentation, and orchestrates the recruitment and polarization of myeloid cells toward immunosuppressive phenotypes. Beyond these local effects, NOTCH signaling is further reinforced through crosstalk with developmental pathways such as Wnt and Hedgehog, which amplify transcriptional programs associated with immune evasion and tumor progression. This networked signaling architecture enables GBM cells to sustain a stable yet adaptable immunosuppressive niche, even under therapeutic pressure. At the same time, emerging evidence indicates that NOTCH activity contributes to systemic immune regulation, suggesting that its effects extend beyond the tumor site to influence broader immune competence. Importantly, the function of NOTCH signaling is highly context-dependent. While sustained activation predominantly reinforces immune suppression, its complete inhibition may disrupt interferon signaling, impair antigen presentation, and weaken effective T-cell responses. This duality highlights NOTCH not simply as an oncogenic pathway, but as a dynamic regulatory system that balances immune activation and suppression depending on cellular and environmental conditions. The diverse immunosuppressive roles of NOTCH signaling in the glioma TME are summarized in Table 3, showing its interactions with tumor, immune, and stromal components that modulate immune escape and therapy resistance. Taken together, these observations support a model in which NOTCH signaling serves as an integrative hub linking tumor plasticity, immune evasion, and therapeutic resistance. This conceptual framework has important implications for treatment strategies, suggesting that selective modulation rather than complete inhibition of NOTCH signaling may be required to effectively restore anti-tumor immunity in glioma. Importantly, the integrative role of NOTCH signaling across tumor, immune, and stromal compartments converges on a central cellular determinant, GSCs. As such, GSCs function not only as a reservoir of tumor plasticity but also as an active signaling hub that links NOTCH-driven stemness programs with coordinated immune suppression, as detailed in the following section.

Table 3.

Immunosuppressive roles of NOTCH signaling in the glioma TME and systemic immunity.

Mechanism	Key findings	Molecular/cellular players	Functional consequences	Clinical/therapeutic implications	Ref
Local immunosuppression in TME
Immune Checkpoint Upregulation	NOTCH induces PD-L1, B7-H4 via JAG1-PI3K-Akt/VDR.	NOTCH1, Dll4, JAG1, Hes1/Hey1	T-cell exhaustion, impaired cytotoxicity	Combined NOTCH + ICI therapy may overcome resistance	(56, 62, 135, 136)
Antigen Presentation Suppression	Downregulates MHC-I and co-stimulatory molecules.	NOTCH inactivation (NOTCH1/2 loss)	Reduced CD8+ T-cell recognition	NOTCH modulation (not broad inhibition) required	(165)
Myeloid Polarization (TAMs/Microglia)	Drives M2-like phenotypes via Jag1/NF-κB.	TAMs, microglia, CCL2, MMPs	IL-10/TGF-β secretion, angiogenesis, ECM remodeling	GSIs reverse polarization; enhance immunotherapy	(21, 124, 125)
Endothelial–Glial Crosstalk	Dll4-NOTCH promotes vascular-immune niche.	Dll4, VEGF, HIF-1α	Stemness, immune evasion, angiogenesis	Anti-Dll4 therapies disrupt niche	(72, 130)
Treg/MDSC Recruitment	NOTCH activates CCL2/JAG1, recruiting immunosuppressive cells.	JAG1, WNT5A, CCL2	T-cell suppression, tumor progression	JAG1 inhibition + ICIs may improve responses	(61, 165)
Systemic immune dysregulation
Thymic Atrophy & Altered T-cell Repertoire	Glioma-derived Jag1 disrupts thymopoiesis via NOTCH1.	Jag1, NOTCH1, HO-1	Depletion of DP thymocytes, Treg expansion	Preserving HO-1 may mitigate lymphopenia	(126–128)
Peripheral Lymphopenia	Jag1-NOTCH1 axis reduces naïve T-cell output.	CD4+CD8+ DP thymocytes, CD4+CD25+Foxp3+ Tregs	Impaired T-cell priming, systemic immunosuppression	Bone marrow-targeted NOTCH modulation	(128)
Therapy resistance mechanisms
Oncolytic Virotherapy Escape	oHSV induces Jag1 in TAMs, recruiting MDSCs.	Jag1, CCL2, MDSCs	Limits oHSV efficacy	GSIs + oHSV combination therapy	(61)
WNT5A-NOTCH Crosstalk	Co-activates NOTCH to drive Treg/M2 polarization.	WNT5A, IL-12/IL-23, TNF-α	“Immune-hot but suppressed” phenotype	Dual WNT5A + NOTCH blockade	(165)
NUP205-Mediated Immune Evasion	Enhances PD-L1 and M2 infiltration via NOTCH.	NUP205, PD-L1, NOTCH1	Immune checkpoint activation	Target NPC-NOTCH axis	(152)

MHC-I, major histocompatibility complex class I; NF-κB, nuclear factor kappa-light-chain-enhancer of activated B cells; VEGF, vascular endothelial growth factor; HIF-1α, hypoxia-inducible factor 1-alpha; IL-10, interleukin-10; TGF-β, transforming growth factor-beta; CCL2, chemokine (C-C motif) ligand 2; oHSV, oncolytic herpes simplex virus; ICI, immune checkpoint inhibitor; TAMs, tumor-associated macrophages; MDSCs, myeloid-derived suppressor cells; Tregs, regulatory T cells.

NOTCH signaling in coordinating GSC maintenance and immunosuppression

NOTCH as a driver of GSC stemness and plasticity

GSCs provide an important cellular interface between NOTCH signaling and the immune microenvironment. Rather than functioning solely as a self-renewing tumor cell population, GSCs can actively participate in remodeling the surrounding niche through cytokine, chemokine, and intercellular signaling. In this context, NOTCH signaling may connect GSC-associated programs with immune regulation and therapeutic adaptation (12). Targeting NOTCHless homolog 1 (NLE1), a negative regulator of NOTCH activity, impairs NOTCH signaling, modulates stem cell marker expression, and sensitizes GSCs to radiation-induced cell death (175). Inhibition of NOTCH signaling appears to reduce the growth and self-renewal capacity of CD133⁺ GSCs both in vitro and in vivo, indicating its potential role in depleting the stem-like tumor cell reservoir (12). According to Tang et al., GSCs mediate tumor progression, immune evasion, and resistance to therapy, with NOTCH signaling acting as a regulatory pathway linking these processes (176). In the TME, NOTCH cooperates with other developmental pathways, like Wnt/β-catenin and Hedgehog, allowing GSCs to thrive under hypoxic conditions (177). These interactions activate transcriptional programs that maintain stem-like identity and promote phenotypic plasticity and metabolic adaptation, both of which contribute to tumor progression and recurrence (177, 178). In sum, these data suggest NOTCH signaling as a central modulator of GSC plasticity and persistence, forming the foundation for downstream immune modulation.

NOTCH-driven GSC–immune crosstalk

Beyond maintaining stemness, NOTCH signaling enables GSCs to actively remodel the immune microenvironment through cytokine-mediated communication and immune cell reprogramming. Rather than functioning as passive responders, GSCs leverage NOTCH activation to restructure the immune milieu through the secretion of immunomodulatory factors and engagement of signaling cascades. For instance, interferon regulatory factor 7 (IRF7) amplifies NOTCH output by inducing IL-6, CXCL1, and CCL2, thereby promoting angiogenesis, recruiting TAMs, and activating the JAK/STAT3–Jagged1–NOTCH signaling axis (179). This feed-forward loop reinforces both the GSC phenotype and the establishment of an immune-tolerant TME. Importantly, silencing IRF7 attenuates stemness and immune suppression simultaneously, highlighting its role as a convergent regulatory node (179). These observations demonstrate that NOTCH signaling functions as a bidirectional interface through which GSCs both shape and are reinforced by the immunosuppressive microenvironment.

Hypoxia–NOTCH–Hedgehog axis in immune evasion

Under hypoxic conditions, NOTCH signaling cooperates with developmental pathways to couple stemness maintenance with immune evasion. Chang and Lai demonstrated that dual activation of NOTCH and Hedgehog signaling defines an aggressive GSC phenotype characterized by elevated stemness markers (EZH2, SUZ12, REST) and localization within hypoxic niches (180). Mechanistically, hypoxia stabilizes HIF-1α, which enhances NOTCH–Hedgehog signaling interactions, thereby sustaining cancer stem cell maintenance. This signaling axis promotes immune escape by increasing Treg infiltration, suppressing MHC class I expression, and reinforcing VEGF-driven angiogenesis. The resulting feedback loop supports tumor survival under stress while limiting immune recognition and therapeutic responsiveness. Together, these findings identify hypoxia-driven NOTCH crosstalk as a key mechanism linking metabolic adaptation to immune suppression in GSCs.

Transcriptional modulators of NOTCH in GSCs

Fine-tuning of NOTCH signaling in GSCs is mediated by transcriptional regulators that integrate developmental and immune signaling programs. Among these, HES6 emerges as a non-canonical NOTCH effector that modulates transcriptional output and balances glioma stemness with immune regulation. Unlike HES1, HES6 suppresses HES1/ASCL1 activity while enhancing DLL1 expression, thereby reshaping NOTCH signaling dynamics (181). Notably, HES6 is enriched in both CD8⁺ T cells and TAMs and correlates with gene signatures associated with epithelial–mesenchymal transition, Wnt/KRAS signaling, and both pro-inflammatory and immunosuppressive pathways (182). These dual associations suggest that HES6 integrates tumor-intrinsic and immune-related signals within the TME. Overall, transcriptional regulators such as HES6 refine NOTCH activity to coordinate stemness programs with immune modulation.

Therapeutic targeting of the NOTCH–GSC axis

Targeting NOTCH signaling in GSCs represents a strategy to simultaneously disrupt tumor maintenance and restore immune surveillance. Evidence indicates that Hedgehog signaling is increased following NOTCH pathway inhibition, and a direct interaction between Hes1, a transcriptional repressor, and Gli1, a major downstream effector of Hedgehog signaling, has been observed in both GBM neurosphere lines and primary GBM cells. Simultaneous targeting of both the NOTCH and Hedgehog pathways significantly enhances apoptosis and more effectively inhibits cell growth and colony formation compared to monotherapy (183). These findings revealed that targeting NOTCH either directly or in combination with Hedgehog inhibitors, hypoxia modulators, and/or checkpoint blockade therapies offers a potential strategy to simultaneously disrupt GSC maintenance and restore immune surveillance within the glioma TME (184). Together, these approaches underscore the therapeutic potential of targeting the NOTCH–GSC axis within a broader network context.

Primary cilia as a platform for NOTCH signaling integration

Subcellular signaling platforms such as primary cilia further contribute to the regulation of NOTCH-dependent pathways in GSCs. Primary cilia are microtubule-based organelles that mediate cellular signaling by both receiving and integrating extracellular cues within the TME, thereby contributing to oncogenesis across multiple cancer types (185). They contribute significantly to regulating neural stem cell proliferation and migration, hosting important pathways such as NOTCH, Hedgehog, and WNT (186). In GBM, the impact of primary cilia may depend on the abundance of ciliated tumor cells and their signaling functions, with evidence suggesting that glioma cilia can either restrict or promote gliomagenesis (187). By integrating extracellular signals, primary cilia fine-tune NOTCH pathway activity and coordinate NOTCH-driven communication between tumor cells and the surrounding microenvironment, thereby linking microenvironmental sensing to stem cell adaptation and therapeutic resilience. These observations suggest that subcellular signaling architecture adds a regulatory layer to NOTCH-mediated control of GSC behavior and immune interactions. As detailed in the following section, this regulatory framework enables GSCs to exploit NOTCH signaling to survive genotoxic stress and maintain an immunosuppressive niche that ultimately impairs therapeutic efficacy.

The NOTCH–GSC–immune axis in glioma therapy resistance

Therapeutic resistance in GBM reflects the interaction between resilient tumor cell populations and a protective TME. Evidence suggests that the NOTCH signaling pathway contributes to this interaction by connecting GSC-associated survival programs with immune and stromal mechanisms (177, 188). Specifically, the activation of NOTCH1 sustains GSC persistence through downstream PI3K/AKT/mTOR pathway stimulation, enabling tumor repopulation following radiotherapy and correlating with poor clinical outcomes (189). Beyond contributing to intrinsic resistance, NOTCH signaling plays a role in bridging therapy resistance with immune evasion. Zhang et al. reported that JAG1 mediates radiotherapy resistance and is downregulated when radiotherapy is combined with colony-stimulating factor 1 (CSF-1) blockade, which depletes immunosuppressive TAMs (190). Moreover, high expression of JAG1, NOTCH1, and DLL1 correlates with poor prognosis in glioma patients undergoing radiotherapy (62). This resistance is reinforced by YAP1, a Hippo pathway effector, that transcriptionally activates JAG1, forming a YAP1–JAG1–NOTCH1 loop that sustains both tumor progression and immune suppression (191).

In a complementary model, Chockalingam et al. showed that overexpression of macrophage colony-stimulating factor in glioma cells promotes the recruitment of TAMs while simultaneously inducing resistance to 5-fluorouracil via NOTCH1-mediated EMT and expansion of a CD24high/CD44low CSC population (130, 192). These findings point to a dual mechanism whereby TAM-derived immunosuppressive cytokines enhance NOTCH activation, which in turn promotes chemoresistant CSC expansion, creating a mutually reinforcing feedback loop. Transcriptomic analyses further reveal that GSCs display heightened NOTCH dependency, contributing to resistance against TMZ and radiotherapy (109, 193, 194). Although NOTCH signaling is widely dysregulated in gliomas, clinical translation of NOTCH inhibitors remains challenging (193). For instance, RO4929097, a γ-secretase inhibitor, showed limited efficacy as a monotherapy in glioma models, despite robust in vitro activity and frequent NOTCH pathway activation (195). Its benefit emerged only when combined with genotoxic agents like TMZ or irradiation, highlighting the context-specificity of NOTCH dependency (195). Furthermore, MDSCs use complex immunosuppressive strategies adapted to the TME, making their interaction with GSCs a key target for therapy (102). Notably, treatment response correlated inversely with IL-6 expression, linking NOTCH signaling to IL-6-mediated MDSC recruitment and TAM polarization. Failure of RO4929097–mTOR combination therapy further underscores the resilience of the immunosuppressive niche in vivo. Collectively, the data support NOTCH signaling as a molecular integrator of CSC-driven resistance and immune escape, and suggest that effective treatment will require combinatorial strategies targeting NOTCH alongside immune-modulatory agents, such as checkpoint inhibitors, CSF-1R blockers, or TAM reprogramming therapies (196). The MEOX2-NOTCH axis refers to the interaction between the transcription factor MEOX2 (mesenchyme homeobox 2) and the NOTCH signaling pathway, which together influence self-renewal and stemness maintenance of GSCs. Targeting this axis facilitates tumor invasion, metastatic progression, and immune evasion via the enhancement of inflammatory and proteolytic signaling, presumably via the modulation of the NF-κB signaling pathway (197). These data collectively position Notch signaling as a molecular integrator of CSC-driven resistance and immune escape, and suggest that effective treatment will require combinatorial strategies targeting Notch alongside immune-modulatory agents, such as checkpoint inhibitors, CSF-1R blockers, or TAM reprogramming therapies.

MicroRNA regulation of NOTCH-driven GBM stemness and immune evasion

Directly targeting NOTCH signaling has proven effective in diminishing therapy-resistant GSC populations and disrupting reprogramming of the immunosuppressive TME; however, the effects of modulating the upstream regulatory networks that control NOTCH activity are still under investigation. Among these regulators, microRNAs (miRNAs) have emerged as critical modulators of NOTCH-driven phenotypes in glioma (198). These small non-coding RNAs can act as tumor suppressors or oncogenic miRNAs (oncomiRs), influencing key processes in GBM pathology, such as GSC maintenance, immune evasion, and therapy resistance (199). In this part, we addressed the complex interplay between miRNAs and NOTCH signaling in GBM, highlighting their dual role as biomarkers and therapeutic targets in overcoming treatment resistance.

In GBM, tumor-suppressive miRNAs, like the miR-34 family (miR-34a, miR-34c-3p, miR-34c-5p), are frequently downregulated. Their expression inversely correlates with NOTCH1 and NOTCH2 activity and glioma invasiveness. Restoration of miR-34 or miR-31 levels induces apoptosis, cell cycle arrest, and loss of stem-like traits, effectively disrupting NOTCH-driven oncogenic programs (200, 201). On the other hand, glioma-derived oncomiRs, such as miR-92a, are secreted to suppress natural killer cell-mediated immunity by downregulating the expression of key immune molecules involved in killing target cells and orchestrating immune responses, such as interferon-γ (IFN-γ), perforin, and Fas ligand (202). Khwaja et al. identified miR-146b, an immune-modulatory microRNA, as significantly upregulated in post-therapy recurrent GBM and markedly associated with the activation of NOTCH, Wnt, and VEGF signaling pathways (203). miR-146b disrupts anti-tumor immune responses by influencing NF-κB signaling in myeloid cells, which can enhance the NOTCH-mediated recruitment of immunosuppressive TAMs and contribute to an immune-evasive TME (136, 203). Furthermore, the co-activation of NOTCH signaling and miR-146b in recurrent GBM supports GSC populations by promoting NOTCH-dependent self-renewal and intrinsic resistance to chemo-radiation (203, 204). Therapy-induced stresses (Such as hypoxia) upregulate miR-146b and reactivate NOTCH, together promoting an immunosuppressive TME that drives tumor recurrence. Targeting this axis, for example, by combining NOTCH inhibitors and miR-146b antagonists could disrupt the feedback loop between glioma stemness, immune evasion, and therapeutic resistance (203). Furthermore, downregulation of miR-146b-5p enhances glioma cell proliferation, inhibits apoptosis, and is associated with unfavorable survival outcomes. Acting as a tumor suppressor, miR-146b-5p mediates TMZ resistance in GBM through negative regulation of tumor necrosis factor receptor-associated factor 6. miR-146b-5p overexpression restores TMZ sensitivity in resistant GBM cells (205).

Similarly, miR-33a acts as an oncogene in glioma by downregulating its target SIRT6, which leads to reduced apoptosis through activation of the JAK2/STAT3 pathway, promoting glioma cell survival (206). miR-33a has been defined as a master regulator of GSC self-renewal through concurrent activation of the NOTCH and PKA pathways. Elevated miR-33a expression correlated with poor prognosis in GBM patients, and its knockdown suppressed GSC tumorigenicity, while its overexpression conferred stem-like properties to non-GICs. miR-33a overexpression downregulates phosphodiesterase 8A and a UV radiation resistance-associated gene, while upregulating their downstream effectors p-CREB and the NOTCH intracellular domain in glioma cells (207). It is important to note that the expression of various miRNAs varies among different GBM tissue samples. miR-148a is frequently upregulated in GBM and shows a strong correlation with genes involved in hypoxia response and extracellular matrix remodeling. In contrast, elevated values of miR-31 are observed in a small subset of GBM and are associated with genes related to cell proliferation and immune responses (201). Dual pharmacological inhibition of NOTCH by DAPT and PKA by KT5720 synergistically impaired GSC growth, whereas combined activation of both pathways (via Forskolin and JAG1) induced stemness in non-GICs (207). These findings position miR-33a as a central node linking NOTCH-dependent self-renewal and PKA-mediated proliferation, offering a compelling therapeutic target. Given that both NOTCH and PKA pathways contribute to PD-L1 upregulation and T-cell exclusion, targeting miR-33a or its downstream effectors may offer dual therapeutic advantages, disrupting GSC self-renewal and reprogramming the immune microenvironment. Taken together, miRNA dysregulation, particularly miR-34, miR-146b, and miR-33a, contributes to NOTCH hyperactivation and immunosuppressive rewiring in GBM. These interactions enable glioma recurrence and resistance to therapy. Targeted modulation of the miRNA–NOTCH axis provides a novel therapeutic approach, particularly in recurrent or immune-resistant GBM.

ECM-driven modulation of NOTCH signaling in glioma immunosuppression and therapy resistance

ECM is no longer viewed as a passive structural scaffold but rather as a dynamic and bioactive component of TEM that contributes to immune modulation, tumor progression, and therapeutic resistance (208). Recent evidence has highlighted an ECM–NOTCH–immune regulatory axis, in which dynamic changes in matrix composition reciprocally affect both tumor cell behavior and immune cell function (209). CD47 is a cell surface protein that acts as an immune regulator by providing a “don’t eat me” signal and thereby inhibiting the phagocytosis of healthy cells (210). CD47 loss-of-function enhances phagocytosis while paradoxically activating NOTCH signaling. This activation induces tenascin-C (TNC) upregulation, an extracellular matrix glycoprotein overexpressed in GBM and correlated with poor clinical outcomes (211). The NOTCH-mediated TNC secretion drives M2-like TAM polarization, promotes ECM remodeling, impairs T cell infiltration, and exacerbates immunosuppression. Furthermore, autocrine TNC-NOTCH signaling loops in GSCs reinforce stemness properties and therapy resistance (76). These observations imply that CD47-targeted monotherapies may trigger compensatory activation of NOTCH signaling, pointing to the potential of combinatorial approaches, such as combining CD47 blockade with γ-secretase inhibitors, to prevent adaptive immune evasion in GBM (211). In a complementary study, Li et al. characterized ZR30, an engineered fibulin-3 (EFEMP1) derivative that possesses both ECM–modulating and NOTCH-inhibitory properties. Using orthotopic GBM models, evidence indicates that ZR30 effectively targets the EGFR/NOTCH/AKT signaling axis, a central hub of therapeutic resistance, while disrupting critical tumor-ECM interactions (212). At the molecular level, ZR30 suppressed MMP2 activity, a NOTCH-dependent metalloproteinase pivotal for invasion, ECM degradation, and immune exclusion, and modulated downstream effectors by downregulating the proliferative transcription factor FOXM1 while restoring expression of the anti-invasive semaphorin SEMA3B (194, 212). These coordinated actions support the capacity of ZR30 to simultaneously address tumor heterogeneity and immunosuppressive ECM remodeling. The multimodal effects of ZR30, including inhibition of NOTCH signaling and angiogenesis, may facilitate blood vessel normalization, enhance T cell infiltration, and reduce immunosuppressive cytokine production. This positions ECM-targeting agents like ZR30 as promising candidates for combination therapies designed to disrupt the GSC-protective niche and prevent immune evasion (213). These findings establish the ECM as an active enhancer of NOTCH-driven immunosuppression and therapy resistance, indicating the potential of targeting ECM–NOTCH–immune interactions to improve GBM treatment.

Epigenetic crosstalk regulating NOTCH-mediated immune evasion in glioma

Noting the established role of miRNAs in post-transcriptional regulation of NOTCH signaling, recent investigations have indicated the importance of epigenetic modifiers, including chromatin remodelers, histone-modifying enzymes, and methylation-dependent regulators, in forming NOTCH-mediated immune evasion in GBM (214). These epigenetic mechanisms influence NOTCH signaling either directly or through modulation of immune-related gene expression, and contribute to glioma stemness, immunosuppression, and therapy resistance (215). A key player in this process is the Polycomb group protein PCGF1, a component of the PRC1 complex, which substantially contributes to epigenetically enforcing NOTCH-dependent immune suppression. PCGF1 is upregulated in IDH-wildtype GBM and correlates with activation of NOTCH, PI3K-Akt, and JAK-STAT pathways. Tumors with high PCGF1 expression exhibit Th2-biased immunity and T cell depletion, a phenotype consistent with NOTCH-induced immunosuppression (216). PCGF1 is implicated in regulating several oncogenic signaling pathways, including NOTCH signaling, which governs cancer-related processes, tumor immunity, and glioma development. Through Polycomb-mediated silencing, PCGF1 represses antigen presentation and inflammatory cytokine genes, contributing to a hostile immune TME (216). Furthermore, WD40 repeat (WDR) domain-containing proteins are among the largest families in the human proteome that act as versatile scaffolds for protein–protein interactions (217) and maintain oncogenesis in various human cancers (218). WDR76, a WD40 domain chromatin-binding protein, is highly expressed in LGG. Elevated WDR76 expression is linked to more aggressive disease features and independently predicts poorer survival outcomes in LGG patients. This high expression, regulated by DNA methylation, leads to the upregulation of oncogenes such as PRC1 and NUSAP1, as well as activation of the NOTCH signaling pathway. Furthermore, WDR76 is implicated in LGG tumor immunity by facilitating the infiltration of immune cells, including M2 macrophages, and the expression of immune checkpoints, such as PDCD1 (which encodes PD-1) (219).

Beyond direct chromatin regulators, several oncogenic drivers have been identified as epigenetically regulated genes that affect NOTCH signaling and immune dynamics. CCDC34 is a coiled-coil domain-containing protein involved in tumor growth, apoptosis, and invasion, and has been implicated in the development and progression of several cancers (220). CCDC34 is overexpressed in LGGs and is governed by DNA methylation. It activates multiple oncogenic pathways, including NOTCH, PI3K-Akt, and HIF-1, contributing to glioma progression and potentially TAM-mediated immune evasion (221). Although the direct immunological role of CCDC34 is currently under investigation, its regulation and pathway enrichment analyses suggest that it may serve as a link between epigenetic regulation and immune responses. Similarly, the non-canonical NOTCH ligand DLL3, traditionally considered a cis-inhibitor of NOTCH receptors, is recognized for its context-dependent functions in glioma. DLL3 expression is epigenetically regulated, with its levels closely associated with promoter methylation and copy number variation (58, 59). Moreover, IDH mutations in gliomas are predictive of improved survival and are associated with epigenetic changes, including DNA hypermethylation, which alter gene expression and tumor behavior (222). In IDH1-mutant gliomas, cell surface NOTCH ligand DLL3 is upregulated and associated with improved prognosis and distinct immune infiltration patterns, suggesting a potentially immune-permissive function (86). Furthermore, DLL3 expression correlates with immune cell infiltration and negatively with primary cilia-related genes, indicating that DLL3 and the TME together influence LGG outcomes (59). These results demonstrate that DLL3 could serve as a subtype-specific biomarker in gliomas, with important implications for patient grouping (223). Moreover, Cao et al. demonstrated that the histone methyltransferase G9a (EHMT2) promotes NOTCH pathway activation by silencing FBXW7, a ubiquitin ligase that degrades NICD in GSCs. Through H3K9me2-dependent repression, G9a stabilizes NICD, driving sustained NOTCH activity, upregulating PD-L1, and downregulating MHC-I. These changes collectively suppress cytotoxic T cell infiltration and enhance MDSC and M2 macrophage expansion. Notably, inhibition of G9a reversed these effects and restored immune responsiveness, highlighting the therapeutic potential of targeting the G9a–FBXW7–NOTCH axis in combination with immune checkpoint inhibitors ¹⁹³. In sum, Chromatin modifiers like G9a, PCGF1, and WDR76 promote NOTCH-driven immune evasion by suppressing antigen presentation and cytokines, while epigenetically regulated genes such as DLL3 and CCDC34 modulate NOTCH signaling and immune balance, highlighting the epigenetic–NOTCH–immune axis as a promising target to enhance antitumor immunity in glioma. Table 4 shows a summary of the complex and interconnected epigenetic network that governs NOTCH signaling–immune crosstalk in gliomas.

Table 4.

Epigenetic modulators and epigenetically regulated genes influencing NOTCH-driven immune evasion in glioma.

Gene/Protein	Epigenetic Role	Connection to NOTCH	Immune Impact	Therapeutic Relevance	Ref
G9a (EHMT2)	Histone methyltransferase (H3K9me2)	Silences FBXW7, stabilizes NICD	↑ PD-L1, ↓ MHC-I, ↑ MDSCs/M2	G9a inhibitors + ICIs	(224)
PCGF1	Polycomb group protein (PRC1 complex)	Reinforces NOTCH via gene silencing	↓ γδ T cells, Th2 skewing	Polycomb-targeted therapy	(216)
WDR76	Chromatin-associated modulator	Activates NOTCH–NUSAP1–PRC1 axis	↓ IL-15, ↑ PD-1, ↑ M2 TAMs	Potential immune-modulating target	(219)
CCDC34	Epigenetically regulated (DNA methylation)	Activates NOTCH, PI3K, HIF-1	Implicated in TAM polarization	Biomarker, NOTCH/HIF co-target	(221)
DLL3	Epigenetically regulated ligand (methylation, CNV)	Inhibits NOTCH (cis-inhibition)	Immune equilibrium regulator	Prognostic marker in IDH-mutant LGG	(58, 59)

FBXW7, F-box and WD repeat domain-containing 7; PD-L1, programmed death-ligand 1; MDSCs, myeloid-derived suppressor cells; TAMs, tumor-associated macrophages; NICD, NOTCH intracellular domain; PRC1, Polycomb repressive complex 1; CNV, copy number variation.

Inflammation, coagulation, and NOTCH–immune crosstalk in gliomas

The TME is dominated by an excess of pro-angiogenic factors from cancerous, stromal, and immune cells, leading to abnormal blood vessel formation. These abnormal vessels create an inflammatory environment that promotes coagulation, thrombosis, and impaired blood flow, facilitating tumor growth and metastasis (225). Inflammation and coagulation are two distinct pathways that have evolved to function together, providing a strong defense against infection, damaged cells, and irritants, but when dysregulated or corrupted by tumors, like GBM, they can enhance tumor proliferation, invasion, dissemination, and survival (226, 227). Recent findings highlight that NOTCH signaling interacts with inflammation and coagulation pathways. In particular, the complement–coagulation cascade (CCC) has emerged as a novel partner of NOTCH signaling in promoting immune dysfunction and therapy resistance in glioma ¹⁹⁷. Pro-inflammatory cytokines, such as TNFα, alter NOTCH signaling in vascular endothelial cells by decreasing NOTCH4 and increasing NOTCH2 expression, which leads to reduced NOTCH activity and promotes apoptosis. These changes, mediated through NF-κB and PI3-kinase pathways, suggest that inflammation-induced dysregulation of NOTCH contributes to endothelial and vascular dysfunction (55). There is evidence that specific targeting of these pathways may present significant potential for the future treatment of multiple tumor types, including GBM (228, 229).

Yang et al. developed a CCC-based risk signature including genes like F2R, SERPINA1, and C1Q, capable of grouping patients with LGG into high- and low-risk subtypes. The high-risk group exhibited co-activation of NOTCH, JAK-STAT, and PI3K-Akt signaling pathways, promoting stemness, immune suppression, and resistance to immunotherapy. Notably, C3AR1, a CCC effector, was implicated in MDSC recruitment and M2 macrophage polarization, phenocopying NOTCH-driven immunosuppression observed in GBM. Furthermore, the proteasome inhibitor MG-132, which is selectively effective in CCC-high LGGs, has been reported to inhibit both NF-κB and NOTCH signaling, thereby reversing thrombo-inflammatory responses. These data indicate the therapeutic potential of targeting the CCC–NOTCH axis in immune-refractory glioma subtypes (229).

NOTCH mutations as predictive biomarkers for immunotherapy in GBM

Mutations affecting various components of the NOTCH signaling may play a role in the development of various malignancies, including GBM (230). An adeno-associated virus-mediated, autochthonous CRISPR genetic screen in GBM revealed frequent co-mutations of the NOTCH1 receptor and B2m, a main component of the MHC-I antigen presentation complex, suggesting a potential link between Notch signaling disruption and impaired immune recognition in glioma (132). NOTCH1 mutations are found in oligodendrogliomas, low-grade gliomas, and anaplastic astrocytoma that have an IDH mutation and 1p/19q codeletion. However, these mutations are infrequently seen in gliomas with wild-type IDH or in low-grade gliomas with an IDH mutation without 1p/19q codeletion (88).

Mutations in NOTCH receptor genes impact various regions of the protein with distinct functional consequences. Activating mutations are commonly found in the negative regulatory region, which regulates receptor cleavage, or in the C-terminal PEST domain, which controls NICD degradation. Mutations in the PEST domain often lead to prolonged NICD stability, resulting in extended or uncontrolled NOTCH signaling. These alterations lead to ligand-independent or sustained receptor activation (231). Conversely, inactivating NOTCH mutations primarily affect the extracellular epidermal growth factor-like repeats, disrupting ligand binding or causing truncated proteins. Such mutations often result in reduced expression of NOTCH target genes in glioma (11, 232). NOTCH2 protein expression is significantly higher in GBM compared to oligodendroglioma, where NOTCH2 loss-of-function mutations or deletions contribute to tumor development, suggesting that oligodendroglioma progression is linked to reduced or lost NOTCH2 activity (233).

Conversely, mutations in other NOTCH family members may exert antitumor effects through distinct mechanisms. NOTCH4 mutations are associated with significantly better outcomes, including higher objective response rates, durable clinical efficacy, and longer progression-free and overall survival outcomes, in individuals suffering from various cancers, including glioma. Multiomics analysis revealed that NOTCH4 mutations correlate with increased tumor immunogenicity and enhanced antitumor immune activity. These findings point to NOTCH4 mutation as a potential marker for predicting improved response to treatment with ICI (234). Tumors with NOTCH4 mutations were characterized by an intense immune micromilieu, including enhanced CD8⁺ T cell infiltration, high tumor mutational burden, and upregulated immune checkpoints, such as PD-1, PD-L1, and CTLA-4. This immunologically active profile is associated with improved survival following ICI therapy, independent of PD-L1 status or tumor mutational burden thresholds. While the link between NOTCH4 dysfunction and immune activation needs to be elucidated, it is hypothesized that loss-of-function mutations in NOTCH4 disrupt immunosuppressive NOTCH signaling circuits and enhance T cell recruitment and activation. These findings align with prior studies that indicate that NOTCH1 activity can suppress Th1-type inflammation (235), and with the growing use of synthetic NOTCH receptors in engineered immune cells to increase antigen-specific T cell responses (236, 237). In sum, NOTCH signaling mutations play a complex, context-dependent role in glioma. Certain mutations can enhance immune activation, making NOTCH a potential therapeutic target and a biomarker for guiding immunotherapy, while others promote immune evasion and tumor progression.

Future perspectives

NOTCH signaling plays an important role in glioma tumorigenesis. Dysregulation of this pathway promotes tumor growth, stemness, and survival, while fostering an immunosuppressive microenvironment through interactions with immune cells. By modulating immune evasion and supporting tumor cell survival, NOTCH-immune crosstalk contributes to the aggressiveness and therapeutic resistance of gliomas. This review describes how aberrant activation of NOTCH receptors and ligands in tumor and stromal cells facilitates immune evasion via mechanisms like M2 macrophage polarization, Treg expansion, impaired antigen presentation, and reduced cytotoxic T cell infiltration. Furthermore, NOTCH signaling intersects with oncogenic pathways such as STAT3, HIF-1α, PI3K/AKT, and epigenetic regulators, collectively sustaining a therapy-resistant and immune-exclusive tumor phenotype. By maintaining stemness and promoting immune escape, NOTCH represents a promising therapeutic target in GBM. However, several key challenges remain unresolved. The context-dependent nature of NOTCH signaling, where both activation and suppression can produce immunosuppressive outcomes, complicates therapeutic targeting. In addition, the field lacks a clear mechanistic understanding of how tumor-intrinsic NOTCH activity integrates with systemic immune alterations (e.g., thymic atrophy and Treg bias), and why promising preclinical strategies have largely failed in clinical settings. Addressing these gaps will be essential for translating NOTCH biology into effective therapies. Despite advances in understanding NOTCH signaling and its mechanism of action, clinical translation of NOTCH-targeted therapies has been challenging. For instance, clinical trials evaluating first-generation pan-NOTCH inhibitors, such as RO4929097 (a GSI), have yielded minimal therapeutic benefit. In a Phase II study of 47 individuals with recurrent GBM, RO4929097 achieved a 6-month progression-free survival rate of just 4% and failed to inhibit neurosphere formation in 6 out of 7 resected patient tumor specimens (114). Moreover, the combination of RO4929097 and bevacizumab provided only limited long-term clinical activity (114). Earlier Phase 0/I clinical investigation showed pharmacodynamic evidence of NOTCH pathway inhibition, including reductions in NICD and CD133⁺ GSCs, but this molecular activity was not accompanied by substantial tumor control (111). The physical and metabolic microenvironment of the brain presents a significant challenge to effective NOTCH inhibition. First, the blood-brain barrier severely restricts the delivery of GSIs and other small molecules, leading to marked pharmacokinetic variability and heterogeneous drug penetration within different areas of the tumor in vivo. Consequently, therapeutic substances fail to achieve sustained and effective concentrations at the highly infiltrative tumor region. Second, first-generation inhibitors suffer from metabolic instability and auto-clearance (114). Specifically, RO4929097 has been shown to auto-induce its own metabolism via the cytochrome P450 CYP3A4 enzyme, rapidly reducing its systemic bioavailability and therapeutic window during continuous dosing regimens (114). Beyond pharmacology, the highly context-dependent, dual role of the NOTCH pathway within the TME poses a major biological challenge (165). While sustained NOTCH activity promotes an immunosuppressive niche, complete pathway suppression can paradoxically weaken anti-tumor immunity. Preclinical models demonstrated that the complete loss of NOTCH signaling impairs antigen presentation by downregulating MHC-I expression. Furthermore, complete NOTCH deficiency induces resistance to IFN-γ and triggers compensatory, immunosuppressive myeloid cell activation (165). This shows that effective clinical intervention should favor context-dependent and precisely regulated changes in signaling activity over complete or indiscriminate pathway inhibition. Together, these data point to a substantial preclinical–clinical disconnect and indicate that biological evidence supporting NOTCH as a therapeutic target should not be equated with clinically validated efficacy. Future investigations should focus on translating preclinical insights into clinically effective NOTCH–immune therapies while addressing treatment resistance, as outlined in the following suggestions:

1. Precision targeting strategies: Given the differential expression of NOTCH isoforms across distinct glioma subtypes, therapeutic interventions should carefully account for receptor specificity to enhance treatment efficacy and minimize adverse effects. NOTCH1 is significantly upregulated in the Classical and Proneural subtypes compared to the Mesenchymal subtype. Conversely, NOTCH2 and NOTCH3 exhibit higher expression in both the Classical and Mesenchymal subtypes but remain low in the Proneural subtype. NOTCH4 expression varies markedly between the Classical and Proneural subtypes. This indicates the complex and subtype-specific roles of NOTCH signaling in GBM biology (17). Furthermore, specific interactions between Notch and other pathways, such as STAT3 signaling in the mesenchymal GBM subtype, have been identified (85). Adapted approaches improve treatment effectiveness by specifically targeting GBM cells using modulators of distinct NOTCH subtypes. Future studies should define when NOTCH acts as a driver versus a regulator of immune balance, enabling more precise therapeutic modulation.

2. Predictive biomarkers for treatment efficacy: The identification and validation of robust predictive biomarkers are required for refining prognostic assessments and optimizing therapeutic strategies. For example, the NOTCH Score a quantitative metric derived from gene expression data measures Notch pathway activity in tumor tissues. Elevated NOTCH Scores are associated with poorer prognosis, potentially due to enhanced tumor immune evasion via upregulation of immune checkpoint molecules (62). Similarly, NOTCH4 mutations correlate with a less differentiated tumor phenotype and decreased overall survival in glioma patients, underscoring their potential as prognostic biomarkers for patient grouping and personalized treatment planning (238). Additionally, DLL methylation status functions as a key epigenetic biomarker in gliomas; aberrant methylation of DLL family members modulates gene expression, influencing tumor differentiation and Notch signaling dynamics. Integrating DLL3 methylation profiling with comprehensive NOTCH mutation analyses may offer a more nuanced molecular characterization of gliomas, thereby improving prognostic precision and facilitating targeted therapeutic interventions (136, 230). Collectively, the identification and validation of predictive biomarkers are essential for transitioning from broad NOTCH inhibition to a personalized, biomarker-guided treatment framework. Considering the molecular and cellular diversity of gliomas, a clinically meaningful patient selection strategy must integrate several biological layers, including: (i) stratification should incorporate NOTCH receptor/ligand expression profiles, distinguishing subtype-specific targets, such as NOTCH3 in classical/mesenchymal tumors or epigenetically regulated DLL3 in IDH-mutant gliomas; (ii) to capture dynamic pathway signaling, quantifying functional NOTCH activation via nuclear NICD localization is pivotal. This profile must be integrated with GSC-associated signatures (such as the transcriptomic NOTCH Score, the TME immune composition (including CD8⁺ T-cell infiltration, myeloid checkpoint landscapes, or loss-of-function NOTCH4 mutations), and hypoxia-related signatures that drive aggressive HIF-1α/NOTCH crosstalk; and (iii) aligning these features with relevant molecular subtypes and IDH status, particularly by combining DLL3 promoter methylation profiling with mutation typing, will establish a multi-dimensional stratification procedure. This framework will enable clinicians to design context-aware combination therapies designed to the unique microenvironmental and genetic landscape of each tumor.

3. Rational combination treatments: Future therapeutic success will likely depend on the development of rational combination regimens. Synergistic approaches that integrate NOTCH pathway inhibitors with immunotherapeutic modalities, such as PD-1/PD-L1 checkpoint blockade, CSF-1R antagonists, or engineered synNOTCH chimeric antigen receptor-T cells, have significant potential to overcome tumor immune evasion mechanisms and induce persistent anti-tumor responses (21, 239–241). Furthermore, concomitant targeting of key co-regulators of the NOTCH signaling pathway in glioma, including WNT5A, IRF7, and ZNF480, may concurrently disrupt cancer stemness properties and immunosuppressive signaling pathways and improve overall treatment outcomes (151, 242, 243). A key unresolved challenge is identifying which pathway combinations are most effective in specific tumor and immune contexts. Future therapeutic success will likely depend on the development of rational and context-dependent combination regimens. Synergistic approaches that integrate NOTCH pathway inhibitors with immunotherapeutic modalities, such as PD-1/PD-L1 checkpoint blockade, CSF-1R antagonists, or engineered synNOTCH chimeric antigen receptor-T cells, may have the potential to counter tumor immune evasion mechanisms and promote more durable anti-tumor responses (21, 228–230). Given the molecular and immunological heterogeneity of these tumors, the optimal combination is unlikely to be uniform across various forms of gliomas. Thus, biomarker-guided selection of combination partners and treatment strategies will be essential. Furthermore, concomitant targeting of key co-regulators of the NOTCH signaling pathway in glioma, including WNT5A, IRF7, and ZNF480, may concurrently disrupt cancer stemness properties and immunosuppressive signaling pathways and improve overall treatment outcomes (88, 146, 231). A key unresolved challenge is identifying which pathway combinations are most effective in specific tumor and immune contexts.

4. ECM remodeling strategies to enhance NOTCH-targeted therapies: As highlighted in the ECM section, the tumor matrix actively reinforces NOTCH-driven immune suppression and therapy resistance. However, the extent to which ECM remodeling can enhance immunotherapy responses remains unclear. Agents such as ZR30 demonstrate the potential to disrupt ECM–NOTCH interactions, but further work is needed to determine how these strategies influence immune cell infiltration and function in vivo (212, 213) (244). Understanding this interface represents an underexplored but promising therapeutic avenue.

5. Addressing mechanisms of resistance to NOTCH-targeted therapies: Investigate the molecular mechanisms underlying intrinsic and acquired resistance to NOTCH inhibitors in GBM. Understanding how tumor cells bypass NOTCH pathway inhibition, whether through compensatory signaling pathways, epigenetic alterations, or MTE interactions, will guide the development of next-generation inhibitors or combination strategies to overcome resistance. Several approaches targeting Notch-1 are under preclinical and clinical investigation as potential strategies to overcome GBM resistance. Investigations on these pathways may expand therapeutic options to combat this resistance (24, 245, 246).

6. Using computational models to improve Notch pathway targeting: Integrating approaches like artificial intelligence (AI) and machine learning into the study of Notch signaling enables the development of personalized, precise, and effective therapies. AI facilitates the elucidation of complex molecular mechanisms within the Notch pathway in specific tumor contexts, supporting more accurate therapeutic targeting (37). Moreover, AI-driven computational design tools can engineer synthetic molecules with the exact geometry and multivalency needed to activate Notch signaling. These advances can be broadly applied to optimize in vitro cell differentiation and accelerate the development of novel immunotherapies (247). The synergy between computational and biological sciences is transforming current approaches to GBM treatment and is instrumental in fully realizing the therapeutic potential of NOTCH pathway targeting (248).

Conclusion

NOTCH signaling remains a highly compelling biological axis in glioma pathogenesis and plays integral roles in tumor stemness, immune evasion, and therapeutic resistance. However, a critical distinction must be made between this strong preclinical rationale and its clinically validated therapeutic efficacy, which remains limited to date. Rather than presenting broad, non-selective NOTCH inhibition as an immediately actionable clinical strategy, future translational efforts must shift toward context-aware, systems-level modulation. The field should prioritize isoform-, cell type-, and context-specific modulation of NOTCH signaling to reduce off-target toxicity and compensatory pathway activation. Integrating NOTCH targeting with biomarker-guided combination therapies, including co-targeting immune checkpoints, tumor-associated macrophages, or interacting developmental pathways, may provide a feasible strategy for the treatment of various forms of gliomas.

Glossary

ADAM10/17

α-secretases

Ab

Antibody

Akt

Protein kinase B

B7-H4

B7 Homolog 4

CB-103

NOTCH transcriptional activation inhibitor

CCDC34

Coiled-coil domain-containing protein 34

CCL2

Chemokine (C-C motif) ligand 2

CHI3L1

GBM-associated glycoprotein

CNV

Copy number variation

CREB

cAMP response element-binding protein

CSF-1

Colony-stimulating factor 1

CSF-1R

Colony-stimulating factor 1 receptor

CXCL1

C-X-C motif chemokine ligand 1

DAPT

γ-secretase inhibitor drug

DLL1

Delta-like ligand 1

DLL3

Delta-like ligand 3

DLL4

Delta-like ligand 4

DP thymocytes

Double-positive thymocytes

ECM

Extracellular matrix

EFEMP1

Epidermal growth factor-containing fibulin-like extracellular matrix protein 1

EZH2

Enhancer of zeste homolog 2

FBXW7

F-box and WD repeat domain-containing 7

FOXM1

Forkhead box protein M1

Forskolin

Adenylate cyclase activator

GBM

Glioblastoma

GI

Gastrointestinal

GSC

Glioma stem cell

GSI

γ-secretase inhibitor

HES1

Hairy and enhancer of split 1

HES5

Hairy and enhancer of split 5

HES6

Hairy and enhancer of split 6

HEY1

Hes-related family BHLH transcription factor with YRPW motif 1

HEY2

Hes-related family BHLH transcription factor with YRPW motif 2

HIF-1α

Hypoxia-inducible factor 1-alpha

HO-1

Heme oxygenase 1

ICI

Immune checkpoint inhibitor

IDH

Isocitrate dehydrogenase

IDH1

Isocitrate dehydrogenase 1

IKKα/IKKβ

Inhibitor of nuclear factor kappa-B kinase subunit alpha/beta

IFN-γ

interferon-γ

IL-10

Interleukin-10

IL-12

Interleukin-12

IL-23

Interleukin-23

INCB3619

Dual ADAM10/17 inhibitor

IRF7

Interferon regulatory factor 7

JAG1

Jagged1

JAG2

Jagged2

JAK

Janus kinase

JAK/STAT

Janus kinase/signal transducer and activator of transcription

LIF

Leukemia inhibitory factor

MAML

Mastermind-like transcriptional coactivator

MDSCs

Myeloid-derived suppressor cells

MHC-I

Major histocompatibility complex class I

MMP2

Matrix metalloproteinase-2

MMPs

Matrix metalloproteinases

MRK003

γ-secretase inhibitor drug

NF-κB

Nuclear factor kappa-light-chain-enhancer of activated B cells

NICD

NOTCH intracellular domain

NUP205

Nucleoporin 205

NUSAP1

Nucleolar and spindle-associated protein 1

PD-L1

Programmed death-ligand 1

PI3K

Phosphoinositide 3-kinase

PKA

Protein kinase A

PRC1

Polycomb repressive complex 1

RBP-J

Recombination signal binding protein for immunoglobulin kappa J region

REST

RE1-silencing transcription factor

RO4929097

γ-secretase inhibitor drug

Ror1

Receptor tyrosine kinase-like orphan receptor 1

SEMA3B

Semaphorin 3B

STAT3

Signal transducer and activator of transcription 3

SUZ12

Suppressor of zeste 12

TAMs

Tumor-associated macrophages

TGF-β

Transforming growth factor-beta

TME

Tumor microenvironment

TMZ

Temozolomide

TNF-α

Tumor necrosis factor-alpha

TP1

Tumor phenotype subtype 1

Tregs

Regulatory T cells

VDR

Vitamin D receptor

VEGF

Vascular endothelial growth factor

WDR76

WD repeat domain 76

WNT5A

Wnt family member 5A

YAP1

Yes-associated protein 1

YKL-40

GBM-associated glycoprotein

ZNF480

Zinc finger protein 480

ZR30

Engineered fibulin-3 derivative

mTOR

Mammalian target of rapamycin

oHSV

Oncolytic herpes simplex virus.

Funding Statement

The author(s) declared that financial support was not received for this work and/or its publication.

Author contributions

SE: Conceptualization, Investigation, Methodology, Supervision, Writing – original draft, Writing – review & editing. NM: Investigation, Methodology, Writing – original draft. AA: Investigation, Methodology, Writing – original draft. LY: Investigation, Methodology, Writing – original draft. MK: Investigation, Methodology, Writing – original draft. ZK: Investigation, Methodology, Writing – original draft. AG: Conceptualization, Supervision, Writing – review & editing.

Conflict of interest

The author(s) declared that this work was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Generative AI statement

The author(s) declared that generative AI was not used in the creation of this manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fonc.2026.1936605/full#supplementary-material