Metabolic-transcriptional rewiring by NFE2L2 promotes M2 macrophage polarization and anti-PD-L1 resistance in glioma Authors: Qiusi Tian, Qun Zhang, Rongzhen Deng, Shengnan Dong, Zhijun Bao, Qingjuan Chen Published: 2025 ABSTRACT AbstractGlioma is a highly malignant brain tumor, and its characteristic immunosuppressive microenvironment is closely associated with poor patient prognosis. In this study, we aimed to investigate the regulatory role of the transcription factor NFE2L2 (nuclear factor erythroid 2-related factor 2) in glioma and its underlying molecular mechanisms. Analysis based on the TCGA database and clinical samples revealed that NFE2L2 is significantly overexpressed in glioma tissues, and its expression level is closely correlated with tumor immune infiltration, particularly showing a positive association with M2-like tumor-associated macrophage (TAM) infiltration. By establishing an in vitro TAM induction model, we observed that NFE2L2 expression was upregulated during M2-like macrophage polarization. Further gene overexpression experiments demonstrated that NFE2L2 significantly enhances the migratory and invasive capabilities of glioma cells. To explore its in vivo function, we constructed an NFE2L2-deficient mouse model. The results showed that NFE2L2 deficiency led to the remodeling of the tumor immune microenvironment, thereby significantly improving the therapeutic response to PD-1 immune checkpoint inhibitors and suppressing glioma progression. Combined transcriptomic and metabolomic analyses identified differential genes and metabolites associated with NFE2L2, revealing its critical role in promoting M2-like macrophage polarization and modulating the tumor microenvironment. In summary, NFE2L2 not only plays a pivotal role in glioma development but may also serve as a potential therapeutic target. Future studies will further explore the specific mechanisms by which NFE2L2 regulates macrophage polarization and its clinical application prospects in immunotherapy.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12935-025-04113-1.Keywords: NFE2L2, Glioma, Tumor-associated macrophages, Metabolome Introduction Glioma is a highly malignant brain tumor, and its complex immune microenvironment plays a crucial role in tumorigenesis and progression. Glioma cells evade host immune surveillance through multiple mechanisms. Firstly, they secrete various cytokines and chemokines to recruit immunosuppressive cells such as regulatory T cells (Tregs) and myeloid-derived suppressor cells (MDSCs), thereby suppressing the function of effector T cells [1]. Secondly, they express immune checkpoint molecules like PD-L1 to inhibit T cell activation and proliferation, further enhancing immune evasion [2]. Additionally, the hypoxic conditions within the glioma microenvironment are considered a key factor in promoting immune escape. Under hypoxia, metabolic alterations in tumor cells modulate specific signaling pathways, leading to the upregulation of immunosuppressive factors [3]. In glioma research, NFE2L2 (nuclear factor erythroid 2-related factor 2) has garnered significant attention in recent years. Upon activation by oxidative stress or chemical stress stimuli, NFE2L2 translocates to the nucleus and binds to antioxidant response elements (AREs), initiating the expression of a series of antioxidant and detoxification genes, including NAD(P)H quinone oxidoreductase 1 (NQO1) and heme oxygenase-1 (HO-1) [4]. These genes enable tumor cells to counteract oxidative stress, thereby promoting their survival and proliferation within the reactive oxygen species (ROS)-rich tumor microenvironment [5]. Accumulating evidence indicates that NFE2L2 expression is elevated in glioma tissues compared to normal brain tissues [6–8]. Moreover, activation of NFE2L2 enables glioma cells to counteract oxidative stress and enhances their survival and proliferative capacity [9, 10]. Elevated NFE2L2 expression is also associated with chemoresistance in gliomas, particularly in patients treated with chemotherapeutic agents such as temozolomide (TMZ). NFE2L2 upregulation may reduce chemotherapeutic efficacy by augmenting cellular detoxification capabilities [11]. Although existing studies have demonstrated the significance of NFE2L2 in gliomas, its specific mechanisms in tumor-associated macrophage (TAM) polarization remain insufficiently elucidated. Current research predominantly focuses on its role in individual cell types, while systematic analyses of its impact on the interactions among diverse immune cell populations within the tumor microenvironment are lacking. Furthermore, whether NFE2L2 deficiency induces immune activation and enhances tumor sensitivity to immunotherapy represents an important avenue for investigation [12]. Therefore, in-depth exploration of NFE2L2’s role in the glioma immune microenvironment, particularly its involvement in tumor-associated macrophages and related mechanisms, holds significant implications for the development of novel therapeutic strategies.While macrophage activation is now understood as a spectrum of functional states, we use the terms "M1/M2" here operationally to refer to cells induced by classical (e.g., LPS/IFN-γ) or alternative (e.g., IL-4/IL-13) stimuli, characterized by the expression of well-established marker sets. This study employed in vivo and in vitro experiments, combined with mass cytometry, transcriptomics, and metabolomics approaches, to systematically analyze the role of NFE2L2 in the glioma microenvironment. Through functional experiments involving NFE2L2 overexpression and knockout, we investigated its effects on tumor cell migration, macrophage polarization, and tumor immune infiltration. The study aims to elucidate the function of NFE2L2 in glioma and its underlying mechanisms, providing a theoretical foundation for the development of novel therapeutic strategies. Materials and methods Ethical statements Ethical approval for this study was granted by the Ethics Committee of 3201 Hospital (Hanzhong, China; Approval No. TYYLKYJJ-2022-014, see supplementary materials). All glioma specimens were collected from patients at this institution between 2014 and 2024, following the acquisition of written informed consent. Relevant clinical characteristics of the included patients are summarized in Table S1. Animals NFE2L2-/- mice on a C57BL/6 background were obtained from Cyagen Biosciences Inc. All animal procedures were approved by the Animal Management Committee of Chongqing University and were conducted in compliance with relevant ethical guidelines. To establish the intracranial glioma model, five-week-old mice were anesthetized with isoflurane and intracranially implanted with 1×104 luciferase-expressing GL261 cells. The mice were euthanized three weeks post-implantation. Tumor growth was monitored throughout the study period using bioluminescence imaging. Following extraction, the brains were fixed in paraformaldehyde (PFA) for subsequent sectioning and analysis. Bioinformatics analysis The relationship between NFE2L2 expression and the tumor immune microenvironment was systematically evaluated. Immune cell infiltration levels were quantified using the immuneeconv algorithm, and their correlations with NFE2L2 were subsequently visualized using the ggpubr and ggExtra packages. Survival analyses (including overall survival and disease-free survival) were performed using clinical and transcriptomic data from The Cancer Genome Atlas (TCGA) glioma cohort (GBMLGG).Furthermore, Spearman correlation analysis between NFE2L2 expression and specific macrophage markers was performed primarily through the GEPIA2 database, with validation provided by additional platforms including Sangerbox, the Clinical Bioinformatics Assistant, and TIMER. Cell culture and treatment The human glioma cell lines U251 and LN229, along with the monocytic THP-1 cell line, were acquired from the Chinese Academy of Sciences Cell Bank. Cells were maintained under standard culture conditions: THP-1 cells were cultured in Pricella medium, while U251 and LN229 cells were grown in high-glucose DMEM. To generate NFE2L2-overexpressing macrophages, THP-1 cells were first differentiated into a macrophage-like state. Subsequently, 2μg of either a NFE2L2 overexpression plasmid (pCDNA3.1-NFE2L2, constructed and stored in-house) or an empty vector control (pCDNA3.1) was transfected into the cells using Lipofectamine 3000. Transfected cells were harvested 48 hours post-transfection. The efficiency of overexpression was confirmed at both the protein and transcript levels by Western blotting and quantitative PCR (qPCR), respectively. All experiments were performed in triplicate, and data are expressed as mean ± standard deviation (SD). For the polarization of M2-like macrophages, THP-1-derived M0 macrophages (induced by 24-hour PMA stimulation) were treated with recombinant human IL-4 and IL-13 (20 ng/mL each) for 48 hours. Tumor-associated macrophages (TAMs) were modeled by co-culturing THP-1 cells with conditioned media collected from U251 and LN229 glioma cell cultures, after which the resulting TAMs were collected for downstream analysis. BMDM was extracted from mice Bone marrow-derived macrophages (BMDMs) were isolated from 6–8-week-old tumor-bearing wild-type and NRF2-knockout C57BL/6J mice. Following euthanasia via cervical dislocation, the hind limbs were disinfected with 75% ethanol for 5 minutes. Femurs and tibiae were aseptically dissected from both legs, and both ends of each bone were removed. Bone marrow was flushed from the shafts using ice-cold PBS in a 1 ml syringe. After lysing red blood cells, the harvested cells were cultured for 6–7 days in complete medium supplemented with macrophage colony-stimulating factor (M-CSF) to induce macrophage differentiation. Purity of differentiated macrophages was confirmed by flow cytometry, with over 95% of cells exhibiting F4/80⁺CD11b⁺ double positivity, thus qualifying them for use in subsequent experiments. Real-time qPCR analysis Total RNA was extracted using RNAex Pro Reagent, after which cDNA synthesis was performed employing a reverse transcription kit for confirmation purposes. Subsequently, RT-qPCR was executed in accordance with established protocols [13]. β-actin was employed as the reference gene, and results were expressed as 2-△△CT. The primers used, along with their respective sequences, are presented in Table S2. Western blotting analysis Following protein extraction, concentrations were determined using a BCA assay. Proteins were separated by SDS-PAGE and transferred to PVDF membranes. After blocking for 1 hour, membranes were incubated overnight with primary antibodies, followed by a 1-hour incubation with secondary antibodies. Protein bands were visualized using an ECL detection kit. Antibody details are provided in Table S3. Immunofluorescence Immunofluorescence staining of cells and tumor tissues was performed following standard procedures. Samples were fixed, permeabilized, and blocked, then incubated overnight with primary antibodies followed by secondary antibody staining. Nuclei were counterstained, and images were acquired using a digital slide scanner. Antibody information is provided in Table S4. Immunohistochemistry and HE staining Immunohistochemical analysis was performed on 5-μm-thick sections following deparaffinization and rehydration. Antigen retrieval was conducted using EDTA, endogenous peroxidases were quenched with hydrogen peroxide, and sections were incubated with specific primary antibodies followed by DAB development. H&E staining, mouse brain tissues were fixed, paraffin-embedded, and sectioned at 10 μm before being stained with hematoxylin and eosin. All slides were imaged using a digital slide scanner. Transwell assay U251 or LN229 cells and macrophages were co-cultured in transwell chambers for 48 hours. After removing non-invaded cells, those that had invaded were fixed, stained with crystal violet, and microscopically quantified to assess invasion. Wound healing assay U251 or LN229 cells at 90% confluence were subjected to a wound healing assay. A scratch was made using a pipette tip, washed with PBS, and then cultured in macrophage conditioned medium for 24 hours before imaging. Transcriptome metabolomics sequencing analysis Collect cell samples and flash-freeze in liquid nitrogen. RNA extraction: TRIzol method was used with quality control requirements of RIN≥8.0 (Agilent 2100). Metabolite extraction was performed using pre-chilled methanol-water (4:1) for lysis, followed by nitrogen blow concentration and reconstitution in acetonitrile-water (1:1). Library construction and sequencing: For transcriptome library construction, the Illumina TruSeq strand-specific kit was selected. mRNA was enriched by Oligo-dT, fragmented, and double-stranded cDNA was synthesized, followed by PE150 mode sequencing on NovaSeq 6000 (≥20M clean reads/sample). Metabolomics analysis employed Ultimate UHPLC-C18 column separation and detection by Q-Exactive HF mass spectrometer in full scan mode (m/z50-1000) with positive/negative ion switching, using data-dependent acquisition (Top10 MS/MS). Quality control and integration: Transcriptome Q30 >85%, metabolomics QC sample RSD 1, p1, p<0.05) were subjected to integrated KEGG pathway analysis.; raw sequencing data were analyzed on OmicShare. Mass spectrometry Glioma tissues obtained from NFE2L1-/+ and wild-type (WT) C57BL/6 mice were subjected to mass cytometry (CyTOF) analysis through Puluting Health Technology (Hangzhou). Tissue digestion was performed using a Miltenyi dissociation kit, followed by filtration and staining with metal-tagged antibodies and a DNA intercalator (193Ir). Intracellular targets were detected following cell permeabilization. After normalization in FlowJo, CD45⁺ immune cells were clustered via the X-shift algorithm in R to characterize distinct immune subsets. Detailed antibody information is provided in Table S5. Statistical analysis Statistical analyses were performed using GraphPad Prism 9.5.1 and R software. Data from molecular and cellular experiments are presented as mean ± standard deviation (SD). Comparisons between two groups were analyzed using an unpaired, two-tailed Student’s t-test. For the bioinformatics analyses, the associations between NFE2L2 expression and immune cell infiltration levels or gene markers were assessed using Spearman’s rank correlation coefficient. This non-parametric method was chosen because it does not assume a normal distribution of the data and is robust to outliers, making it suitable for analyzing gene expression data. P< 0.05 was considered statistically significant. Data availability ScRNA-seq data were obtained from the Tisch database [14]. Results NFE2L2 is significantly upregulated in tumor tissues of glioma patients and closely associated with immune infiltration levels Preliminary analysis revealed that NFE2L2 could serve as a prognostic and immune biomarker, with higher expression observed in macrophages. Notably, the survival predictive capability of NFE2L2 was most pronounced in low-grade gliomas (LGG) (Supplementary Figure 1 A). Furthermore, eight distinct single-cell sequencing datasets consistently demonstrated significant upregulation of NFE2L2 in macrophages (Supplementary Figure 1B). Therefore, this study focused on gliomas (GBMLGG) to investigate the role of NFE2L2 in tumor immunity. Analysis of the TCGA database indicated that mRNA expression of NFE2L2 was significantly higher in patient tumor samples compared to normal brain tissues (Fig 1A). Examination of five matched pairs of tumor and adjacent normal tissue samples confirmed that both mRNA and protein expression levels of NFE2L2 were significantly elevated in GBMLGG patients relative to normal tissues (Figs 1B-C). Table 1 presents the correlations between NFE2L2 expression in GBMLGG and various clinicopathological characteristics. Significant upregulation of NFE2L2 in glioma tumor tissues and its close correlation with immune infiltration levels. A In the TCGA database, the expression level of NFE2L2 in glioma tissues was significantly higher than in normal tissues; B RT-qPCR analysis revealed a marked increase in NFE2L2 mRNA expression levels in human glioma samples; C Western blot experiments further confirmed elevated NFE2L2 protein expression in human glioma tissue specimens; D Association between NFE2L2 copy number variation (CNV) and immune cell infiltration levels in LGG. This analysis was conducted to test the hypothesis that genetic alterations of NFE2L2 contribute to shaping the tumor immune microenvironment. Spearman correlation analysis was used. E The copy number of NFE2L2 also demonstrated significant correlation with immune cell infiltration levels in low-grade glioma; F NFE2L2 exhibited the strongest correlation with macrophage infiltration in low-grade glioma; G The expression level of NFE2L2 was evaluated for its association with immune cell infiltration levels in GBM using Spearman correlation analysis. The correlation coefficients and statistical significance are shown.; H The expression level of NFE2L2 also showed significant correlation with immune cell infiltration levels in low-grade glioma (LGG).The experiments were performed in triplicate, and the data are presented as the means ± SDs, *P < 0.05, ***P < 0.001 vs. the control group Table 1Correlations between NFE2L2 expression and clinicopathological characteristics in GBMLGG patients(Data source: TCGA database)CharacteristicsLow expression of NFE2L2High expression of NFE2L2P valuen349350WHO grade, n (%)< 0.001 G2139 (21.8%)85 (13.3%) G3109 (17.1%)136 (21.4%) G453 (8.3%)115 (18.1%)IDH status, n (%)< 0.001 WT77 (11.2%)169 (24.5%) Mut267 (38.8%)176 (25.5%)1p/19q codeletion, n (%)< 0.001 Non-codel239 (34.5%)281 (40.6%) Codel108 (15.6%)64 (9.2%)Primary therapy outcome, n (%)< 0.001 PD43 (9.2%)69 (14.8%) SD73 (15.7%)75 (16.1%) PR45 (9.7%)20 (4.3%) CR93 (20%)47 (10.1%) Correlations between NFE2L2 expression and clinicopathological characteristics in GBMLGG patients(Data source: TCGA database) Furthermore, survival analysis revealed that patients with high NFE2L2 expression exhibited significantly worse overall survival (Supplementary Figure 1 C, P<0.001) and lower disease-free survival (Supplementary Figure 1D, P<0.001). Collectively, these results indicate that NFE2L2 expression is associated with multiple key clinical features of GBMLGG and may serve as a diagnostic and prognostic biomarker for GBMLGG. Having established that NFE2L2 expression is correlated with macrophage infiltration,we sought to investigate whether genetic alterations of the NFE2L2 gene itself could be a driving force behind this phenomenon. Gene copy number variation (CNV) is a common genetic event in cancer that can directly lead to the dysregulated expression of oncogenes or tumor suppressor genes. Therefore, we hypothesized that gains or losses in the NFE2L2 copy number might systematically influence its expression level, thereby shaping the immune cell composition in the tumor microenvironment. To test this, we analyzed the association between NFE2L2 copy number variation and the infiltration levels of various immune cells (Fig 1D). Additionally, Spearman correlation analysis revealed that NFE2L2 expression levels showed significant positive correlations with macrophage markers CD163 (Supplementary Figure 2 A, r=0.387, P<0.001), CSF1 (Supplementary Figure 2B, r=0.388, P<0.001), and CD86 (Supplementary Figure 2 C, r=0.460, P<0.001), but only a weak correlation with CD68 (Supplementary Figure 2D, r=0.096, P<0.001). Copy number variations of NFE2L2 were associated with the infiltration levels of various immune cells in the tumor microenvironment of LGG (Fig 1E) and glioblastoma (GBM) (Fig 1F). Using the TIMER database, we further analyzed the correlation between NFE2L2 expression and immune infiltration levels in gliomas. The results showed that in glioblastoma, NFE2L2 expression was positively correlated with the infiltration levels of B cells, CD4+ T cells, neutrophils, macrophages, and dendritic cells, but negatively correlated with CD8+ T cells (Fig 1G). Notably, in low-grade gliomas, NFE2L2 expression exhibited positive correlations with the infiltration levels of all the aforementioned immune cells, with stronger correlation coefficients (Fig 1H). In summary, NFE2L2 is significantly overexpressed in GBMLGG and holds prognostic significance, while also playing a crucial role in the regulation of the tumor immune microenvironment. NFE2L2 drives an immunosuppressive phenotype in tumor-associated macrophages To validate the above findings, researchers established an in vitro model of tumor-associated macrophages (TAMs). Human THP-1 monocytes were differentiated into M0 macrophages by PMA treatment for 24 hours, followed by co-culture with glioma cells (U251 or LN229) for 48 hours to obtain TAMs (Fig 2A).Real-time quantitative PCR results confirmed that the mRNA expression of canonical M2-associated markers CD163, CD206, and ARG1 was upregulated in these TAMs (Fig 2B), the in vitro-induced TAMs exhibited an immunosuppressive phenotype, characterized by the upregulation of canonical M2-associated markers (CD163, CD206, ARG1).Western blot analysis (Fig 2C) revealed a significant increase in NFE2L2 expression in both IL-4/IL-13-induced macrophages and TAMs. These results demonstrate that NFE2L2 expression is strongly associated with the formation of immunosuppressive TAMs within the glioma microenvironment..To investigate the role of NFE2L2 in macrophage polarization, we overexpressed NFE2L2 in macrophages, which resulted in a significant invasive branched morphology (Fig 2D). Concurrently, the expression levels of M2-associated markers CD163, CD206, ARG1, and IL10 were markedly upregulated following NFE2L2 overexpression (Fig 2G). Immunofluorescence assays further confirmed that NFE2L2 colocalized with CD206 in macrophages (Fig 2E), but no colocalization was observed with CD86. The expression level of NFE2L2 is significantly elevated in M2-like tumor-associated macrophages, and its high expression state promotes the polarization of macrophages toward the M2-like phenotype. A Schematic diagram of the tumor-associated macrophage generation experiment; B The relative expression levels of M2-associated markers (including CD163, CD206, ARG1, and IL-10) are significantly increased in IL-4/IL-13-induced macrophages and TAMs after 48-hour co-culture with U251 or LN229 cells. C Western blot analysis reveals that NFE2L2 protein levels are markedly upregulated in IL-4/IL-13-induced macrophages and TAMs. D Macrophages overexpressing NFE2L2 exhibit typical morphological characteristics of M2-like macrophages. E Partial colocalization is observed between NFE2L2 and the M2-associated marker CD206 within macrophages, whereas no colocalization is detected with the M1 marker CD86. F The relative expression levels of M2-associated markers (such as CD163, CD206, ARG1, and IL-10) are significantly increased in macrophages overexpressing NFE2L2. All experiments were performed in triplicate, and data are presented as mean ± standard deviation; ***P < 0.001 compared with the control group Macrophages overexpressing NFE2L2 promote glioma cell migration and invasion We further investigated the functional impact of NFE2L2-overexpressing macrophages on glioma cells. Western blot analysis confirmed successful transfection of the NFE2L2 plasmid in these cells (Fig 3A). Transwell assays demonstrated that the conditioned medium promoted the invasive capacity of U251 and LN229 cells (Fig 3B). Scratch wound assays revealed that the conditioned medium (CM) from NFE2L2-overexpressing macrophages significantly enhanced the migratory ability of U251 and LN229 cells (Fig 3C). Additionally, given the critical role of epithelial-mesenchymal transition (EMT) in glioma invasion and metastasis, treatment with CM from NFE2L2-overexpressing macrophages led to a marked decrease in the expression of the EMT marker E-cadherin, while significantly increasing the levels of N-cadherin and Snail2 (a transcription factor, noted at first mention)(Fig 3D). These results indicate that NFE2L2-overexpressing macrophages promote glioma migration and invasion, although the underlying mechanisms require further investigation. The promoting effect of NFE2L2 gene overexpression in macrophages on glioma cell migration. A Western blot analysis confirmed the successful establishment of THP-1 cell lines with NFE2L2 overexpression. B Transwell assay demonstrated that macrophages overexpressing NFE2L2 significantly enhanced the invasive abilities of U251 and LN229 cells. C Wound healing assay results indicated that NFE2L2-overexpressing macrophages promoted the migratory capacity of U251 and LN229 cells. D Macrophages overexpressing NFE2L2 regulated the expression of epithelial-mesenchymal transition (EMT)-related markers in U251 and LN229 cells. All experiments were performed in triplicate, and data are presented as mean±standard deviation; *P < 0.05 indicates statistically significant differences compared with the control group NFE2L2 deficiency suppresses glioma development Based on the potential role of NFE2L2 in the malignant progression of glioma, we hypothesized that NFE2L2 may promote glioma development in vivo. To test this hypothesis, we constructed a whole-body NFE2L2 knockout mouse model (Supplementary Figure 3A-B) to simulate the immune microenvironment with NFE2L2 deficiency in vivo. A GL261 glioma cell line stably expressing the luciferase reporter gene was used. Subsequently, GL261-Luc cells were implanted into the brains of wild-type C57BL/6 mice and NFE2L2-knockout C57BL/6 mice (Fig 4A). In vivo bioluminescence imaging revealed that the luminescence signal in the NFE2L2 homozygous group was significantly weaker than that in the control group 21 days post-implantation (Figs 4B and C). Furthermore, the measured tumor diameter and surface area in NFE2L2-/- mice were markedly lower than those in the control group (Figs 4D and E). Histological analysis via hematoxylin-eosin (HE) staining and Ki67 proliferation assays demonstrated that tumor growth and cell proliferation were significantly suppressed in NFE2L2-deficient C57BL/6 mice after GL261 glioma cell implantation (Fig 4F). NFE2L2 deficiency inhibits glioma growth in vivo. A Experimental design for constructing an orthotopic glioma mouse model. B GL261-luc cells were injected into the brains of wild-type C57BL/6 mice and NFE2L2-/- C57BL/6 mice. C Bioluminescence imaging analysis of tumor growth was performed on days 7 and 21 post-injection, with quantitative statistics (mean ±SD; n=4). D Observation of the effect of NFE2L2 knockout on tumor growth, E quantitative analysis of tumor diameter and surface area (mean±SD; n=4). F Representative images of HE staining and immunohistochemical staining (Ki67 and PD-L1),the wild-type control mice were employed simultaneously in this study and in a parallel NFE2L1 study NFE2L2 deficiency induces immune activation and enhances the sensitivity of gliomas to anti-PDL1 therapy In subsequent analyses, we sought to investigate whether alterations in NFE2L2 expression correlate with changes in the glioma immune microenvironment. Mass spectrometry-based proteomic analysis was performed on samples from wild-type and NFE2L2-/- tumor model mice, identifying 31 distinct cell subpopulations through specific cellular markers and mapping the expression profiles of 41 markers across these subpopulations to elucidate cellular functional states (Fig. 5A-C). The study revealed that NFE2L2 deficiency leads to reduced infiltration of macrophages expressing M2-associated markers and exhausted CD8+ T cells within tumors, while macrophages expressing M1-associated markers exhibit an increasing trend (Fig. 5D). Notably, a significant elevation of MHC class II molecules (major histocompatibility complex class II molecules) was observed in the NFE2L2-/- group, which is typically associated with enhanced anti-tumor immune responses and improved clinical prognosis (Fig. 5E). Mass spectrometry analysis reveals that NFE2L2 deficiency enhances immune activation in gliomas and increases sensitivity to PD-L1 therapy. A-B TSNE plots illustrate the distribution of 31 cell clusters in their respective samples. C The 31 cell clusters are classified, with each cluster defined accordingly. D Histograms display the quantity of respective cell clusters across different groups. E TSNE plots demonstrate the expression distribution of CD163, MHCⅡ, CD8, PD-L1, CD4, and CD19 in subcutaneous glioma tumors from different groups Screening and functional analysis of NFE2L2-related differentially expressed genes To elucidate the potential mechanisms by which NFE2L2 regulates the phenotypic switching of tumor-associated macrophages, we isolated Bone Marrow-Derived Macrophages (BMDMs) from both wild-type and NFE2L2-/- tumor model mice. These BMDMs were induced to differentiate using Macrophage Colony-Stimulating Factor (M-CSF) and subsequently subjected to transcriptomic and metabolomic sequencing. Principal Component Analysis (PCA) of the samples demonstrated reliable reproducibility, and 464 differentially expressed genes (DEGs) were identified (73 upregulated and 391 downregulated) (Figs 6A-B). The qPCR validation results were consistent with the RNA-seq data (Fig 6C), confirming the reliability of the sequencing results for subsequent analyses. Heatmap analysis of the top DEGs revealed that significantly altered genes included endocytosis-associated protein Epn2, Cluster of Differentiation 36 (CD36, a member of the class B scavenger receptor family, SR-B2), and Endostatin Collagen Type XVIII Alpha 1 Chain (Col18a1). These genes are closely associated with immune system endocytosis [15], pathogen recognition and clearance [16], and inflammatory angiogenesis regulation [17] (Fig 6). Notably, among the top 20 DEGs, the only upregulated gene was ST8 alpha-N-acetyl-neuraminide alpha-2,8-sialyltransferase 6 (St8sia6), which plays a critical role in regulating immune cell migration and adhesion. St8sia6 catalyzes the addition of α−2,8-linked polysialic acid (PSA) to glycoproteins such as NCAM. PSA modification reduces intercellular adhesion and promotes the migration of immune cells (e.g., T cells and macrophages) within tissues, which is essential for immune cell infiltration at inflammatory sites [18]. To explore the functional implications of the DEGs, we performed Gene Ontology (GO) and KEGG enrichment analyses. The DEGs were enriched in 7,322 GO terms, with 1,170 showing significant enrichment (P < 0.05) (Fig 6E). Further analysis revealed that GO terms related to immune regulation or immune responses were prominently enriched, including "Regulation of immune system process," "Immune system process," "Positive regulation of immune system process," "Innate immune response," "Neutrophil-mediated immunity," and other biological processes closely associated with immune modulation (Fig 6F). KEGG pathway analysis demonstrated significant enrichment of pathways related to macrophage polarization, including the PI3K/AKT, Hippo signaling pathway, and HIF-1 signaling pathway (Fig 6G). Therefore, NFE2L2 may regulate macrophage polarization in vivo through the PI3K/AKT, Hippo signaling pathway, and HIF-1 signaling pathway, thereby influencing the malignant proliferation of gliomas. Analysis of differentially expressed genes: A PCA indicates sample variance contributions, with yellow for controls and blue for NFE2L2 knockout; B Volcano plot displays log2 fold changes and FDR values, highlighting significant genes; C qPCR validates RNA-seq results; D Heatmap illustrates expression of the top 20 genes; E GO enrichment histogram shows gene counts per secondary GO term; F GO terms related to adipogenesis with P < 0.05 are mapped, showing gene ratios Differential metabolites and their functional analysis Following NFE2L2 knockout, the metabolic profile of mouse BMDMs exhibited significant alterations, and the affected metabolites could be clearly distinguished (Fig. S2). The screening criteria for differential metabolites were VIP ≥ 1 and a t-test P-value < 0.05. In positive ion mode (POS) 138 differential metabolites were identified, with 101 upregulated and 37 downregulated (Fig. 7A). In negative ion mode (NEG), 139 differential metabolites were identified, including 135 upregulated and 4 downregulated (Fig. 7B). KEGG pathway analysis of the selected differential metabolites in POS and NEG modes revealed that in POS mode, the metabolic pathways affected by NFE2L2 knockout primarily included choline metabolism in cancer, inflammatory mediator regulation of TRP channels, ether lipid metabolism, and riboflavin metabolism (Fig. 7C; S3A). In NEG mode, the impacted pathways encompassed various amino acid metabolisms (including D-amino acid metabolism, histidine metabolism, arginine and proline metabolism), purine metabolism, and central carbon metabolism in cancer (Fig. 7D). Furthermore, based on the pathway enrichment information of differential metabolites, we constructed a network to illustrate the interactions among different pathways. The core pathways in POS mode included apoptosis (ko04210), vitamin B₆ metabolism (ko00750), and glycerophospholipid metabolism (ko00564) (Fig. 7E). In NEG mode, the core pathways comprised fatty acid biosynthesis (ko00061), alanine, aspartate, and glutamate metabolism (ko00250), and sulfur metabolism (ko00920) (Fig. 7F). MSEA analysis revealed that in POS mode, pathways related to amino acid metabolism were enriched, such as valine, leucine, and isoleucine degradation; phenylalanine and tyrosine metabolism; tyrosine metabolism; glutamate metabolism; and amino sugar metabolism (Fig. 7G). In NEG mode, pathways associated with vitamin and amino acid metabolism were broadly enriched, with significant enrichment observed in vitamin B6 metabolism, glycine, and serine metabolism (Fig. 7H). Thus, NFE2L2 may regulate macrophage immune responses by altering the secretion of different metabolites and interacting with related pathways. Differential metabolites and their functional analysis. A Under POS conditions, the volcano plot and quantitative statistical chart of differential metabolites are displayed. The abscissa represents the log2 transformation of the fold change in metabolite abundance for each comparison group, while the ordinate shows the -log10 transformation of the P-value after t-test. The dashed line on the Y-axis indicates the P-value threshold for screening differential metabolites. Red dots denote upregulated differential metabolites (VIP≥1 and P<0.05, FC>1), whereas blue dots represent downregulated differential metabolites (VIP≥1 and P<0.05, FC<−1). The size of the dots is proportional to the VIP value of the metabolites. B The volcano plot and quantitative statistical chart of differential metabolites under NEG conditions are shown. C The bubble plot of KEGG enrichment under POS conditions is presented, with the ordinate indicating pathway names and the abscissa representing the enrichment factor (i.e., the ratio of the number of differential metabolites in a pathway to the total number of metabolites in that pathway). The size of the bubbles corresponds to quantity, and darker red indicates smaller Q-values. D The bubble plot of KEGG enrichment under NEG conditions is provided. E The interaction network diagram of KEGG pathways under POS conditions is illustrated. F The interaction network diagram of KEGG pathways under NEG conditions is shown. G The MSEA enrichment map under POS conditions is displayed, with the ordinate showing the names of enriched metabolite sets and the abscissa indicating the degree of enrichment. H Finally, the MSEA enrichment map under NEG conditions is presented Integrated analysis of transcriptome and metabolome We found that DEGs and DEMs significantly influenced multiple key biological processes, including Chemical carcinogenesis-DNA adducts, Alcoholic liver disease, Regulation of actin cytoskeleton, PI3K-Akt signaling pathway, Hippo signaling pathway, Glioma, and Wnt signaling pathway (Fig 8). Specifically, Comprehensive analysis of differentially expressed genes and their associated differentially expressed motors using KO genes and KEGG modules. A Displays the relative expression levels of identified KO genes across four pathways. B Modules illustrate representative KO genes through classical KEGG pathway diagrams, where green or dark dots indicate differentially expressed genes or intermediates, while blue rounded rectangles represent indirectly affected biological processes. Red boxes highlight those KO modules that underwent changes (corresponding to differentially expressed genes). Bar charts present the relative abundance of each KO module In chemical carcinogenesis-DNA adducts, changes in three KO modules were observed, including NAD+-dependent farnesol dehydrogenase (Cyp1b1) and Glutathione S-Transferase Mu 1/Mu 2 (Gstm1/Gstm2). In the regulation of actin cytoskeleton, we observed downregulation of nine modules involving various enzymes such as fibronectin 1 (Fn1), farnesyl diphosphate synthase (pdgfra), and isopentenyl diphosphate Delta-isomerase (itgad), which collectively influenced the regulation of actin cytoskeleton. In the PI3K/AKT signaling pathway, we observed downregulation in 9 out of 10 KO modules, involving multiple receptor tyrosine kinases and their associated molecules, such as Pdgfrb and Pdgfra, which are the β and α subunits of platelet-derived growth factor receptors. These receptors enhance glutaminase expression through the Ras-MAPK pathway, supporting tumor biosynthetic demands [19]. Lrs1, as a leucyl-tRNA synthetase, directly influences the utilization efficiency of branched-chain amino acids (leucine, isoleucine) [20]. Leucine activates the mTORC1 pathway, promoting nucleotide and lipid synthesis, and its aberrant expression is associated with metabolic disorders [21–23]. Following NFE2L2 knockout, the expression levels of these kinases were consistently reduced, suggesting that NFE2L2 may inhibit tumor progression by restricting amino acid metabolism in tumor cells. This hypothesis requires further experimental validation. Discussion NFE2L2 plays a pivotal role in the immune microenvironment of gliomas. Its regulatory effects not only influence tumor cell growth and survival but also profoundly impact immune cell function. In recent years, increasing studies have revealed the dual role of NFE2L2 in gliomas: on one hand, it enhances the antioxidant capacity and viability of glioma cells; on the other hand, it may also affect tumorigenesis and progression by remodeling the immune microenvironment. This complex interplay positions NFE2L2 as an emerging therapeutic target in glioma biology research. In this study, we found that NFE2L2 was significantly upregulated in tumor tissues of glioma patients, and its expression level was closely associated with the immune infiltration status of the tumor, particularly in regulating the polarization of tumor-associated macrophages (TAMs). This suggests that NFE2L2 may play a critical role in the tumor microenvironment. As a key factor for immunosuppressive TAMs, our data indicate that NFE2L2 drives macrophages toward a functional state commonly associated with immunosuppression and tumor progression (often characterized by the expression of M2-associated markers). In our experiments, macrophages overexpressing NFE2L2 significantly enhanced the migration and invasion of glioma cells, indicating that NFE2L2 may regulate tumor malignancy by influencing macrophage phenotype and function. These findings not only reveal the importance of NFE2L2 in glioma development but also provide new therapeutic targets, particularly for modulating immune cells within the tumor microenvironment. Programmed death-1 (PD-1) and its ligand programmed death-ligand 1 (PD-L1) are key molecules mediating tumor immune escape, with PD-L1 expression levels closely associated with the formation of an immunosuppressive tumor microenvironment [24]. Although PD-1/PD-L1 inhibitors have demonstrated significant efficacy in various solid tumors, clinical studies indicate their therapeutic response rates in gliomas remain suboptimal [25]. We found that this therapeutic challenge is linked to the immune microenvironment characteristics regulated by NFE2L2: NFE2L2 deficiency not only suppresses the initiation and progression of gliomas but also enhances tumor sensitivity to PD-L1 inhibitors. The underlying mechanism involves reduced infiltration of macrophages with an M2-like phenotype and increased CD8+ T cells and M1 macrophages upon NFE2L2 deletion. This finding is particularly significant, as previous studies have confirmed that the accumulation of such immunosuppressive macrophages is directly associated with PD-L1 therapy resistance [26]. More importantly, NFE2L2 modulates the immune microenvironment through multi-target regulation: in addition to influencing macrophage polarization and T-cell function, it enhances B-cell immune responses (promoting IgM production while suppressing activation marker expression) [27] and improves the antitumor activity of NK cells by regulating metabolism and antioxidant responses [28]. These findings demonstrate that NFE2L2 regulates immune cell function in the tumor microenvironment through multiple mechanisms, impacting tumor progression and immune evasion. They also lay the foundation for developing novel immunotherapeutic strategies, particularly for patients with NFE2L2-deficient tumors, who may benefit more from immunotherapy. Finally, through integrated transcriptomic and metabolomic analyses, we identified NFE2L2-associated differentially expressed genes and their roles in macrophage polarization. Enrichment analysis of these differentially expressed genes revealed that NFE2L2 may influence macrophage polarization states by regulating the PI3K/AKT, Hippo, and HIF-1 signaling pathways. For instance, NFE2L2 activation upregulates the expression of multiple antioxidant genes, thereby reducing intracellular oxidative stress levels and indirectly enhancing PI3K/AKT pathway activity [29, 30]. Studies also indicate that NFE2L2 activation can suppress the activity of MST1/2, the core kinases of the Hippo pathway, leading to YAP dephosphorylation and activation, which promotes cell proliferation [31]. Furthermore, NFE2L2 can indirectly modulate Hippo signaling by influencing cellular redox status, particularly within the tumor microenvironment [30]. Additionally, NFE2L2 enhances AKT phosphorylation through its interaction with PI3K, thereby promoting cell survival and proliferation [32, 33]. This interaction is particularly critical in the tumor microenvironment, where cancer cells frequently encounter dual challenges of hypoxia and high oxidative stress, and NFE2L2-mediated regulation provides a biological advantage for tumor cell survival [34, 35]. The findings suggest that these signaling pathways are not only involved in the metabolic regulation of tumor cells but also closely associated with immune responses. The metabolic characteristics of gliomas play a crucial role in their microenvironment. Studies have demonstrated that glioma cells reprogram metabolic pathways to meet the energy and biosynthetic demands required for their rapid proliferation [36, 37]. For instance, glioma cells often exhibit enhanced glycolysis and lactate metabolism [38–40], which not only provide energy for tumor cells but also suppress the function of surrounding immune cells through lactate accumulation. Additionally, significant alterations occur in the amino acid metabolism of glioma cells, with tryptophan metabolism playing a pivotal role in modulating immune responses and promoting tumor growth [41]. Metabolomic analysis in this study revealed that NFE2L2 deficiency can induce metabolic reprogramming by downregulating multiple amino acid metabolic pathways, such as the degradation of valine, leucine, and isoleucine, phenylalanine and tyrosine metabolism, glutamate metabolism, and amino sugar metabolism, thereby blocking the energy supply to gliomas and inhibiting tumor proliferation. These metabolic features not only influence the biological behavior of tumors but may also serve as novel therapeutic targets to improve glioblastoma treatment outcomes through metabolic intervention. Thus, the regulatory role of NFE2L2 may provide new insights into understanding the immune mechanisms within the tumor microenvironment and their role in tumor progression, offering an important theoretical foundation for future clinical applications. Although current research findings support the potential of NFE2L2 as a therapeutic target for glioma, there are certain discrepancies in its mechanisms of action and effects across different studies. For instance, some studies suggest that excessive activation of NFE2L2 may lead to immunosuppression, thereby promoting tumor progression, while others indicate that NFE2L2 can enhance immune responses under specific conditions. Therefore, future research should focus on elucidating the interaction mechanisms between NFE2L2 and the glioma immune microenvironment, which is crucial for understanding its role in tumorigenesis and progression. In summary, NFE2L2 plays a pivotal role in the initiation and progression of glioma, with its high expression closely associated with tumor malignancy and immune infiltration. The findings demonstrate that NFE2L2 may promote glioma growth and invasion by modulating the polarization states of various immune cells, including macrophages. These results provide substantial evidence supporting NFE2L2 as a potential therapeutic target and offer new directions for future research to further explore its application prospects in glioma immunotherapy. Supplementary Information Supplementary material 1. Supplementary material 2. Supplementary material 3. Author contributions Qiusi Tian and Qun Zhang contributed to formal analysis, data curation, conceptualization, investigation, methodology, and the writing and editing process of the review. Qiusi Tian was responsible for the collection of clinical data, acquisition of funding, and conducting animal experiments. Methodological contributions were made by Rongzhen Deng,Qingjuan Chen,Zhijun Bao and Shengnan Dong. Qun Zhang took the lead in project administration and oversaw the writing and review process. Funding This study was supported by grants from the Natural Science Foundation of Shaanxi Province (Project 2023-JC-QN-0992 awarded to Tian Qiusi) and the Key Project of General Medical Research (Project TYYLKYJJ-2023-014 awarded to Tian Qiusi). Data availability ScRNA-seq data were obtained from the Tisch database [14]. The publication, along with the supplementary materials available online, contains all the necessary data required for assessing the conclusions drawn in the paper. Furthermore, any additional data pertinent to this study can be obtained by reaching out to the corresponding author, who can be contacted via the following email addresses: bjzhangqun@126.com Declarations Competing interests The authors declare no competing interests. Footnotes Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Qiusi Tian and Qun Zhang These authors contributed equally to this work. References Associated Data Supplementary Materials Supplementary material 1. Supplementary material 2. Supplementary material 3. Data Availability Statement ScRNA-seq data were obtained from the Tisch database [14]. ScRNA-seq data were obtained from the Tisch database [14]. The publication, along with the supplementary materials available online, contains all the necessary data required for assessing the conclusions drawn in the paper. Furthermore, any additional data pertinent to this study can be obtained by reaching out to the corresponding author, who can be contacted via the following email addresses: bjzhangqun@126.com