Integrated single-cell and transcriptome sequencing data reveal the value of AOC1 in gliomas tumor microenvironment and prognosis Authors: Jun Liu, Jiayu Chen, Liang Chen, Yang Wang, Jinxin Xin, Lei Mou, Zhiming Chen Published: 2025 ABSTRACT BackgroundGlioma is a primary brain tumor that is frequently encountered in humans, with malignant gliomas representing an exceptionally aggressive subtype. The role of AOC1 in the progression of various human tumors has been documented; however, its involvement in gliomas (including distinct WHO grades, namely WHO Grade IV glioblastoma multiforme [GBM] and WHO Grades II–III low-grade gliomas [LGG]), remains largely unexplored.MethodsWe analyzed RNA sequencing data from The Cancer Genome Atlas (TCGA) and Genotype-Tissue Expression (GTEx) databases to assess AOC1 expression. We performed transcriptomic analysis and differential gene expression studies based on AOC1 levels in TCGA sequencing data. Functional analyses exploring the signaling pathways involved in AOC1. CIBERSORT, single-sample Gene Set Enrichment Analysis (ssGSEA), and ESTIMATE were employed to estimate the tumor microenvironment in gliomas. Analyzing the single-cell sequencing data further revealed the value of AOC1 played in intercellular communication. Finally, the mechanism of AOC1 involvement in gliomas was verified by cellular and molecular experiments.ResultsHigh tumor expression of AOC1 is associated with lower patient survival. Our analyses revealed that cytokine signaling pathways are significantly suppressed in AOC1-high GBMLGG tissues. Q-PCR results corroborated that AOC1 inhibits IL17, IL6 and IL1β secretion in GBM cells. The CIBERSORT analysis showed a significant increase in the proportion of M2 macrophages in AOC1-high cancer tissues. Cell-cell communication analysis indicated that the MIF signaling network between AOC1-high GBMLGG tumor cells and macrophages is notably weakened. Additionally, AOC1 silencing directly delays the proliferation and migration of GBMLGG cells in vitro.ConclusionsTumor AOC1 levels can serve as a molecular marker for diagnosis and molecular subtyping. As a secreted protein, AOC1 holds promise as a new target for non-invasive molecular diagnosis and precision therapy in GBMLGG.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12885-025-15457-2. Introduction Glioma is central nervous system tumor that originate from glial cells, which serve as non-neural support cells in the brain [1]. As a prevalent primary brain tumor, glioma is highly aggressive, exhibiting an average incidence of 6 per 100,000 individuals annually [2–4]. Malignant gliomas are particularly devastating due to their high malignancy, extremely elevated mortality rates and significant risk of recurrence, imposing a considerable burden on both society and families [5]. Malignant gliomas include the WHO Grade IV glioblastoma multiforme (GBM) and WHO Grades II–III low-grade gliomas (LGG). Meanwhile, the incidence of gliomas has been steadily increasing, with a study from Japan indicating a significant rise in malignant glioma rates among the elderly from 1989 to 2008 [6]. Despite extensive research on therapeutic approaches to glioma over the past few decades, our understanding of the disease remains limited [4]. Previous studies have highlighted that the onset and progression of glioma are closely linked to its genetic and molecular characteristics [3, 4]. Furthermore, latest research has revealed that the development and advancement of gliomas are closely related with the alterations in the tumor microenvironment and mechanisms of immune evasion [2]. However, the underlying mechanism of gliomas pathogenesis and progression remains largely unknown. Consequently, elucidating the molecular mechanisms underlying these genetic variations and abnormal expressions is crucial for the development of novel targeted therapies and therapeutic strategies for treating glioma. The amine oxidase copper-containing (AOC) family comprises copper-dependent amine oxidases, primarily including AOC1, AOC2, and AOC3 [7]. AOC1, in particular, is a secreted diamine oxidase capable of degrading polyamines such as putrescine and histamine [7]. Previous studies have indicated that AOC plays a significant role in regulating various biological processes, including cellular proliferation, differentiation, apoptosis, and oxidative stress [8–11]. Elevated AOC1 expression was observed across several malignancies. Notably, the increased tumoral expression level of AOC1 strongly associates with tumor differentiation status, directly influencing colorectal cancer patient prognosis [11]. Similarly, AOC1 activates the AKT signaling pathway and EMT process to exacerbate gastric cancer [12]. While AOC1 has been demonstrated to promote tumorigenesis in colorectal and gastric cancers, its role in glioma remains poorly understood [8, 9, 11–13]. Thus, further investigation into the AOC1 gene may elucidate the underlying mechanisms of glioma and potentially identify new therapeutic targets for its treatment. In this study, we first performed transcriptomic analysis using the TCGA database and discovered that AOC1 is highly expressed in cancer tissues of GBM and LGG patients, with its levels progressively increasing with cancer grade. High tumor expression of AOC1 is associated with lower patient survival. Differential gene expression and enrichment analyses based on AOC1 levels in TCGA bulk sequencing data revealed that cytokine signaling pathways are significantly suppressed in AOC1-high GBMLGG tissues. In vitro, Q-PCR results further corroborated that AOC1 inhibits cytokine secretion in GBM cells. Using the CIBERSORT algorithm to analyze bulk sequencing data, we found a significant increase in the proportion of M2 macrophages in AOC1-high cancer tissues. Subsequent cell-cell communication analysis of single-cell data indicated that the MIF signaling network between AOC1-high GBMLGG tumor cells and macrophages is notably weakened. Concurrently, AOC1 directly promotes the proliferation and migration of GBMLGG cells. Finally, we identified that tumor AOC1 levels can serve as a molecular marker for diagnosis and molecular subtyping. As a secreted protein, our findings suggest that AOC1 holds promise as a new target for non-invasive molecular diagnosis and precision therapy in GBMLGG. Methods RNA sequencing data processing and filtering The comprehensive RNA sequencing (RNA-seq) data encompassing various cancer types were systematically processed and acquired from the UCSC XENA database [14, 15]. For our analyses, we specifically retrieved HTSeq-FPKM and HTSeq-Count data for GBM and LGG samples directly from the TCGA portal (https://portal.gdc.cancer.gov/). Additionally, transcriptome data for GBM and LGG patients were sourced from GSE4290, while single-cell transcriptome data from four glioma patients were obtained from GSE270109. It is important to note that this study was conducted in strict accordance with the guidelines established by TCGA, GEO, and GTEx. Differentially expressed gene (DEG) analysis To identify differentially expressed genes (DEGs), we employed the DESeq2 R package to analyze DEGs according to the expression levels of AOC1 in GBM and LGG samples. A cutoff threshold of 50% was established to categorize samples into high and low expression groups, as previously reported [16]. Functional enrichment analysis We conducted functional enrichment analyses of the differentially expressed genes via the ClusterProfiler R package (version 3.14.3). DEGs were selected based on the criteria of |logFC| > 1 and adjusted p-value (padj) < 0.05 [17]. The analysis included an examination of Gene Ontology (GO) categories, focusing on Cellular Component (CC), Molecular Function (MF), and Biological Process (BP). Furthermore, we performed pathway analysis utilizing the Kyoto Encyclopedia of Genes and Genomes (KEGG) to elucidate the biological pathways associated with the identified DEGs. Gene set enrichment analysis We performed Gene Set Enrichment Analysis (GSEA) to investigate the functional and pathway differences, comparing the GBM and GBM and LGG samples with different AOC1 levels [17]. For each analysis, the genomes were permuted 1,000 times. Statistical significance was assessed based on an adjusted p-value (padj) of less than 0.05 and a false discovery rate (FDR) q-value of less than 0.25. Analysis of immune infiltration in bulk sequencing data Immune cell infiltration abundance, functional scoring of immune cells, and tumor purity are widely utilized methodologies for conducting immune-related analyses. In this study, we employed CIBERSORT, single-sample Gene Set Enrichment Analysis (ssGSEA), and ESTIMATE scoring approaches to investigate the relationship between AOC1 expression and the tumor microenvironment. To compare the enrichment scores of high and low AOC1 expression groups, we utilized the Wilcoxon rank-sum test. Predictive model generation and prediction We utilized the RMS R package to generate personalized predictions of overall survival (OS) for patients with GBMLGG [18]. To assess the calibration of our model, we constructed a calibration curve by plotting the predicted probabilities from the nomogram against the observed outcomes. An ideal prediction would align with the 45° line on the graph, indicating optimal predictive accuracy. To evaluate the effectiveness of the nomogram, we calculated the concordance index (C-index) using the bootstrap method with 1,000 resamples. Additionally, we conducted Decision Curve Analysis (DCA) and constructed receiver operating characteristic (ROC) curves to compare the predictive accuracy of the nomograms against individual prognostic factors comprehensively. A significance threshold of P < 0.05 was established for all statistical analyses. The DeLong test was used to statistically compare the area under the curve (AUC) between models (two-tailed, P < 0.05). Single-cell sequencing data processing The glioma single-cell sequencing data was acquired from GSE270109, consisting of 1 normal tissue and 3 glioma tissues. R software and relevant function packages were utilized to process the data, extracting information and generating Seurat objects. Quality control measures included monitoring mitochondrial gene expression above 6%, excluding two-cell samples from analysis, omitting erythrocyte gene expression surpassing 5% in the dataset, and considering and mitigating the impact of cell cycle effects on results during analysis. The integration of data for further analysis involved utilizing the “harmony” package, a specialized software tool designed for this purpose. To effectively visualize clustered cells in a 2D plot, we employed the Unified Metamorphic Approximation and Project (UMAP) technique known for its ability to reduce dimensionality while preserving important features. Cell clusters were annotated using “SingleR” along with references from previous literature reports [19–21]. Prediction and analysis of intercellular communication Compared to traditional organizations, the communication between cells within the tumor microenvironment undergoes modifications. CellPhoneDB functions as a publicly accessible database that encompasses tissue receptors, ligands, and their intricate interactions. CellChat aids in predicting pathways of communication between cells by analyzing complexes formed by proteins, ligands, and receptors. Cells are categorized into distinct subgroups based on differences in levels of AOC1 expression. The dissimilarities in communication patterns between cell subgroups with high/low AOC1 expression and other cellular phenotypes are evaluated using correlation heatmaps. Cell line The tumor cell lines used in this study were U251MG and U87MG, obtained from China Center for Type Culture Collection (CCTCC). Cells were cultured in DMEM medium containing 10% fetal bovine serum and 1% penicillin-streptomycin. Cells were cultured in an incubator at 37 °C in 5% CO2. RT-qPCR We employed the LightCycler 96 System (Roche, Switzerland) along with the PerfectStart Green qPCR SuperMix (TransGen, China) for quantitative real-time RT-PCR (qRT-PCR) analysis to measure the mRNA levels accurately. The primers used in the study were: AOC1-F: 5'-TGTCCACGCAACCTTCTACA- 3'AOC1-R: 5'-ACTGGGTCTGCTCAAGTGTG- 3'β-Actin-F: 5'-CCCGAGCCGTGTTTCCT- 3'β-Actin-R: 5'-GTCCCAGTTGGTGACGATGC- 3' AOC1-F: 5'-TGTCCACGCAACCTTCTACA- 3' AOC1-F: 5'-TGTCCACGCAACCTTCTACA- 3' AOC1-R: 5'-ACTGGGTCTGCTCAAGTGTG- 3' AOC1-R: 5'-ACTGGGTCTGCTCAAGTGTG- 3' β-Actin-F: 5'-CCCGAGCCGTGTTTCCT- 3' β-Actin-F: 5'-CCCGAGCCGTGTTTCCT- 3' β-Actin-R: 5'-GTCCCAGTTGGTGACGATGC- 3' β-Actin-R: 5'-GTCCCAGTTGGTGACGATGC- 3' CCK8, wound healing and transwell assay Cells (U87MG) were transfected with either control siRNA, siRNA targeting AOC1 (siRNA-AOC1), and a non-coding siRNA (siRNA-NC) using Lipofectamine 3000 reagent according to the manufacturer’s instructions. After 24 h post-transfection, a sterile 200 µL pipette tip was used to create a uniform scratch (wound) across the cell monolayer in each well. The cells were then washed gently with phosphate-buffered saline (PBS) to remove detached cells and debris. Fresh serum-free medium was added to the wells to prevent cell proliferation and to ensure that observed changes were due to migration alone. The initial scratch (0 hour) was photographed using an inverted microscope equipped with a digital camera. Images were taken at multiple points along the scratch to ensure consistency. The cells were then incubated at 37 °C in a 5% CO2 humidified atmosphere. After 24 hours, the same areas of the scratch were photographed again to monitor cell migration into the wound area. The images were analyzed using ImageJ software to measure the distance migrated by the cells. The wound closure was quantified by comparing the width of the wound at 0 hours and 24 hours. The percentage of wound closure was calculated for each group (control, siRNA-AOC1, and siRNA-NC). Transwell inserts were placed into a 24-well plate. The lower chamber (the well) was filled with 600 µL of complete medium containing 10% FBS as a chemoattractant. After 24 h post-transfection, 200 µL of serum-free medium containing 5 × 104 cells was added to the upper chamber of each Transwell insert. The cells were allowed to migrate through the membrane towards the lower chamber. The plate was incubated at 37 °C in a 5% CO2 humidified atmosphere for 24 h to allow cell migration. After the incubation period, non-migratory cells remaining on the upper side of the membrane were gently removed with a cotton swab. The cells that had migrated to the lower side of the membrane were fixed with 4% paraformaldehyde for 15 min and then stained with 0.1% crystal violet solution for 30 min. The membranes were washed with PBS to remove excess stain. The number of migrated cells was counted under an inverted microscope at multiple fields of view per membrane. Representative images were captured for documentation. Statistical analysis The R software (version 3.6.2) was utilized for statistical analyses and graphing. Wilcoxon rank sum test was conducted to assess AOC1 expression in unpaired samples. We employed both Cox regression analysis and the Kaplan-Meier method to evaluate prognostic factors in our investigation. To compare the effect of AOC1 expression on survival with other clinical characteristics, we conducted multivariate Cox analysis. We determined the cutoff value for AOC1 expression based on its median level, and statistical significance was defined as p-values below 0.05.To evaluate the diagnostic efficacy of AOC1 expression in distinguishing between GBMLGG and healthy samples, ROC analysis was performed using the pROC package [18]. The AUC values, ranging from 0.5 to 1.0, indicated a recognition power of 50–100%. Results AOC1 expression is aberrantly upregulated in GBMLGG The RNA-seq data for TCGA and GTEx were downloaded from UCSC XENA (https://xenabrowser.net/datapages/) in a uniform format processed by the toil process. We compared the expression levels of AOC1 between normal samples from the TCGA and GTEx databases and the corresponding tumor samples from the TCGA database. Results showed significant upregulation of AOC1 in 25 cancer types, including GBM and LGG (Fig. 1A). Further analysis of the GBM and LGG datasets revealed that AOC1 expression was markedly increased in the GBM and LGG (termed as GBMLGG) samples (Fig. 1B). In order to explore the association between AOC1 expression and clinical features in GBMLGG patients, we conducted an analysis of AOC1 mRNA expression levels across different clinical categories using data from the TCGA database. The findings indicate a significant association between elevated AOC1 expression and WHO grade, 1p/19q codeletion, and age over 60 years (Fig. 1C-E). Simultaneously, we conducted an analysis of AOC1 expression in the human normal glial cell line HEB and three glioma cell lines HS683, U251, and U87 (Fig. 1F). Fig. 1AOC1 aberrantly overexpresses in pan-cancer and GBMLGG. A Expression levels in AOC1 pan-cancer samples versus paired normal samples. B AOC1 expression levels in normal samples and TCGA-GBMLGG samples. C-E The expression level of AOC1 in each subgroup of GBMLGG. F the expression of AOC1 in human normal glial cell line HEB and three glioma cell lines HS683, U251 and U87. Between-group analysis: Wilcoxon rank sum test; NS: P ≥ 0.05; *P < 0.05; **P < 0.01; ***P < 0.001 AOC1 aberrantly overexpresses in pan-cancer and GBMLGG. A Expression levels in AOC1 pan-cancer samples versus paired normal samples. B AOC1 expression levels in normal samples and TCGA-GBMLGG samples. C-E The expression level of AOC1 in each subgroup of GBMLGG. F the expression of AOC1 in human normal glial cell line HEB and three glioma cell lines HS683, U251 and U87. Between-group analysis: Wilcoxon rank sum test; NS: P ≥ 0.05; *P < 0.05; **P < 0.01; ***P < 0.001 High expression of AOC1 predicts poor survival in GBMLGG patients We conducted ROC curve analysis to evaluate the potential of AOC1 in differentiating GBMLGG patients from healthy individuals. The calculated AUC of 0.801 suggests that AOC1 holds promise as a biomarker in this context (Fig. 2A). We employed Kaplan-Meier analysis to investigate the association between AOC1 expression and prognosis in GBMLGG patients. As depicted in the Figure, patients with high AOC1 expression exhibited a poorer prognosis compared to those with low AOC1 expression, as evident by the overall survival (OS) outcomes. (hazard ratio [HR], 2.73 (2.12–3.52); P < 0.001) (Fig. 2B), disease specific survival (DSS) (hazard ratio [HR], 2.72 (2.08–3.55); P < 0.001) (Fig. 2C), progress free interval (PFI) (hazard ratio [HR]], 1.88 (1.52–2.33); P < 0.001) (Fig. 2D), and we analyzed the prognosis in each subgroup of GBMLGG, Kaplan-Meier analysis showed that AOC1 was highly expressed in the Female subgroup (HR = 2.12(1.44–3.13), P < 0.001), age ≤ 60 years old subgroup (HR = 2.47(1.80–3.39), P < 0.001), WGO grade: G3 subgroup (HR = 2.88(1.79–4.62), P < 0.001), 1p/19q non-condel subgroup (HR = 2.61 (2.00–3.41.00.41), P < 0.001), Astrocytoma subgroup (HR = 1.91 (1.03–3.54), P = 0.041) were associated with poor prognosis (Fig. 2E-I). We utilized univariate Cox proportional hazards regression to evaluate the factors influencing overall survival (OS), and it was shown that AOC1, WHO grade, IDH status, 1p/19q codeletion, age, and histological type were predictors of poor OS. These factors were then added to multivariate cox regression, showing that AOC1 and IDH status were independent prognostic factors for worse OS (P < 0.001). Collectively, our analysis provides preliminary evidence that the aberrantly increased expression of AOC1 serves as a significant marker for poor prognosis in patients with GBMLGG. Fig. 2High expression of AOC1 was associated with poor OS in GBMLGG patients. A ROC analysis of the diagnostic effect of AOC1 on GBMLGG. B-D Kaplan-Meier curves of OS, DSS and PFI in GBMLGG. E-I Kaplan-Meier curves of GBMLGG patients in the subgroup High expression of AOC1 was associated with poor OS in GBMLGG patients. A ROC analysis of the diagnostic effect of AOC1 on GBMLGG. B-D Kaplan-Meier curves of OS, DSS and PFI in GBMLGG. E-I Kaplan-Meier curves of GBMLGG patients in the subgroup High AOC1 expression leads to the suppressed cytokines release in GBMLGG To elucidate how elevated AOC1 expression contributes to poor prognosis in GBMLGG patients, we identified DEGs associated with low and high AOC1 expression in GBMLGG samples and performed functional enrichment analysis, as Fig. 3A illustrated. We analyzed the differences in median mRNA expression between the high-AOC1 expression group and the low-AOC expression group in gene expression profiles. Comparing the high-AOC1 group and the low-AOC group, we identified a total of 1279 differentially expressed genes (DEGs), which were deemed statistically significant (|logFC| >1 and padj < 0.05) (Fig. 3A), with the majority (872) of the genes showing increased expression in the low-AOC group. Fig. 3Functional enrichment analysis reveals the signaling pathways and functions in which AOC1 may be involved. A Patients were divided into two groups, high expression and low expression, based on the expression level of AOC1 for differential analysis and enrichment analysis. B, C Select differentially expressed genes for GO and KEGG enrichment analysis, and present the results using a bubble chart. D Display of GSEA enrichment analysis results Functional enrichment analysis reveals the signaling pathways and functions in which AOC1 may be involved. A Patients were divided into two groups, high expression and low expression, based on the expression level of AOC1 for differential analysis and enrichment analysis. B, C Select differentially expressed genes for GO and KEGG enrichment analysis, and present the results using a bubble chart. D Display of GSEA enrichment analysis results We utilized the clusterProfiler and GSEA package to perform feature enrichment analysis on the DEGs mentioned above, focusing on GO and KEGG. The enriched biological processes (BP) were cell chemotaxis, leukocyte migration, myeloid leukocyte migration, response to chemokine, and chemokine-mediated signaling pathway (Fig. 3B). The enriched cellular components (CC) encompassed various structures, such as the external side of plasma membrane, the immunoglobulin complex, T cell receptor complex, collagen trimer, and so on (Fig. 3B). Additionally, the enriched molecular functions (MF) included antigen biding, cytokine activity, immunoglobulin receptor binding, among others. In the KEGG analysis, the enriched pathways comprised diverse pathways, such as cytokine-cytokine receptor interaction, viral protein interaction with cytokine and cytokine receptor, IL-17 signaling pathway, among others. (Fig. 3C) (Supplementary Table e1). To better understand the biological pathways of GBMLGG at different expression levels of AOC1, we used GSEA to analyze the profiles. AOC1 was inversely correlated with the immune activities according to GSEA analysis (Fig. 3D) (Supplementary Table 2). Given that recent studies indicate that inflammatory signaling pathways influence patient prognosis by affecting both the tumor microenvironment and the intrinsic characteristics of tumor proliferation and migration [22], we hypothesize that the alterations in the top-ranked inflammatory signaling pathways may be a key mechanism through which AOC1 impacts patient outcomes. Given that the cytokine signaling pathways regulated by AOC1 plays pivotal roles in tumor growth and shaping an immunosuppressive tumor microenvironment, we further investigate the impacts of AOC1 on pro-inflammatory pathways. We knocked down AOC1 by siRNA in U87MG and U251MG, and Fig. 4A and B shows the siRNA knockdown efficiency. Western blot showed that knocking down AOC1 leads to the decrease in IL-17 levels (Fig. 4C, D). The results of RT-qPCR also showed that knockdown of AOC1 decreased the expression of IL6 and IL17 (Fig. 4E). Meanwhile, AOC1 overexpression significantly increased IL-17 expression in both U87MG and U251MG cell lines (Fig. 4F, G). Collectively, our results that AOC1 shows promotive effects on pro-inflammatory cytokines (IL-17, IL6) synthesis in vitro. Fig. 4AOC1 promotes the pro-inflammatory cytokines synthesis. A Validation of the efficiency of siRNA knockdown of AOC1 in U87MG. B Validation of the efficiency of siRNA knockdown of AOC1 in U251MG. C-D Knockdown of AOC1 was followed by a decrease in IL17 expression in U87MG (C) and U251MG (D). E-F Overexpression of AOC1 was followed by a increase in IL17 expression in U87MG (E) and U251MG (F). G RT-qPCR verified that mRNA expression of IL17 was reduced after knockdown of AOC1. (Original protein blot in Supplementary Material) AOC1 promotes the pro-inflammatory cytokines synthesis. A Validation of the efficiency of siRNA knockdown of AOC1 in U87MG. B Validation of the efficiency of siRNA knockdown of AOC1 in U251MG. C-D Knockdown of AOC1 was followed by a decrease in IL17 expression in U87MG (C) and U251MG (D). E-F Overexpression of AOC1 was followed by a increase in IL17 expression in U87MG (E) and U251MG (F). G RT-qPCR verified that mRNA expression of IL17 was reduced after knockdown of AOC1. (Original protein blot in Supplementary Material) AOC1 level positively correlates with the infiltration of M2-polarized TAMs in GBMLGG The CIBERSORT algorithm was used to assess the extent of immune cell infiltration in each sample. We generated a heatmap illustrating the differential genes and proportions of immune cell infiltration, and observed a strong correlation between AOC1 and M2 macrophages (Fig. 5A-B). Among all immune cells, M2 macrophages exhibited the highest proportion within the samples, with significantly higher infiltration proportions observed in the high expression group of AOC1 compared to the low expression group (Fig. 5C-D). The functionality of immune cells was further assessed, revealing a significant enhancement in the function of various inhibitory immune cells within the high-expression group of AOC1. The results of the ESTIMATE algorithm indicate an elevated expression of various components in the tumor microenvironment in the high AOC1 expression group, while the low AOC1 expression group exhibits the highest tumor purity (Fig. 5E-H). This finding shows an increase in immune suppressive cell components within the high AOC1 expression group, implying compromised immune function within the tumor microenvironment. Fig. 5The AOC1 affects the functionality of immune cells in the tumor microenvironment of glioblastoma. A, B The CIBERSORT evaluates the abundance of immune cell infiltration in tumor samples and demonstrates the correlation between differentially expressed genes and immune cell infiltration abundance. C The differences in the proportions of immune cell infiltration and cellular functions between the high expression group and low expression group of AOC1 were compared based on calculations using CIBERSORT and ssGSEA. E-H Based on the results of the ESTIMATE algorithm, we compared the differences in tumor sample purity between the high expression group and low expression group of AOC1 The AOC1 affects the functionality of immune cells in the tumor microenvironment of glioblastoma. A, B The CIBERSORT evaluates the abundance of immune cell infiltration in tumor samples and demonstrates the correlation between differentially expressed genes and immune cell infiltration abundance. C The differences in the proportions of immune cell infiltration and cellular functions between the high expression group and low expression group of AOC1 were compared based on calculations using CIBERSORT and ssGSEA. E-H Based on the results of the ESTIMATE algorithm, we compared the differences in tumor sample purity between the high expression group and low expression group of AOC1 Elevated tumoral AOC1 expression suppresses MIF signaling network in tumor-macrophage intercellular communication The GSE270109 dataset includes single-cell sequencing results of both normal tissue and glioma tissue. After conducting quality control, we utilized the ‘harmony’ package for data integration. Based on reliable literature, we annotated cell subtypes accurately. In comparison to normal brain tissue, the proportions of macrophages and T cells in the tumor microenvironment of gliomas were found to be decreased (Fig. 6). Malignant cells were classified into two clusters based on AOC1 expression levels: a high-expression cluster and a low-expression cluster. The cellular composition of normal organization and glioma has undergone significant alterations, particularly in B cells, macrophages, and T cells. Fig. 6Analysis of the GSE270109 single-cell dataset demonstrating the tumor microenvironment of gliomas. A Annotated cell clusters based on previous literature. B, C Combining single-cell data and cellular annotation results. D Bubble diagram of AOC1 expression among cell clusters. E Percentage of different cell types between normal and tumor groups Analysis of the GSE270109 single-cell dataset demonstrating the tumor microenvironment of gliomas. A Annotated cell clusters based on previous literature. B, C Combining single-cell data and cellular annotation results. D Bubble diagram of AOC1 expression among cell clusters. E Percentage of different cell types between normal and tumor groups AOC1 is predominantly expressed in neoplastic cells. Based on the expression level of AOC1, malignant cells are categorized into two types: AOC1-positive and AOC1-negative cell types. The intercellular communication among the BMP signaling pathway, MIF signaling pathway, PTN signaling pathway, and SPP1 signaling pathway varies in AOC1-positive and AOC1-negative cell types (Fig. 7A-F). Notably, according to the results from the MIF signaling network analysis, AOC1-positive tumor cells exhibited a marked reduction in intercellular communication intensity with macrophages and T cells (Fig. 7D). This finding is consistent with our earlier analysis, which demonstrated that AOC1 inhibits signaling pathways associated with inflammatory factors. Fig. 7Cell-to-cell communication was analyzed using Cellchat. A, B The overall graph of proportions and weights of intercellular communication. C, D, E, F The intercellular communication among the BMP signaling pathway, MIF signaling pathway, PTN signaling pathway, and SPP1 signaling pathway varies in AOC1-positive and AOC1-negative cell types Cell-to-cell communication was analyzed using Cellchat. A, B The overall graph of proportions and weights of intercellular communication. C, D, E, F The intercellular communication among the BMP signaling pathway, MIF signaling pathway, PTN signaling pathway, and SPP1 signaling pathway varies in AOC1-positive and AOC1-negative cell types Prognostic predictive model of AOC1 in GBMLGG To translate our research into clinical practice, we have developed a prognostic nomogram that utilizes both clinical variables and AOC1 levels in patients. Our findings from multivariate COX regression analysis indicate that AOC1 and IDH status are independent risk factors for patient prognosis (Fig. 8A-B). Based on the above findings, we developed a nomogram that accurately predicts the outcomes for patients at 1-year, 3-year, and 5-year (Fig. 8C). Additionally, we employed various methodologies to assess the clinical utility of the nomogram. The calibration curve demonstrated excellent agreement between our model’s predictions and real-world observations (Fig. 8D). Furthermore, both the C-index and time-dependent ROC curve metrics exceeded 0.70, indicating strong predictive performance of the nomogram (Fig. 8E). Consistent conclusion was observed based on CGGA database (Fig. 8F, G). Ultimately, compared to other variables, our nomogram exhibited superior clinical advantages. Fig. 8AOC1 combined with clinical variables for prognostic modeling for clinical patient management. A Multivariate COX regression analysis demonstrated independent prognostic factors affecting patients. B Clinical prognostic modeling predicts patient survival probabilities at 1, 3, and 5 years. C-E Multiple indicators were applied to evaluate the capacity of the model based on TCGA database. F-G Multiple indicators were applied to evaluate the capacity of the model on CGGA database AOC1 combined with clinical variables for prognostic modeling for clinical patient management. A Multivariate COX regression analysis demonstrated independent prognostic factors affecting patients. B Clinical prognostic modeling predicts patient survival probabilities at 1, 3, and 5 years. C-E Multiple indicators were applied to evaluate the capacity of the model based on TCGA database. F-G Multiple indicators were applied to evaluate the capacity of the model on CGGA database RNA silencing of AOC1 inhibits tumor cell proliferation and migration Based on these findings, we further demonstrated the effect of AOC1 on the biological behavior of malignant cells by phenotypic assays. We found that knocking down AOC1 significantly reduced wound closure in glioma cells at 24 h. At the same time, Transwell assays showed that glioma migration rate was reduced after knocking down AOC1 (Fig. 9A-C). These findings strongly suggest that AOC1 may play a complex role in GBMLGG. Fig. 9Knockdown of AOC1 inhibits tumor cell proliferation and migration. A CCK8 assay showed that knockdown of AOC1 inhibited the proliferative activity of tumor cells. B Representative images of wound healing experiments of tumor cells in the control, siRNA-AOC1, and siRNA-NC groups at 0 and 24 h. Initial and post-migration images highlight differences in wound healing between groups. C Transwell assay shows that cell migration in the siRNA-AOC1 group is significantly reduced Knockdown of AOC1 inhibits tumor cell proliferation and migration. A CCK8 assay showed that knockdown of AOC1 inhibited the proliferative activity of tumor cells. B Representative images of wound healing experiments of tumor cells in the control, siRNA-AOC1, and siRNA-NC groups at 0 and 24 h. Initial and post-migration images highlight differences in wound healing between groups. C Transwell assay shows that cell migration in the siRNA-AOC1 group is significantly reduced Discussion AOC1 is a secreted diamine oxidase that degrades polyamines such as putrescine and histamine [7]. Previous studies have shown that AOC is involved in the regulation of various biological processes, including cell proliferation, differentiation, apoptosis and oxidative stress [7–13]. Previous research has demonstrated that AOC1 exhibits elevated expression in colorectal cancer and gastric cancer, and is correlated with unfavorable prognosis [9–13]. However, the expression and prognostic value of AOC1 in glioma have not been clearly studied. Our analysis of GBMLGG data in UCSC XENA revealed a significant upregulation of AOC1 expression in GBMLGG samples. Concurrently, we conducted an analysis of AOC1 expression across diverse subgroups in GBMLGG and identified a significant correlation between elevated AOC1 expression and WHO grade, 1p/19q coding loss, as well as an age exceeding 60 years. By comparing the difference in mRNA expression of gene expression profiles between the AOC1 high-expression group and low-expression group, we found 407 high-expression DEGs and 872 low-expression DEGs. Through the functional enrichment analysis of these differential genes, it was found that AOC1-related genes were mainly related to chemokine signaling pathway, cytokine-cytokine receptor interaction, cytokine activity and IL-17 signaling pathways (Fig. 3A-D). GSEA enrichment analysis provided additional evidence that AOC1 is associated with chemokines and cytokines. We found that the IL-17 signaling pathway is closely related to the high expression of AOC1 in GBMLGG. Dysregulation of the IL-17 signaling pathway can promote cytokine secretion, alter the tumor microenvironment, and correlate with poor prognosis in a variety of tumors. Studies have shown that IL-17 can induce Steap4 to promote TRAF4 expression through p63 + cell expansion, promote keratinocyte proliferation and tumor formation [23]. In skin cancer, specific knockdown of IL-17 reduces Stat3-associated proliferation as well as epidermal cell proliferation [24]. In colorectal cancer, IL-17 A induces the expression of PLET1 to promote colon carcinogenesis [25]. In our research, we found the increased M2 TAMs infiltration in AOC1-high GBMLGG samples. Therefore, we hypothesized that AOC1 may participate in the occurrence and development of GBMLGG through the IL-17 signaling pathway, and further research is needed to explore the specific regulatory mechanism of AOC1 in GBMLGG. The macrophages exert a dichotomous role in tumor initiation and progression [26–28]. On the one hand, macrophages can promote tumor regression by ingesting and degrading tumor cells. Studies have found that when macrophages are stimulated by activation signals, they will release a series of cytokines and chemical substances, attracting other immune cells such as T cells and natural killer cells to the tumor focus to form an immune response, thereby inhibiting tumor growth [26, 29, 30]. Meanwhile, macrophages can also promote the anti-tumor effect of immune cells through mechanisms such as presenting tumor antigens and activating T cells [26–28]. Tumor cells can educate tumor-associated macrophages (TAMs) towards M2 polarization through different mechanisms, to avoid being attacked by the body’s immune system. Studies have shown that tumor cells secrete a variety of immunosuppressive factors, such as tumor necrosis factor β (TGF-β), interleukin-10 (IL-10) and prostaglandin E2 (PGE2), which inhibit the activity of macrophages and reduce its ability to present tumor cells-originated antigens [27, 31–35]. Recently, with the research on TAMs, emerging evidence suggests that TAMs play a critical role in both tumor progression and therapeutic interventions [29, 30, 35, 36]. TAMs have complex functions that can both promote tumor growth and progression and inhibit tumor growth and metastasis. In some cases, TAMs may promote tumor growth, metastasis, and angiogenesis by releasing growth factors, inflammatory factors, and cytokines [32–34, 37–39]. In addition, TAMs can suppress the activity of other immune cells, hindering their attack on tumor cells [28, 31, 37–40]. These functions make M2 TAMs one of the important mechanisms for tumor immune escape, as it can directly facilitate tumor cells growth, or indirectly by repressing the activity of other immune cells [35]. High expression of AOC1 was associated with increased M2 macrophages infiltration, according to current results. Indeed, the latest studies also revealed that AOC1 knockdown inhibits non-small cell lung cancer (NSCLC) cell proliferation and migration, and reduces M2 macrophage polarization [41]. Research findings have indicated that there is a significant upregulation of AOC1 in colorectal cancer, which is linked to an unfavorable prognosis [11]. The expression of AOC1 is increased in gastric cancer, knocking down AOC1 can inhibit the phosphorylation of AKT to inhibit the proliferation, invasion and migration of gastric cancer cells [12]. In nasopharyngeal carcinoma, LINC00839–TAF15–AOC1 axis accelerates tumor growth [42]. At the same time, inhibiting AOC1 can increase the expression of epithelial marker E-cadherin and reduce the expression of mesenchymal Expression of stromal markers N-cadherin, SNAIL and Slug to inhibit epithelial-mesenchymal transition (EMT) in human gastric cancer [12]. Therefore, as a secreted protein, our results implied that AOC1 has the potential to be an early diagnostic indicator of GBMLGG. Further experiments are required to clarify the effect of AOC1 on GBMLGG and the cryptic regulatory mechanism. The outcomes derived from the COX regression analysis demonstrated that AOC1 has the potential to act as an autonomous predictor of adverse prognosis in GBMLGG, functioning independently from other factors. Furthermore, the multivariate Cox regression analysis revealed that elevated AOC1 expression independently contributed to a poorer overall survival (OS). The construction of the nomogram prediction model served as additional confirmation of the prognostic significance of AOC1 expression, affirming its predictive value for prognosis. The calculated area under the curve (AUC) values for the nomogram’s predictive accuracy were higher than 0.7 within the 1 to 5 year projection range. Although the nomogram offers an intuitive graphical representation, its interpretability and utility can be constrained in certain scenarios. In fact, although we performed a comprehensive analysis of AOC1 by bioinformatics, our study was limited to a small sample size. In order to ensure the reliability and rigor of the research results, the sample should be further expanded for research. We should also further investigate the effect of AOC1 on IL-17 signaling in GBMLGG. At present, the laboratory has developed a large number of experimental programs. Conclusion This study revealed for the first time the increased expression of AOC1 in GBMLGG, suggesting that high AOC1 expression was associated with poor prognosis. IL-17 signaling pathway may be an important pathway involved in the regulation of AOC1 in GBMLGG. AOC1 has the potential to be called an early diagnostic indicator of GBMLGG, we need further experiments to study the effect of AOC1 on GBMLGG and explore its potential regulatory mechanism. Supplementary Information Supplementary Material 1. Supplementary Material 1.