Abstract Glioblastoma (GBM) harbors a highly inflammatory microenvironment driven predominantly by activated innate immune cells, despite being classified as an immunologically cold tumor due to limited T cell infiltration. Tumor-associated macrophages (TAMs) are key contributors to disease progression, in part through their production of interleukin-1β (IL-1β), a pro-inflammatory cytokine with tumor-promoting functions. However, the precise role of IL-1β⁺ TAMs in GBM remains incompletely understood. This study aimed to elucidate the functional contributions of IL-1β⁺ TAMs to GBM malignancy and to explore their therapeutic relevance. Single-cell RNA sequencing (scRNA-seq) analysis revealed that IL-1β⁺ TAMs were enriched in GBM tissues compared to normal brain tissue, with their elevated infiltration correlating with aggressive tumor phenotypes and poor prognosis. Functionally, IL-1β stimulated GBM cells to secrete inflammatory mediators such as PGE2 and TNFα. These mediators, in turn, upregulated C/EBPβ expression in macrophages, thereby enhancing IL-1β transcription. Mechanistically, tumor-derived PGE2 and TNFα synergistically activated C/EBPβ via EP4 receptor signaling, initiating a self-sustaining IL-1β-PGE2/TNFα-EP4-C/EBPβ-IL-1β feedback loop that amplified pro-inflammatory crosstalk between GBM cells and TAMs. Disruption of PGE2/EP4 signaling effectively suppressed IL-1β^+ TAM generation and attenuated tumor growth in preclinical models. Our finding highlights how GBM cells induce macrophages to secrete IL-1β through the synergistic action of PGE2 and TNFα via the EP4 receptor and C/EBPβ activation. This feedback loop between tumor cells and macrophages fosters a pro-inflammatory TME that drives GBM progression. Targeting the PGE2/EP4-C/EBPβ signaling axis may therefore present a promising immunotherapeutic strategy to disrupt tumor-TAM crosstalk and suppress GBM progression. Supplementary Information The online version contains supplementary material available at 10.1186/s12974-025-03551-y. Keywords: Glioblastoma microenvironment, IL-1β^+ macrophages, PGE2, TNFα, C/EBPβ Introduction Glioblastoma (GBM), the most common and aggressive primary central nervous system tumor, may originate from astrocytes, neural stem cells or oligodendrocyte progenitor cells [[44]1]. The standard treatment for GBM includes maximal safe surgical resection, concurrent radiotherapy, and adjuvant chemotherapy (such as temozolomide). Emerging therapies, such as tumor-treating fields (TTFields) and targeted therapies, have also been introduced. Despite these interventions, the prognosis for GBM remains poor, with a median survival of approximately 15 months and a 5-year survival rate below 6% [[45]2, [46]3]. A critical factor contributing to the failure of current therapies is the complex tumor microenvironment (TME), which is rich in non-tumor cells that drive cancer progression [[47]4–[48]6]. The TME of GBM is a highly heterogeneous ecosystem composed of tumor cells, non-tumor cellular components, and vascular structures [[49]4, [50]7, [51]8]. Among the non-tumor cells, TAMs, derived from microglia and bone marrow-derived macrophages (BMDMs), constitute the largest cellular population, accounting for up to 50% of the total cells within the TME [[52]9, [53]10]. TAMs play a pivotal role in GBM progression by releasing a diverse array of growth factors, chemokines, and cytokines [[54]8, [55]11]. TAMs can also contribute to matrix metalloproteases expression and activity, which can promote tumor cell invasion and metastasis [[56]12, [57]13]. Notably, targeting TAMs in preclinical GBM models has been shown to modulate malignant progression, positioning TAMs as promising therapeutic targets. However, the molecular mechanisms underlying the contribution of TAMs to GBM malignancy remain poorly understood. Inflammation plays a significant role in various cancers, including GBM, where the inflammatory cytokine IL-1β is often elevated in the TME [[58]14, [59]15]. Excessive expression or sustained release of IL-1β within the TME can lead to chronic inflammation, which promotes tumor initiation and progression [[60]16–[61]18]. IL-1β has been shown to induce the expression of adhesion molecules, such as ICAM-1 and VCAM-1, in GBM, facilitating adhesive interactions between GBM cells and monocytes [[62]19]. Furthermore, Chen et al. reported a feedforward paracrine circuit involving IL-1β and its receptor (IL-1R1) between tumor cells and monocyte-derived macrophages, creating an interdependence that drives PDGFB-mediated GBM progression [[63]20]. Thus, both GBM cells and macrophages contribute to elevated IL-1β levels within the TME. However, the precise origin of IL-1β remains unclear, as the TME is predominantly anti-inflammatory and immunosuppressive under normal conditions. In this study, we identified a significant association between increased IL-1β-expressing TAMs and reduced patient survival in GBM. Our findings indicate that GBM cells remodel macrophages to produce IL-1β through the secretion of tumor-derived prostaglandin E2 (PGE2) and tumor necrosis factor (TNF). Specifically, we discovered that PGE2 programmed macrophages to express IL-1β by signaling through the EP4 receptor, leading to upregulation of the C/EBPβ transcription factor. This programming enhances the ability of macrophages to produce IL-1β, which, in turn, promotes the secretion of key chemokines and cytokines by GBM cells. In preclinical mouse cancer models, blocking EP4 effectively downregulated IL-1β expression in macrophages and delayed tumor growth. These results reveal an inflammatory loop between GBM cells and IL-1β-expressing TAMs that drives tumor progression. Our findings highlight the therapeutic potential of targeting EP4 to disrupt this loop and promote more effective anti-tumor strategies. Materials and methods Reagents The compounds and reagents used in this study include: Phorbol 12-myristate 13-acetate (PMA, P8139, Sigma-Aldrich), COX-2 inhibitor (NS398, HY-13913, MedChemExpress), EP4 antagonist (ONO-AE3-208, HY-50901, MedChemExpress), PKA inhibitor (H-89, HY-15979, MedChemExpress). Recombinant human TNFα (HY-P7058, MedChemExpress), Prostaglandin E2 (PGE2, HY-101952, MedChemExpress), Recombinant human IL-1β (211-11B, PeproTech), lipopolysaccharide(LPS, L439, Sigma-Aldrich). ScRNA-seq analysis ScRNA-seq data matrices and metadata were obtained from nine newly diagnosed IDH wild-type GBM patients (Gene Expression Omnibus accession number [64]GSE182109) [[65]8] and four normal brain tissue samples (Gene Expression Omnibus accession number [66]GSE190453) [[67]21] for sample analysis. After acquisition, all datasets were processed using R-studio (v4.3.2). The Seurat package (v4.4.0) was utilized for quality control, analysis, and clustering of all datasets [[68]22]. Low-quality cells were removed based on manually defined criteria: gene counts between 200 and 70,000, at least 10 cells with more than 10 genes, and mitochondrial fraction exceeding 20%. Remaining cells were then filtered, combined, and converted into a Seurat object. To eliminate batch effects across different samples, the Harmony algorithm (v1.2.0) was applied [[69]23]. The top 30 Harmony embeddings were selected using the “ElbowPlot” function for cell clustering, followed by cluster determination using “FindClusters”. Based on a single-cell dataset, use the FindMarkers Function from the Seurat package to identify cluster-specific marker genes with a resolution parameter ranging from 0.1 to 1. Perform dot plot and feature plot visualizations to examine the expression of genes specific to each cluster. Determine cell types by inspecting the genes that are specifically expressed within each cluster. Finally, to evaluate the correlation between macrophage subpopulation characteristics and survival in GBM patients, we utilized the GBM dataset from The Cancer Genome Atlas (TCGA, [70]http://cancergenome.nih. gov) and applied the BisqueRNA package to calculate the relative abundance of macrophage subpopulations in single cells, mapping these features onto the GBM dataset to obtain survival curve characteristics. Human tissue samples All patient tissue samples were collected from the Department of Neurosurgery at the Affiliated Guangdong Second Provincial General Hospital of Jinan University after signing informed consent between January 2020 and June 2024. Pathological results were diagnosed by at least two senior pathologists from the Affiliated Guangdong Second Provincial General Hospital of Jinan University. This study adhered to the principles outlined in the Declaration of Helsinki and was approved by the Ethics Committee of the Affiliated Guangdong Second Provincial General Hospital of Jinan University (Ethics No. 2023-DW-KZ-080-01). Cell lines and culture conditions The T98G and THP1 cells were gifted by the Department of Neurosurgery, Sun Yat-sen University Cancer Center (Guangzhou, China), while the U251 cell line was purchased from Procell Life Science & Technology Co., Ltd. (Wuhan, China) and authenticated. All cells were cultured in a humidified incubator at 37 °C with 5% CO2 and 95% air. Regular mycoplasma testing was conducted. U251 and T98G cells were cultured in Dulbecco’s Modified Eagle’s Medium (DMEM, C11885500BT, Gibco) supplemented with 10% fetal bovine serum (FBS, MK1123-500 C, MIKX) and 1% penicillin-streptomycin solution (15140122, Gibco). THP1 cells were cultured in RPMI1640 medium (C11875500BT, Gibco) supplemented with 10% FBS and 1% penicillin-streptomycin solution. To induce THP1-derived macrophages (THP1 Macs), THP1 cells were stimulated with 80 ng/mL PMA for 24 h to use for subsequent experiments. Collection of GCM For conditioned medium collection, when T98G or U251 cells density reached about 80%, the culture medium was replaced with serum-free medium supplied with 1% penicillin/streptomycin. 48 h later, the culture supernatant was collected after cellular debris was removed by centrifugation (3000 rpm, 10 min) and filtration with 0.22 μm filters (SLGP033RB, Merck Millipore). This supernatant was designated as GBM-conditioned medium (GCM). For IL-1β-stimulated tumor supernatant collection, and U251 cells were stimulated with recombinant human IL-1β (1ng/ml) for 48 h. After replacing the medium with serum-free medium, the supernatant was collected 48 h later and named as ^IL−1βGCM for subsequent experiments. Enzyme‑linked immunosorbent assay (ELISA) For quantification of IL-1β, THP1 Macs were stimulated for 48 h, as indicated. Supernatants were collected and stored at −80℃ until use. IL-1β (437004, BioLegend) was measured following the manufacturer’s instructions. PGE2 and TNFα levels were quantified in the supernatants of cell lines. Cells were seeded at 2.5 × 10^5 cells per well in 6-well plates and cultured for 48 h in 2 ml of complete medium to collect supernatants. ELISAs were performed to detect PGE2 (JM-03317H1, JINGMEI Biotechnology) and TNFα (JM-03277H1, JINGMEI Biotechnology) following manufacturer’s instructions. When PGE2 levels in tumor tissue were quantified using ELISA, the value was normalized to tissue weight. Flow cytometry To analyze IL-1β expression in macrophages stimulated by tumor supernatants, THP1 Macs were stimulated with GCM for 48 h and then analyzed by flow cytometry. Monensin (420701, BioLegend) was added to the culture medium 5 h before cell harvest. Cells were washed with FACS buffer (PBS containing 2% FBS) and preincubated on ice for 10 min with human Fc receptor-blocking antibodies (anti-CD16/CD32, 101302, BioLegend). Cells were then fixed, permeabilized, and stained with anti-human IL-1β (Pacific Blue, 511709, BioLegend) for 30 min at 4 °C. After staining, the cells were washed to remove unbound antibodies, resuspended in staining buffer, and analyzed using a FACS Aria III (BD Biosciences). Quantitative real-time PCR (RT-PCR) Total RNA was extracted from cells using the RNA Quick Purification Kit (RN001, ES Science), and cDNA was synthesized using HiScript III RT SuperMix (R323-01, Vazyme). RT-PCR was performed using ChamQ SYBR qPCR Master Mix (Q311-03, Vazyme). Gene expression levels were normalized to the endogenous reference gene ACTB using the 2^−ΔΔCt method, and reactions were run on a QuantStudio 6 instrument (Thermo Fisher). Primer sequences are listed in Supplementary Table [71]S1. Western blotting THP1 Macs were collected and lysed using RIPA buffer containing protease inhibitors (BC3710, Solarbio). The lysates were incubated on ice for 30 min and then centrifuged at 12,000 g at 4 °C for 1 h to remove insoluble material. Total protein concentrations were determined using a BCA Protein Assay Kit (CW0014S, Cowin Biotech) to ensure equal protein loading across samples. Equal amounts of protein samples were mixed with 5 × loading buffer (BL502A, Biosharp) and boiled for 10 min to denature the proteins. Precast Gels Plus Bis-Tris gels (PSB2001-12 F, WSHT) were used for electrophoresis, which was performed at a constant voltage of 120 V until the protein marker (26616-1, Thermo Fisher) reached the bottom of the gel. After electrophoresis, the gel was transferred to a PVDF membrane (BS-PVDF-45, Merck Millipore) using a wet transfer apparatus with transfer buffer (BL626A, Biosharp) under conditions of constant current (400 mA) for 1 h. The transferred membrane was then blocked with Western Blot Rapid Blocking Buffer (BL1032A, Biosharp) for 20 min at room temperature on a shaker. After discarding the blocking buffer, the membrane was incubated overnight at 4 °C with primary antibodies against C/EBPβ (83791-6-RR, Proteintech, 1:1000 dilution) or GAPDH (10494-1-AP, Proteintech, 1:1000 dilution) while gently shaking. The membrane was washed three times with TBST buffer (T1087, Solarbio) to remove any unbound primary antibodies. Following this, the membrane was incubated with HRP-conjugated secondary antibodies (7074 S, Cell Signaling Technology, 1:1000 dilution) for 1 h at room temperature. The membrane was washed again three times with TBST buffer, each for 5 min, to eliminate unbound secondary antibodies. Finally, the membrane was treated with ECL chemiluminescent substrate (BL520A, Biosharp) and imaged using a chemiluminescence imaging system. Immunofluorescence staining For human tissue and mouse tumor samples, tissues were fixed in 4% paraformaldehyde (BL539A, Biosharp) for 24 h and dehydrated through a sucrose gradient. Sections were embedded, and antigen retrieval was performed using citrate buffer (G1202, Servicebio). Subsequently, endogenous peroxidase was blocked with 3% H[2]O[2] (CS7730, G-CLONE), and cell permeabilization was facilitated using 0.1% Triton X-100 (T8200, Solarbio). Sections were blocked with 5% BSA (4240GR100, Biofroxx) and incubated with primary antibodies against CD68 (25747-1-AP, Proteintech, 1:500 dilution), IL-1β (66737-1-IG, Proteintech,1:500 dilution), Ki67 (27309-1, Proteintech, 1:500 dilution) overnight at 4 °C. The next day, sections were incubated with fluorescent secondary antibodies (RGAM002 and RGAR004, Proteintech, 1:500 dilution) at room temperature in the dark for 1 h, followed by DAPI (C1002, 1:1000 dilution) staining for 20 min. Sections were then mounted with anti-fade mounting medium (BL701A, Beyotime) and cover slipped. For cell immunofluorescence staining, cells were first seeded on slides, washed with PBS, and fixed with pre-cooled 4% paraformaldehyde. Cells were then permeabilized using 0.1% Triton X-100 to facilitate cell permeabilization and blocked with 5% BSA. Subsequently, diluted primary antibodies against CD68 (25747-1-AP, Proteintech, 1:500 dilution), IL-1β (Proteintech, 1:500 dilution) were applied and incubated overnight at 4 °C. The next day, sections were incubated with fluorescent secondary antibodies (66737-1-IG, Proteintech, 1:500 dilution) at room temperature in the dark for 1 h, followed by DAPI (C1002, Beyotime, 1:1000 dilution) staining for 20 min. Sections were then mounted with anti-fade mounting medium and cover slipped. All fluorescence images were captured using a confocal microscope (Leica) and analyzed with LAS X software (Leica) and ImageJ (NIH) [[72]24]. Bulk RNA-Seq analysis For THP1 RNA-Seq analysis, THP1 Macs were cultured with or without PGE2 (1nM) and TNFα (10ng/ml) or GCM. After 24 h, cells were harvested and lysed in TRIzol (T9424, Sigma). For U251 RNA-Seq analysis, U251 cells were cultured with or without IL-1β (1ng/ml). After 24 h, cells were harvested. Total RNA was extracted from all samples using Trizol. RNA libraries were prepared and sequenced using the Illumina PE150 platform by Guangzhou IGE. Raw data were mapped to the human genome, and transcriptome quantification was performed for each sample group. Pathway analysis between samples was conducted using GO, KEGG, and GSEA. In vivo subcutaneous xenograft model All animal experiments were conducted in strict adherence to the ethical principles and guidelines for animal care and approved by the Animal Ethics Committee of the Second Provincial Hospital of Guangdong Province (Ethics No.2024-DW-KZ-068-01). BALB/c nude mice, aged 6–8 weeks, were purchased from Zhuhai BestBio Biotechnology Co., Ltd. and housed in specific pathogen-free (SPF) animal facilities. For subcutaneous xenograft models, 5 × 10^6 U251 cells were injected subcutaneously into the right flank of BALB/c nude mice. When the tumors reached an approximate volume of 100 mm^3, the mice were randomly divided into three groups (G, GM and GMO). Groups GM and GMO received intravenous injections of 1 × 10^6 THP1 Macs in 50ul PBS every three days. Group GMO also received intraperitoneal injections of ONO-AE3-208 at a dose of 10 mg/kg daily. Tumor size was measured every two days using the formula: volume = length×width^2/2. Mice were euthanized nine days after drug injection, and tumors were collected, weighed, and preserved for further study. Statistical analysis All experiments were repeated at least three times. Data were analyzed using GraphPad Prism 9 (GraphPad Software). Comparisons between two groups were performed using Student’s t-tests. Data are presented as mean ± standard error of the mean (SEM). Correlation and survival analyses using the TCGA dataset were performed using the Log-rank (Mantel-Cox) test. Statistical significance was set at P < 0.05. Results Increased IL-1β^+ TAMs in GBM are associated with reduced patient survival ScRNA-seq data from 13 individuals (9 with IDH wild-type GBM tumor tissues and 4 with normal brain tissues) were analyzed, encompassing 97,951 cells in total—41,177 from normal brain tissue and 56,774 from tumor tissue. Unsupervised clustering algorithms identified eight distinct cellular clusters based on gene expression profiles (Fig. [73]S1A). Through a combination of literature review [[74]8, [75]10, [76]25, [77]26] and analysis of gene expression patterns, we classified these clusters into seven major cell types: GBM cells (cluster 0, expressing EGFR, GFAP, SOX2, and S100B), myeloid cells (cluster 1, expressing PTPRC/CD45, C1QA, C1QB, and C1QC), T cells (cluster 5, expressing PTPRC/CD45, CD3D, CD3E, and CD3G), neurons (clusters 3 and 4, expressing MAP2, RBFOX1, DLGAP2, and CAMD1), pericytes (cluster 6, expressing PDGFRB), endothelial cells (cluster 7, expressing FN1 and SPARC), and oligodendrocytes (cluster 2, expressing MBP, PLP1, MOBP, and CTNNA3) (Fig. [78]1A, Fig. [79]S1B). Consistent with previous studies [[80]27–[81]29], the GBM microenvironment is composed of non-neoplastic cells and is enriched with myeloid cells (35.39%), along with a higher proportion of T cells (9.08%) compared to normal brain tissue (Fig. [82]1B). Fig. 1. [83]Fig. 1 [84]Open in a new tab IL1B^+ TAMs are associated with poor prognosis in GBM patients. (A) t-Distributed Stochastic Neighbor Embedding (t-SNE) plots from single-cell RNA sequencing of normal human brain tissues (n = 4) and tumor tissues (n = 9), with colors representing different cell clusters. (B) Percentage of different subpopulations in normal human brain tissues and GBM tissues. (C) t-SNE plot of myeloid cell subpopulations in GBM tissues (n = 9), with colors representing distinct myeloid cell subclusters. (D) Heatmap of the top 50 genes for each of the 12 myeloid cell subclusters. (E) Pie chart showing the proportions of different myeloid cell subpopulations within the total myeloid cell population in GBM patients (n = 9). (F) Kaplan‒Meier analysis of overall survival rate according to the IL1B^+ Mac expression level in GBM tissues in a TCGA cohort. (G) Representative IF staining of IL-1β (green) and CD68 (red) in tissues from normal individuals (n = 14), LGG (n = 14), and GBM (n = 17) patients, Scale bar, 50 μm. (H) The proportion of macrophages (CD68) among total cells (left panel) and the proportion of IL-1β^+ TAMs among total macrophages (right panel) in tumor tissue from LGG and GBM patients. (I) Correlation analysis of IL-1β and CD68 expression in GBM patient tissues with IF stained. (Data are presented as the Mean ± SEM. R²: Coefficient of Determination. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) The abundant infiltration of myeloid cells within TME suggests their critical involvement in GBM progression. We then explored myeloid cell population to gain deeper insights into their molecular characteristics. Hallmark pathway analysis of myeloid cell clusters revealed significant enrichment in pathways such as IL6_JAK_STAT3_SIGNALING, INFLAMMATORY_RESPONSE, COMPLEMENT, and TNFA_SIGNALING_VIA_NFKB (Fig. [85]S1C). Further analysis of myeloid cells from the 9 GBM patient datasets revealed additional subpopulations, including four microglia clusters (Mic), four macrophage clusters (Mac), two dendritic cell (DC) clusters, one monocyte cluster, and one neutrophil cluster (Fig. [86]1C-D). Among the macrophage subpopulations, distinct gene expression profiles were observed. Specifically, Cluster Mac-1, characterized by high IL1B expression and enriched for inflammation- and immune-suppressive-related genes, was predominant in most patients (Fig. [87]1E). Cluster Mac-2 exhibited proliferation-related gene expression, marked by MKI67, while Cluster Mac-3 was associated with angiogenesis (VCAN expression). Cluster Mac-4 was marked by high APOE expression, associated with lipid metabolism and oxidative phosphorylation (Fig. [88]1D). To assess the clinical relevance of these myeloid subtypes, we integrated bulk RNA sequencing data and survival information from The Cancer Genome Atlas (TCGA) database for GBM patients. Using deconvolution algorithms, we mapped myeloid cell subtype gene signatures to TCGA datasets. Notably, IL1B^+ Macs were significantly correlated with reduced patient survival (Fig. [89]1F). Through scRNA-Seq, we can observe that CD68 is expressed in myeloid cells (Fig. [90]S1D). Immunofluorescence analysis of human GBM specimens confirmed the co-expression of IL1β with CD68, a pan-macrophage marker, further validating the presence of IL1β^+ TAMs in GBM (Fig. [91]1G). GBM patients exhibited a higher percentage of IL-1β^+ CD68^+ cells in TME compared to LGG patients (Fig. [92]1H), suggesting that IL-1β^+ TAMs were positively correlated with the progression of GBM. Analysis of fluorescence specimens in GBM patients reveals that IL-1β expression positively correlates with macrophage infiltration capacity within the TME (Fig. [93]1I). These data indicate that a subpopulation of inflammatory TAMs characterized by high IL-1β expression is positively associated with poor outcome in GBM. Secreted factors from GBM induce an IL-1β-expressing macrophage phenotype Macrophages exhibit significant plasticity, enabling them to modify their phenotype and function in response to the surrounding microenvironment [[94]8, [95]17, [96]30]. Within the GBM microenvironment, macrophages are reprogrammed by GBM-derived factors to promote tumor progression [[97]6, [98]31, [99]32]. To explore how GBM cells induce TAMs to secrete the pro-inflammatory cytokine IL-1β, GBM cell-conditioned medium (GCM) was collected and used to stimulate macrophages derived from THP1 cells (THP1 Macs). LPS (100 ng/mL) served as a positive control. IL-1β mRNA in macrophages was measured at various time points. In response to LPS stimulation, IL-1β mRNA expression increased rapidly within 1 h, peaking at 2 h. In contrast, THP1 Macs exposed to GCM exhibited delayed but robust IL-1β expression, with significantly higher levels observed at 24 h post-exposure to GCM derived from both U251 and T98G cells (Fig. [100]2A). Fig. 2. [101]Fig. 2 [102]Open in a new tab GCM induces macrophages to secrete IL-1β. (A) IL1B mRNA expression in THP1 Macs was measured after stimulation with GCM from U251 and T98G cells, as well as LPS (100 ng/ml) at various time points. Statistical significance is indicated by * compared to the control group (n = 3). (B) ELISA analysis of IL-1β protein levels in the supernatant of THP1 Macs after 48 h of indicated stimulation (n = 6). (C) Flow cytometry analysis showing representative plots (left) and the quantification of IL-1β^+ cells as a percentage of total cells (right) in THP1 Macs after 48 h of stimulation (n = 4). (D) Representative immunofluorescence images of IL-1β and CD68 expression in THP1 Macs after 48 h of stimulation (n = 5). Blue: DAPI, Green: IL-1β, Red: CD68. The fluorescence intensity of IL-1β in different groups (upper right) and the percentage of IL-1β + cells among CD68^+ cells (lower right). (Data are shown as the Mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) Additionally, an ELISA assay revealed a significant increase in IL-1β protein levels in the supernatants of macrophages treated with GCM from U251 or T98G cells compared to controls (Fig. [103]2B). To exclude the possibility that IL-1β originated from the GCM itself, we measured IL-1β levels in GCM and found minimal baseline production by GBM cells (< 10 pg/ml) (Fig. [104]S2A-B), indicating that IL-1β detected in supernatants was derived from the macrophages. Flow cytometry further confirmed an elevated proportion of IL-1β^+ Macs in the GCM-treated group compared to untreated controls (Fig. [105]2C). Immunofluorescence staining and confocal imaging showed morphological changes in GCM-treated macrophages, accompanied by enhanced fluorescence intensity of IL-1β compared to both the control and LPS-treated group (Fig. [106]2D). These findings collectively demonstrate that GBM cells, through the secretion of soluble factors, effectively program macrophages into an IL-1β-secreting state, contributing to the inflammatory TME. IL-1β enhances GBM cells to produce inflammatory factors Given the ability of TAMs in GBM to produce high levels of IL-1β, we investigated the functional impact of IL-1β on GBM cells. U251 cells were treated with recombinant IL-1β protein at varying concentrations. As shown in Fig. [107]3A, IL-1β-treated U251 cells exhibited significantly increased expression of inflammatory cytokines, including IL-1β, IL-6, and CSF3, compared to untreated controls. These results indicate that IL-1β enhances GBM cells’ production of inflammatory factors. Fig. 3. [108]Fig. 3 [109]Open in a new tab IL-1β promotes the production of inflammatory factors in GBM cells. (A) mRNA expression of IL6, IL1B, and CSF in U251 cells stimulated with various concentrations of IL-1β (n = 3). Statistical significance compared to the control group is indicated by *p values. (B) Workflow for RNA sequencing of U251 cells treated with 1ng/ml IL-1β. (C) Venn diagram depicting the overlap between 711 upregulated genes in IL-1β-stimulated U251 cells and 1,891 secreted protein-coding genes from The Human Protein Atlas, resulting in 117 intersecting genes. (D) Heatmap illustrating the expression of inflammation-related, cell invasion-related, and chemokine-related genes among the 117 intersecting genes. (E) List of the top 10 upregulated genes from the 117 intersecting genes. (F) Top 20 KEGG pathways enriched in IL-1β-stimulated U251 cells. (G) PTGS2 mRNA expression in U251 and T98G cells following IL-1β stimulation (n = 3). (H) ELISA analysis of PGE2 secretion in U251 and T98G cells after IL-1β stimulation (n = 3). (I) mRNA expression of PTGS2 in GBM (n = 153), LGG (n = 513), and Normal groups (n = 2642) from the TCGA database (left) and correlation analysis between PTGS2 and IL1B expression in GBM (n = 153) patients (right). (Data are presented as mean ± SEM. NS: Not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) To further elucidate the molecular mechanisms, transcriptome sequencing was performed on U251 cells treated with or without IL-1β (1 ng/mL) (Fig. [110]3B). Compared to the control group, 711 genes were upregulated in IL-1β-stimulated GBM cells. By intersecting these genes with the secreted protein dataset from the Human Protein Atlas (1891 genes), 117 overlapping genes were identified (Fig. [111]3C). Notably, the IL-1β-stimulated U251 cells exhibited increased expression of genes associated with tumor-promoting inflammation (e.g., IL1B, IL1A, IL23A, IL36G, IL24), cell migration (e.g., MMP1, MMP3, MMP7, MMP10, MMP19, MMP24), myeloid cell recruitment (e.g., CCL2–CCL5, CXCL1–CXCL2, CXCL5, CXCL8, CXCL10–CXCL11, CXCL20), and prostaglandin synthesis (PTGS2) (Fig. [112]3D-E). Pathway enrichment analysis of the upregulated genes using KEGG revealed significant enrichment in pathways related to cytokine-cytokine receptor interactions and TNF signaling (Fig. [113]3F). Among the top 10 upregulated genes, PTGS2 (cyclooxygenase-2), a critical enzyme in PGE2 synthesis [[114]33–[115]35], was highly expressed in IL-1β-stimulated cells (Fig. [116]3E). Previous studies have shown that PGE2 derived from pancreatic cancer, in cooperation with TNFα, can induce the IL-1β^+ TAM phenotype [[117]14]. Based on these findings, we hypothesize that GBM-derived PGE2 might play a central role in the generation of IL-1β^+ TAMs within the TME. To validate these findings, U251 and T98G cells were treated with IL-1β for 24 h, and PGE2 expression levels were quantified using RT-PCR and ELISA. Following IL-1β treatment, GBM cells showed a significant increase in PGE2 expression (Fig. [118]3G-H). Further analysis using the TCGA database revealed a positive correlation between PTGS2 and IL1B expression in GBM (Fig. [119]3I). Moreover, PTGS2 expression was significantly higher in GBM patients compared to lower-grade glioma (LGG) patients (Fig. [120]3I). Overall, IL-1β treatment significantly increases the production of inflammation-related factors in U251 and T98G cells, and transcriptome sequencing identifies multiple upregulated genes, including PTGS2. Further analysis through the TCGA database shows that PTGS2 expression is significantly higher in GBM patients compared to LGG patients, indicating that IL-1β and PTGS2 play a crucial role in maintaining an inflammatory TME and promoting tumor development. PGE2 and TNFα derived from GBM cells cooperatively elicit the IL-1β^+ TAM state Given the elevated levels of soluble factors in supernatants from GBM cells treated with IL-1β, we hypothesized that these supernatants (termed ^IL−1βGCM) might further induce macrophages to produce IL-1β. To test this, U251 cells were incubated in the medium containing IL-1β (1 ng/mL) for 48 h. The medium was then replaced with serum-free culture medium, and the cells were cultured for an additional 48 h to collect the supernatant devoid of exogenous IL-1β protein (Fig. [121]4A). THP1 Macs were subsequently stimulated with either U251 ^IL−1βGCM or control U251 GCM. Notably, U251 ^IL−1βGCM treated macrophages showed much higher IL1B mRNA expression compared to the cells with GCM treatment (Fig. [122]4B), suggesting that IL-1β-stimulated GBM cells secrete factors capable of further enhancing IL-1β induction in macrophages. Fig. 4. [123]Fig. 4 [124]Open in a new tab PGE2 and TNFα secreted by GBM cells collaboratively induce the IL-1β^+ TAM phenotype. (A) Experimental setup for collecting conditioned media from U251 cells treated with IL-1β and from control U251 cells. (B) IL1B mRNA expression in macrophages stimulated with U251 GCM or U251 ^IL−1βGCM for 24 h (n = 3). (C) IL1B mRNA expression in macrophages treated with various concentrations of PGE2 for 24 h (n = 3). (D) TNFA mRNA expression in GBM (n = 153), LGG (n = 513), and Normal groups (n = 2642) from the TCGA database (left) and correlation between TNFA and IL1B expression in GBM (n = 153) patients (right). (E) TNFA mRNA expression in U251 and T98G cells after 24-hour IL-1β stimulation (n = 3). (F) ELISA analysis of TNFα secretion in U251 and T98G cells treated with IL-1β for 48 h (n = 3). (G-H) TNFR1 and TNFR2 mRNA expression in macrophages after stimulation with different concentrations of PGE2 for 24 h (n = 3). (I) IL1B mRNA expression in macrophages stimulated with various concentrations of TNFα and PGE2 for 24 h (n = 3). (J) ELISA analysis of IL-1β protein levels in macrophage supernatants after stimulation with different concentrations of TNFα and PGE2 for 48 h (n = 3). (K) Volcano plot showing the upregulated and downregulated genes in THP1 Macs stimulated with U251 GCM (n = 3). (L) Volcano plot of upregulated and downregulated genes in THP1 Macs stimulated with PGE2/TNFα (n = 3). (Data are expressed as mean ± SEM. NS: Not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) To evaluate whether PGE2 derived from GBM cells play a major role in IL1B expression in macrophages, THP1 Macs were exposed to different concentrations of PGE2. Although PGE2 stimulation increased IL1B transcript levels, the maximal expression induced by 1nM PGE2 was substantially lower than that induced by U251 GCM (7-fold vs. 249-fold; Fig. [125]4B-C). Basal IL-1β secretion by GBM cells failed to enhance PGE₂-driven IL1B expression, as co-treatment with IL-1β(1ng/ml) and PGE₂ (1nM) induced only a modest (~ 6.3-fold) increase (Fig. [126]S2B-C). These results indicate that while PGE2 contributes to IL-1β induction, additional factors are required for robust IL-1β secretion. Recent findings in pancreatic cancer suggested that IL-1β-expressing TAMs could be elicited by the synergistic activity of PGE2 and tumor necrosis factor-alpha (TNFα) [[127]14, [128]36]. TNFα, a pleiotropic cytokine with dual roles in tumor biology, exerts its effects through two receptors, TNFR1 and TNFR2 [[129]37, [130]38]. Analysis of the TCGA database revealed elevated TNFα expression in GBM patients compared to LGG patients, and TNFA expression positively correlated with IL1B levels in tumors (Fig. [131]4D). GBM cells (both U251 and T98G) showed a significant increase in TNFA expression following IL-1β treatment (Fig. [132]4E). Higher levels of TNFα were also detected in the supernatants of GBM cells with IL-1β treatment compared to untreated controls. (Fig. [133]4F). Next, the expression of TNFR1 and TNFR2 were evaluated in THP1 Macs following PGE2 treatment. The results demonstrated that TNFR2, but not TNFR1, was upregulated in response to PGE2 (Fig. [134]4G-H). To examine the potential synergy between PGE2 and TNFα in inducing IL-1β production, macrophages were co-administrated with both factors. Remarkably, co-administration of PGE2 and TNFα resulted in a 20.45-fold increase in IL1B expression compared to PGE2 alone (Fig. [135]4I). Similarly, ELISA results showed significantly elevated IL-1β protein levels in macrophage supernatants treated with the combination of TNFα and PGE2 (Fig. [136]4J). RNA sequencing of macrophages identified numerous transcripts synergistically induced by PGE2 and TNFα. These transcripts were overrepresented in IL-1β^+ TAMs and encoded factors associated with tumor-promoting inflammation (IL1B, IL1A, IL23A), prostaglandin synthesis (PTGES, PTGS2), myeloid cell recruitment (CCL3, CCL24, CXCL1, CXCL2, CXCL3, CXCL5, CXCL6, CXCL8), immunosuppression (IL10, TGFB), and angiogenesis (VEGFA) (Fig. [137]4K-L). Given the role of the inflammasome in processing pro-IL-1β into its secreted form, IL-1β, we examined inflammasome-related gene expression (NLRP3, CASP1/4/5, IL18, CFLAR) in THP1 Macs. NLRP3, CASP1/4, and IL18 expression remained unchanged with either treatment, whereas CASP5 were upregulated by GCM but not by PGE₂/TNFα (Fig. [138]S3A-B). Notably, CFLAR (c-FLIP), a critical regulator required for complete NLRP3 inflammasome assembly [[139]39], was upregulated in response to both GCM and PGE₂/TNFα (Fig. [140]S3A-B), suggesting its potential role in facilitating IL-1β release under these conditions. Analysis of scRNA-seq data in myeloid cell populations from 9 GBM patients (Fig. [141]S3C-H) revealed high NLRP3, IL18, and CFLAR but low CASP1/4, indicating partial NLRP3 inflammasome activation in IL-1β^+macrophages. Taken together, these data identify PGE2 and TNFα as soluble factors secreted by GBM cells that cooperatively induce an inflammatory macrophage phenotype. C/EBPβ is upregulated in macrophages stimulated by PGE2/TNFα or GCM To explore the molecular mechanisms underlying the induction of inflammation by GCM, transcriptome sequencing was performed on macrophages stimulated with GCM or PGE2/TNFα. Pathway analysis revealed significant enrichment of inflammatory response pathways under both conditions. Among the genes enriched in the inflammatory response pathways, over 45% overlapped between the two conditions (Fig. [142]5A-B). Heatmap analysis of the 52 overlapping genes identified nine significantly upregulated gene under both conditions (Fig. [143]5C). IL1B exhibited the most prominent elevation in mRNA expression levels among the nine upregulated genes identified, as evidenced by its highest ranking in Trimmed Mean of M-values (TMM) under both PGE2/TNFα co-stimulation or U251 GCM treatment (Fig. [144]5D). This upregulation pattern suggests that PGE2 and TNFα might serve as key mediators within U251 GCM responsible for inducing robust IL1B expression in THP1 Macs. Fig. 5. [145]Fig. 5 [146]Open in a new tab C/EBPβ is upregulated in macrophages stimulated by PGE2/TNFα or GCM. (A) GO analysis identifies the top 10 biological processes with significant differences comparing PGE2/TNFα vs. control (left) and U251 GCM vs. control (right) (n = 3). (B) Enrichment of inflammatory response-related genes and their overlap between the two experimental groups compared to the control. (C-D) Expression levels of the top 9 most significantly upregulated genes among the 52 intersecting genes. (E) CEBPB mRNA expression in THP1 Macs after 24-hour stimulation with U251 GCM, T98G GCM, or PGE2/TNF-α (n = 3). *p indicates statistical significance compared to the control group. (F) mRNA expression of CEBPB in GBM (n = 153), LGG (n = 513), and Normal groups (n = 2642) from the TCGA database (left) and correlation between CEBPB and IL1B gene expression in GBM (n = 153) patients (right). (G) Expression analysis of CEBPB (left) and IL1B (right) genes in GBM patients (n = 9) from scRNA-seq data. (Data are presented as mean ± SEM. NS: Not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) CEBPB was significantly upregulated in macrophages stimulated with GCM or PGE2/TNFα (Fig. [147]5C-D). C/EBPβ is a critical transcription factor that regulates the expression of pro-inflammatory genes such as IL1B, IL6, TNF, and IL12p40 in macrophages [[148]40–[149]43]. RT-PCR validation further confirmed that CEBPB expression was upregulated in macrophages treated with either PGE2/TNF-α or GCM (Fig. [150]5E). Analysis of bulk RNA-seq data from TCGA dataset showed that CEBPB expression was positively associated with GBM and correlated with IL1B expression in GBM (Fig. [151]5F). Analysis of bulk RNA-seq data from The Cancer Genome Atlas (TCGA) dataset showed that CEBPB expression was positively associated with glioblastoma (GBM) and correlated with IL1B expression in GBM (Fig. [152]5F). These results were further supported by scRNA-seq data (Fig. [153]5G), which confirmed the link between C/EBPβ expression and the inflammatory state of macrophages within the GBM microenvironment. GBM-derived PGE2 and TNFα induce IL-1β expression in macrophages through activation of the PKA-C/EBPβ signaling pathway. Previous studies have established that C/EBPβ plays a pivotal role in the inducible expression of pro-inflammatory cytokines, including IL-1β, IL-6, TNFα, and IL-8 [[154]40, [155]41]. Specifically, PGE2 has been shown to signal via its EP4 receptor, utilizing the cAMP-PKA-C/EBPβ pathway to regulate the expression of IL-8 expression [[156]36]. However, the precise role of C/EBPβ in the transcriptional regulation of IL-1β remains to be fully elucidated. To investigate the temporal effects of PGE2 and TNFα stimulation on IL1B expression, THP1 Macs were exposed to sequential stimulations with PGE2 and TNFα, with the induction period divided into two consecutive 12-hour intervals separated by a medium change (Fig. [157]6A-B). When macrophages were stimulated with either PGE2 or TNFα alone during the first and second 12-hour intervals, both IL1B and CEBPB mRNA induction were modest (2-6-fold). Similarly, if cells were first treated with TNFα and subsequently with PGE2, the level of IL1B and CEBPB mRNA induction remained comparable to single-stimulus treatments. Remarkably, when macrophages were pre-treated with PGE2 for the initial 12 h and then exposed to TNFα, IL1B mRNA induction reached 34-fold and CEBPB mRNA induction reached 14-fold. These findings demonstrate that PGE2 effectively sensitizes macrophages to subsequent TNFα stimulation, enhancing CEBPB transcriptional activation. Western blotting analysis confirmed that co-stimulation with PGE2 and TNFα markedly enhanced C/EBPβ protein expression, while individual stimulation with either PGE2 or TNFα alone resulted in minimal C/EBPβ protein (Fig. [158]6C). Similarly, treatment of THP1 Macs with GCM derived from U251 cells robustly induced C/EBPβ protein expression (Fig. [159]6D). Further corroborating these findings, immunofluorescence analysis revealed that either PGE2/TNFα co-stimulation or GCM treatment significantly upregulated IL-1β and C/EBPβ protein expression in macrophages. Importantly, under these conditions, C/EBPβ was predominantly localized to the nucleus, indicating its activation as a transcription factor responsible for driving inflammatory gene expression (Fig. [160]6E). Fig. 6. [161]Fig. 6 [162]Open in a new tab GBM-derived PGE2 and TNFα indue IL-1β expression in macrophages via the PKA- C/EBPβ signaling pathway. (A-B) Expression of IL1B (A) and CEBPB (B) mRNA in THP1 Macs after 24 h of stimulation with different sequences of PGE2 (1nM) and TNFα (10ng/ml) (n = 3). (C) Expression of C/EBPβ protein in THP1 Macs after stimulation under various conditions (n = 3). (D) Expression of C/EBPβ protein in THP1 Macs after stimulation with PGE2/TNFα or U251 GCM. (E) Immunofluorescence showing co-expression of IL-1β (green) and C/EBPβ (red) in THP1 Macs 6 h post-stimulation under different conditions (n = 3). Blue, DAPI. (F-G) CEBPB mRNA (F) and IL1B mRNA (G) expression in THP1 Macs 24 h after stimulation with U251 or T98G GCM following a 30-minute pre-treatment with varying concentrations of PKA inhibitor (H89) (n = 3). (Data are presented as the Mean ± SEM. NS: Not significant, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) Mechanistically, PGE2 signaling operates through the elevation of intracellular cAMP levels and the subsequent activation of protein kinase A (PKA) via PGE2 receptors [[163]44, [164]45]. To delineate the role of PKA in this pathway, THP1 Macs were pre-treated with H89, a specific PKA inhibitor [[165]46, [166]47], one hour before stimulation with PGE2/TNFα or GCM. Inhibition of PKA significantly suppressed both IL1B and CEBPB mRNA expression in macrophages treated with GCM derived from U251 or T98G cells (Fig. [167]6F-G). These results highlight a critical mechanism in which PGE2 sensitizes macrophages to TNFα stimulation via the PKA-C/EBPβ signaling axis, facilitating robust IL-1β expression. Among the soluble factor-derived from tumor, PGE2 is one of key factors in educating macrophages to an IL-1β-secreting phenotype. Since PGE2 secretion is regulated by the COX-2 enzyme, we utilized NS-398, a selective COX-2 inhibitor, to suppress PGE2 production in GBM cells [[168]34, [169]48, [170]49]. As outlined in Fig. [171]7A, U251 and T98G cells were treated with varying concentrations of NS-398 for 48 h, followed by 24 h of culture in serum-free medium. The supernatants from these cultures, termed ^NS-398GCM, were collected for further use. Quantification of PGE2 in the ^NS-398GCM confirmed a dose-dependent decrease in PGE2 concentration with increasing NS-398 concentrations (Fig. [172]7B). THP1 Macs were then stimulated with ^NS-398GCM. As NS-398 concentrations increased, IL1B and CEBPB expression in macrophages was significantly reduced (Fig. [173]7C-H). These findings highlight the critical role of PGE2 in driving the differentiation of macrophages into an IL-1β-secreting phenotype and underscore the dependence of this process on COX-2-mediated PGE2 production in GBM cells. Fig. 7. [174]Fig. 7 [175]Open in a new tab PGE2 is a key factor among tumor-derived soluble molecules that drives the secretion of IL-1β in THP1 Macs. (A) Workflow for collecting ^NS398GCM from U251 and T98G cells treated with the PTGS2 inhibitor (NS398). (B) ELISA analysis of PGE2 production by U251 and T98G cells treated with NS398 (n = 3). (C-D) Effects of ^NS398GCM collected from different concentrations of NSG398 treated U251(C) or T98G (D) on the expression of IL1B mRNA in macrophages (n = 3). (E-F) Effects of ^NS398GCM collected from different concentrations of NSG398 treated U251(E) or T98G (F) on the expression of IL-1β protein in macrophages (n = 3). (G-H) Effects of ^NS398GCM collected from different concentrations of NSG398 treated U251(G) or T98G (H) on the expression of CEBPB mRNA in macrophages (n = 3). (Data are presented as the Mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) Tumor-derived PGE2 induces C/EBPβ and IL-1β expression through EP4 receptor signaling Prostaglandin E2 may exert its biological effects through four distinct receptors: EP1, EP2, EP3, and EP4. To identify the specific receptor subtype involved in the regulation of IL-1β expression in macrophages, we analyzed scRNA-seq data from GBM patients ([176]GSE182109). The analysis revealed that myeloid cells within the TME predominantly express the EP4 receptor (PTGER4), with minimal expression of EP1(PTGER1), EP2(PTGER2), and EP3 (PTGER3) (Fig. [177]8A). This observation was further validated by RT-PCR, which demonstrated that EP4 is highly expressed in THP1 Macs (Fig. [178]8B). Fig. 8. [179]Fig. 8 [180]Open in a new tab Tumor-derived PGE2 stimulates IL-1β expression via EP4 receptor signaling. (A) t-SNE plot showing the expression levels of PTGER1-4 (EP1-4) mRNA in GBM patients (n = 9) from scRNA-seq data. (B) Comparison of PTGER1-4 mRNA expression in THP1 Macs (n = 3). (C) Workflow for using ONO-AE3-208 to block the EP4 receptor in macrophages. (D-G) Effects of ONO-AE3-208 on the expression of IL1B mRNA and protein in macrophages stimulated with GCM from U251 and T98G cells (n = 3). (H-I) Effects of ONO-AE3-208 on the expression of CEBPB mRNA in macrophages stimulated with U251 and T98G GCM (n = 3). (Data are presented as the Mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) Given the prominent role of IL-1β-mediated inflammation in GBM pathogenesis, we sought to determine whether tumor-derived PGE2 promotes IL-1β-expressing macrophages via the EP4 receptor. To test this, macrophages were pre-incubated with ONO-AE3-208, a selective EP4 receptor antagonist, for 30 min prior to treatment with GCM (Fig. [181]8C). After 24 h, IL1B mRNA expression was significantly reduced in macrophages treated with GCM derived from U251 or T98G cells in a dose-dependent manner when EP4 was blocked (Fig. [182]8D-E). ELISA assays confirmed consistent results at the protein level, with decreased IL-1β secretion following EP4 inhibition (Fig. [183]8F-G). Additionally, EP4 antagonism reduced the expression of C/EBPβ, a key transcription factor driving IL-1β expression, as evidenced by mRNA analyses (Fig. [184]8H-I). Collectively, these findings demonstrate that tumor-derived PGE2 acts predominantly through the EP4 receptor to reprogram macrophages into an inflammatory IL-1β^+ phenotype. EP4 receptor Blockade reduces IL-1 β^+ TAMs and delays GBM growth Given that GBM cells induce IL-1β⁺ TAMs via PGE₂/EP4 signaling and that IL-1β enhances GBM cell proliferation (Fig.S4), we investigated whether EP4 inhibition could mitigate TAM-mediated tumor progression. As ONO-AE3-208 showed no direct cytotoxicity toward U251 cells in vitro or in vivo (Fig.S5), we employed it in a human tumor cell line-derived xenograft (CDX) model, generated by subcutaneous injection of U251 cells into nude mice with macrophage transfer, to assess the functional impact of EP4 blockade on IL-1β⁺ TAM–driven tumor progression. Once tumor sizes reached approximately 100 mm³, the mice were randomly divided into three groups: Group G (control), Group GM (Mac treatment), and Group GMO (Mac + ONO-AE3-208). Starting on day 15, groups GM and GMO received intravenous injections of 1 × 10^6 THP1 Macs every three days. Additionally, Group GMO was intraperitoneally injected ONO-AE3-208 (10 mg/kg) to block EP4 receptor signaling, starting on day 14 (Fig. [185]9A). Fig. 9. [186]Fig. 9 [187]Open in a new tab Inhibition of the EP4 receptor reduces IL-1β^+ TAMs and impedes tumor growth in vivo. (A) Schematic of the subcutaneous GBM xenograft mouse model and drug injection protocol. (B) Tumor growth after injection of ONO-AE3-208 and macrophages (n = 5). (C) Photographs of tumors excised from mice. (D) Tumor weight (n = 5). (E-F) PGE2 concentrations in blood and tumor tissues from different groups of mice (n = 3). (G-H) IL-1β concentrations in blood and tumor tissues from different groups of mice (n = 5). (I) Immunofluorescence staining of IL-1β and CD68 in tumor tissues from mice. Left: Representative images of each group (Blue: DAPI; Green: IL-1β; Red: CD68). Right: Quantification of IL-1β mean fluorescence intensity across groups. (J) Immunofluorescence staining of Ki67 in tumor tissues from different groups of mice. Left: Representative fluorescence images showing Ki67 expression (Blue: DAPI; Red: Ki67). Right: Quantified expression levels of Ki67 across groups (Data are presented as the Mean ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001) Tumors in Group GM were significantly larger than those in Group G (control), indicating that TAMs promote tumor growth (Fig. [188]9B). In contrast, tumor growth was significantly reduced in Group GMO compared to Group GM, demonstrating that EP4 receptor inhibition effectively suppresses tumor progression (Fig. [189]9C-D). To investigate the underlying mechanisms, PGE2 and IL-1β levels were measured in serum and tumor tissues. While serum PGE2 levels were similar across all groups, tumor PGE2 levels were markedly higher in Groups GM and GMO (Figs. [190]9E-H), suggesting that TAMs promote PGE2 production within the TME. As was expected, IL-1β was undetectable in the serum or tumor tissues of Group G (control). However, IL-1β levels were significantly elevated in Group GM compared to Groups G and GMO, highlighting the role of TAMs in IL-1β production and the suppressive effect of EP4 receptor blockade. Immunofluorescence assays further confirmed a marked reduction in IL-1β^+ TAM infiltration in Group GMO following EP4 inhibition (Fig. [191]9I). Additionally, Ki67, a marker of tumor cell proliferation, was significantly upregulated in Group GM, supporting the role of TAMs in promoting GBM growth (Fig. [192]9J). In contrast, EP4 receptor inhibition in Group GMO significantly reduced Ki67 expression, correlating with reduced tumor cell proliferation. Our findings demonstrate that GBM cells remodel macrophages into IL-1β^+ TAMs via PGE2/EP4 signaling, and that IL-1β^+ TAMs contribute to GBM progression in the mouse CDX model. Importantly, EP4 receptor inhibition alleviates these pro-tumorigenic effects, highlighting its potential as a therapeutic target. Discussion TAMs have been demonstrated to be crucial for promoting GBM progression [[193]50, [194]51]. In this study, we demonstrate that the pro-inflammatory cytokine IL-1β is induced and released by macrophages in response to tumor-derived PGE2 and TNFα. IL-1β, in turn, drives tumor cells to secrete more inflammatory mediators, particularly PGE2, thus creating a positive feedback loop. This interaction between macrophages and tumor cells perpetuates the inflammatory microenvironment in GBM and fosters tumor progression. Importantly, in vivo administration of an EP4 inhibitor significantly delayed tumor growth in tumor cell-derived xenograft models, highlighting the therapeutic potential of targeting the PGE2–EP4–IL-1β axis, which emerges as a key driver of IL-1β^+ TAMs in GBM. Our findings provide insights into mechanistic understanding of how IL-1β mediates pro-tumor effects and how tumor-derived PGE2 and TNFα regulate TAMs, which have important implications for GBM therapy. Macrophages are the predominant immune cells within GBM, significantly contributing to tumor progression through the secretion of soluble factors, including various growth factors and cytokines. Thus, TAMs represent a promising therapeutic target in cancer treatment. However, the role of TAMs in tumor progression and therapeutic response is paradoxical, largely due to their pronounced compositional and functional heterogeneity [[195]52, [196]53]. This complexity underscores the necessity for a more nuanced understanding of TAM subsets and their regulation to develop effective treatment strategies that harness their full therapeutic potential. Here we identified IL-1β^+ macrophages in TME and demonstrated their strong association with GBM progression and poor patient outcomes. Notably, IL-1β^+ TAMs have been previously implicated in the progression of other cancers, such as pancreatic and breast cancers [[197]54–[198]56]. Through the production of IL-1β, TAMs contribute to establishing a pro-tumor inflammatory microenvironment and promote tumor growth via multiple mechanisms. For example, IL-1β induces the expression of CCL2 in both TAMs and tumor cells, facilitating the recruitment of myeloid cells into the tumor tissue [[199]19]. Moreover, Chen et al. reported that IL-1β played a pivotal role in driving the expansion and migration of myeloid-derived suppressor cells (MDSCs) in GBM [[200]57]. Our observation revealed that macrophages differentiate into IL-1β^+ TAMs upon exposure to GBM-derived PGE2 and TNFα. IL-1β derived from macrophages enhanced the capacity of GBM cells to produce inflammatory mediators and chemokines, including IL-6, PGE2, TNF-α, and CSF3. This bidirectional interaction established a positive feedback loop that reinforced the IL-1β^+ TAM phenotype and sustained a pro-inflammatory TME, thereby contributing to tumor progression. PGE2 is considered as a bifunctional mediator within TME, where inflammation often coexists with immunosuppression [[201]58]. On one hand, PGE2–EP4 signaling is a key driver of immune suppression in the TME. PGE2 has been shown to impair the Function of conventional dendritic cells type 1 (cDC1) [[202]59] and steers cDC2 maturation [[203]60], thereby hindering effective T cell activation in tumors. Additionally, PGE2 acts synergistically with LPS to enhance IL-10 production in macrophages [[204]61]. On the other hand, PGE2 can activate NF-κB signaling, thus promoting inflammatory responses [[205]62]. Our findings indicated that GBM-derived PGE2 and TNF-α synergistically induced robust IL-1β synthesis in THP1 Macs, aligning with previous studies [[206]14, [207]20]. Notably, PGE2 alone displayed limited capacity to drive IL-1β^+macrophage programming, highlighting its dependence on complementary signals such as TNFα. We observed that sequential stimulation of THP1 Macs with PGE2 (administered during the initial 12 h) followed by TNF-α (administered during the subsequent 12 h) resulted in significantly higher IL-1β mRNA expression. However, the level of IL-1β mRNA expression was very low when cells were first treated with TNFα and subsequently with PGE2. These findings suggest that it is a critical mechanism by which GCM facilitates robust IL-1β production to use PGE2 to sensitize macrophages to TNFα stimulation. In addition, we found that the inhibition of PGE2 production in GBM cells through COX-2 blockade directly attenuated IL-1β production by macrophages. Altogether, these findings highlight that tumor-derived PGE2 plays a crucial role in the interaction between IL-1β^+ macrophage and GBM cells, offering potential therapeutic opportunities to disrupt these proinflammatory pathways in GBM. PGE2 has been identified as an important regulator of TAM function via four known receptors (EP1-EP4) [[208]63]. Here, we found the predominance of the EP4 receptor in THP1 Macs and TAMs in the GBM microenvironment, with minimal expression of the other 3 receptors. We also demonstrated the downregulation of GCM-induced IL-1β mRNA and protein levels upon treatment with an EP4-specific antagonist. Additionally, our finding confirmed the suppression of IL-1β production and tumor growth in a cell-derived xenograft (CDX) model following EP4 signaling ablation. Our finding indicates the therapeutic potential of targeting the PGE2–EP4–IL-1β axis to disrupt these proinflammatory pathways in GBM. PGE2 exerts its effects through EP4 via the cAMP/PKA/CREB signaling pathway. The transcription factor C/EBPβ, a downstream effector of the PGE2/PKA signaling pathway, has emerged as a critical regulator of proinflammatory gene expression, including IL8 [[209]36]. Stimulation of EP2 and EP4 receptors by PGE2 in bone marrow-derived macrophages has been shown to rapidly induce CREB phosphorylation, leading to subsequent C/EBPβ expression. Our findings reveal that GBM-derived PGE2 and TNFα cooperatively drive C/EBPβ-mediated inflammatory responses in macrophages. We hypothesize that PKA-dependent induction of C/EBPβ facilitates IL-1β synthesis, supported by observations that (a)PGE2/TNF-α and GCM upregulated C/EBPβ expression, and (b) downregulated CEBPB and IL1B expression in macrophages treated with PKA inhibitor. Similar signaling dynamics have been observed in PGE2/TNF-α-induced IL-8 expression [[210]36]. Additionally, PGE2 stimulation of EP2 and EP4 in bone marrow-derived macrophages led to rapid CREB phosphorylation and subsequent C/EBPβ expression [[211]44]. Thus, PGE2 is a key factor in educating macrophages to adopt an IL-1β-secreting phenotype in response to soluble factors derived from GBM cells via the PKA-C/EBPβ signaling axis. While this study provides important insights, several limitations need to be addressed in future research. First, although we verified the presence and function of IL-1β^+ TAMs through single-cell sequencing and in vitro experiments, validation in larger, more diverse patient cohorts is essential to confirm these findings and establish their clinical relevance. Second, the precise roles of IL-1β and PGE2 in GBM progression require further investigation, particularly through in vivo models that better replicate the complexity of the TME. Future studies should also explore additional signaling pathways and molecular mediators involved in the IL-1β-driven interactions between GBM cells and TAMs to provide a more comprehensive understanding of this process. In conclusion, our findings reveal a link between increased IL-1β-expressing TAMs and poor survival in GBM patients. We demonstrated that GBM cells reprogram macrophages to produce IL-1β through PGE2 and TNFα, with PGE2 signaling via the EP4 receptor to upregulate the transcription factor C/EBPβ (Fig. [212]10). This signaling cascade enhances IL-1β production in macrophages, establishing a feedback loop that perpetuates tumor inflammation and progression. Importantly, pharmacological inhibition of EP4 in preclinical models effectively reduced IL-1β expression and slowed tumor growth, highlighting the EP4 receptor as a promising therapeutic target to disrupt this inflammatory feedback loop and mitigate GBM progression. Fig. 10. [213]Fig. 10 [214]Open in a new tab A proposed model for this study. Schematic diagram depicting the positive feedback loop (IL-1β-PGE2/TNFα-C/EBPβ-IL-1β) between GBM cells and IL-1β^+ macrophages Supplementary Information [215]Supplementary Material 1.^ (292KB, docx) [216]Supplementary Material 2.^ (6.6MB, xls) [217]Supplementary Material 3.^ (1.3MB, docx) Acknowledgements