Single-Cell and Spatial Transcriptomics Identify Astrocyte Regulation of the Microenvironment and Prognosis in High-Grade Gliomas

Single-Cell and Spatial Transcriptomics Identify Astrocyte Regulation of the Microenvironment and Prognosis in High-Grade Gliomas | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Single-Cell and Spatial Transcriptomics Identify Astrocyte Regulation of the Microenvironment and Prognosis in High-Grade Gliomas Chang Ge, Wenjie Zhang, Pengcheng Zhang, Jingxuan Xu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9156282/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background: High-grade glioma is a highly malignant primary tumor of the central nervous system characterized by a poor clinical prognosis. The regulatory mechanisms of astrocytes within its TME are not fully understood. This study aimed to investigate the gene expression profiles, functional remodeling, prognostic significance, and regulatory role of astrocytes in the immune microenvironment of HGG, thereby identifying novel molecular markers and therapeutic targets for the management of HGG. Methods: ScRNA-seq was conducted on tumor and peritumoral normal tissues from three patients diagnosed with World Health Organization grade III HGG. To characterize astrocyte properties in HGG, spatial transcriptome deconvolution, cell-cell communication analysis, and differential gene enrichment analysis were performed. WGCNA was applied to data from The Cancer Genome Atlas, while Venn intersection combined with LASSO Cox regression was utilized to identify prognostic core genes derived from astrocytes. To assess the prognostic value and immune regulatory function of the identified core genes, Kaplan-Meier survival analysis, receiver operating characteristic curve analysis, and CIBERSORT immune infiltration analysis were executed. Results: In HGG, the proportions of astrocytes, microglia, and tumor cells were significantly elevated, with astrocytes emerging as the predominant cell population within the TME and engaging in extensive signaling interactions with microglia. Astrocytes demonstrated a bidirectional functional remodeling characterized by metabolic activation and the inhibition of matrix interactions. A total of five prognostic core genes derived from astrocytes were identified, among which LTF, NOX4, and HSP90B1 served as independent prognostic risk factors. A risk score model based on these genes effectively differentiated survival outcomes for HGG patients. Furthermore, core genes NDUFB2, HSP90B1, and LITAF were found to regulate the infiltration and polarization of CD8+ T cells and M2 macrophages, thereby contributing to an immunosuppressive TME that facilitates the malignant progression of HGG. Conclusions: Astrocytes are crucial regulators of high-grade glioma HGG TME remodeling, exhibiting distinct functional changes and molecular mechanisms. The identified prognostic core genes derived from astrocytes may serve as potential molecular markers for assessing HGG prognosis and as candidate targets for astrocyte-targeted antitumor therapies, thereby offering new insights for precision treatment of HGG. Trial registration: Not applicable. High-grade glioma Astrocytes Tumor microenvironment Prognostic marker Immune infiltration Cell-cell communication Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Background Glioma is the most common primary malignant tumor of the central nervous system (CNS), characterized by rapid growth, strong invasiveness, high recurrence rate, and poor prognosis [ 1 , 2 ] . According to the World Health Organization (WHO) classification criteria, gliomas are graded from grade Ⅰ to Ⅳ based on malignancy, with WHO grades Ⅲ and Ⅳ categorized as high-grade gliomas (HGG) [ 3 ] . Currently, the standard treatment regimen for HGG remains the Stupp protocol, which consists of surgical resection combined with radiotherapy and temozolomide-based chemotherapy. However, even with standardized treatment, the overall prognosis of HGG patients remains unsatisfactory, with a limited median survival time and persistently high recurrence rate [ 4 , 5 ] . Therefore, in-depth exploration of the molecular mechanisms underlying HGG initiation and progression, as well as identification of potential key regulatory targets, are crucial for developing novel targeted therapeutic strategies and improving patient outcomes. Astrocytes are the most abundant glial cell type in the CNS, playing pivotal roles in nervous system development and homeostasis maintenance. Their functions include maintaining the blood-brain barrier (BBB), storing and distributing energy substrates to support neuronal activity, and promoting neural cell survival and synaptogenesis [ 6 , 7 ] . Additionally, astrocytes regulate inflammatory responses and neurodegenerative processes in the CNS through multiple mechanisms, such as releasing neurotoxic factors, modulating microglial activity, and participating in the recruitment and infiltration of inflammatory cells [ 8 – 11 ] . However, systematic studies on the key genes in astrocytes that are closely associated with HGG patient prognosis and their specific roles in reshaping the tumor immune microenvironment are still lacking. In this study, we integrated single-cell RNA sequencing (scRNA-seq) technology with transcriptomic data from public databases to systematically decipher the gene expression profiles of astrocytes in HGG tissues. Firstly, scRNA-seq was performed on HGG tumor tissues to identify differentially expressed genes (DEGs) in astrocytes. Subsequently, public transcriptomic datasets were integrated, and weighted gene co-expression network analysis (WGCNA) was employed to screen key modules closely related to HGG initiation and progression. On this basis, combined with clinical prognosis information of patients from public databases, we further identified core astrocyte-derived genes significantly associated with HGG patient prognosis. Finally, immune infiltration analysis was conducted to explore the potential regulatory roles of these key genes in the HGG tumor immune microenvironment, aiming to reveal the molecular mechanisms by which astrocytes contribute to HGG progression. This study provides a theoretical basis for a deeper understanding of the composition of the HGG immune microenvironment, exploration of novel targeted therapeutic strategies, and identification of prognostic biomarkers. Methods 1 Human Tumor Specimen Collection Glioma samples and peritumoral normal tissues were collected from patients who underwent surgical treatment at the Second Affiliated Hospital of Xinjiang Medical University. All specimens were collected in accordance with national standards. This study was approved by the Ethics Committee of the Second Affiliated Hospital of Xinjiang Medical University, and written informed consent was obtained from all patients whose specimens were included in the study. 2 Single-Cell RNA Sequencing (scRNA-seq) Data Collection Tumor samples were thoroughly rinsed with phosphate-buffered saline (PBS) to remove blood, then dissected with a surgical blade and washed in pre-chilled RPMI 1640 medium. Tissue dissociation was performed using a collagenase cocktail, followed by red blood cell lysis with a red blood cell lysis buffer (BL503A, Biosharp, China). The cells were collected by centrifugation, and cell count and viability were assessed using a fluorescent cell analyzer (Countstar® Rigel S2, ALIT Life Science, China), after which dead cell removal (Miltenyi, 130-090-101, Germany) was performed as needed. Finally, fresh cells were washed twice with RPMI 1640 medium and resuspended in 1×PBS containing 0.04% bovine serum albumin (BSA). Libraries for single-cell/nucleus RNA sequencing were constructed from the successfully dissociated single-cell/nucleus suspension using the DNBelab C Series High-Throughput scRNA Library Prep Kit V3.0. First, water-in-oil emulsions were generated with the DNBelab C-TaiM 4 system, and reverse transcription was performed on qualified emulsions. Subsequently, demulsification was conducted, and the cDNA products were purified. The target cDNA was subjected to fragmentation, end repair, adaptor ligation, amplification and purification to generate the BGI single-cell/nucleus transcriptome library. The library concentration was quantified using a Qubit 4.0 Fluorometer, and the libraries were stored at -80°C. Paired-end sequencing was performed on the constructed libraries using the DNBSEQ-T7 sequencing platform to obtain high-quality transcriptomic data. 3 Quality Control of scRNA-seq Data scRNA-seq data were processed and analyzed following standard protocols. Raw sequencing reads were aligned to the human reference genome GRCh38 using Cell Ranger (v5.0, 10x Genomics), and gene expression was quantified using unique molecular identifiers (UMIs). For data generated by the BGI sequencing platform, preprocessing was performed using the dnbc4tools software with similar parameters. Downstream analyses were conducted on the filtered gene expression matrix using the Seurat software (v4.0.0). Low-quality cells were filtered out based on the following criteria: number of detected genes 20%, or ribosomal gene transcript proportion > 50%. Meanwhile, genes detected in fewer than 3 cells were excluded. Potential doublets were identified and removed using the DoubletFinder software (v2.0.3), with the expected doublet rate set to 7.5% based on the Poisson distribution. 4 Normalization and Dimensionality Reduction UMI counts were normalized using the LogNormalize method (with a scale factor of 10,000) and log-transformed. The top 2,000 highly variable genes were selected for principal component analysis (PCA). Batch effects between different samples or experimental conditions were corrected using the Harmony software (v1.0). The number of principal components (PCs) used for subsequent analyses was determined via an elbow plot, and the first 30 PCs were selected in this study. Cell clustering was performed using the graph-based Louvain algorithm with a resolution parameter of 0.4 (adjustable from 0.1 to 1 according to dataset size). Marker genes for each cell cluster were identified using the FindAllMarkers function, with screening thresholds set as |avg_log2FC| ≥ 0.5 and adjusted P-value ≤ 0.05. 5 Cell Feature Selection and Annotation To distinguish malignant cells from normal cells, genome-wide copy number variations (CNVs) were inferred from gene expression data using the CopyKat software (v0.1.0). Aneuploid cells were classified as tumor cells, while diploid cells were considered normal stromal or immune cells, and this classification was validated by the expression of known lineage-specific marker genes. Cell type annotation was performed by combining automated tools with manual verification based on classic marker genes. 6 Spatial Transcriptome and Cell-Cell Communication Analysis Public spatial transcriptome data (GSE253080) were used for deconvolution analysis with scRNA-seq data as the reference on spatial transcriptome spots. The cell type proportion of each spot in the spatial transcriptome was obtained after deconvolution, and the cell type with the highest proportion in each spot was defined as the dominant cell type of that spot. Based on the above cell type annotation results, cell-cell communication analysis was performed on the spatial transcriptome data. 7 Differential Expression Analysis To identify differentially expressed genes (DEGs) in astrocytes, the DESeq2 algorithm was applied to analyze the scRNA-seq transcriptomic data. First, preprocessing was performed on the raw gene expression count matrix using the built-in normalization method of DESeq2 following its standard analysis pipeline to eliminate systematic biases caused by factors such as sequencing depth variation. DEGs were then screened via statistical tests with the criteria of adjusted P-value 1. 8 Enrichment Analysis To further elucidate the biological functions of DEGs, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed on the screened DEGs. GO analysis annotated gene functions from three dimensions: Biological Process (BP), Cellular Component (CC) and Molecular Function (MF). DEGs were mapped to each term in the GO database, and the number of genes in each term was counted. Using genome-wide GO annotations as the background, hypergeometric distribution tests were performed to identify significantly enriched GO terms for DEGs. KEGG pathway enrichment analysis was used to characterize the major metabolic and signal transduction pathways involved in DEGs [12–14]. DEGs were mapped to the KEGG database, and the number of genes in each pathway was counted. Significantly enriched KEGG pathways were screened using hypergeometric distribution tests with genome-wide KEGG annotations as the background. For data visualization, bubble plots were generated using the ggplot2 package in R to intuitively display the distribution of significantly enriched GO terms and KEGG pathways. 9 Weighted Gene Co-Expression Network Analysis (WGCNA) Officially batch-corrected TCGA pan-cancer transcriptomic data were downloaded from the UCSC Toil RNA-seq Recompute database ( https://xenabrowser.net/datapages/ ). Transcriptomic data of low-grade glioma (LGG) and glioblastoma (GBM) — the main subtype of high-grade glioma (HGG) — were extracted, including 5 normal control samples, 523 LGG samples and 166 GBM samples. The expression matrix of protein-coding genes was further filtered for subsequent WGCNA analysis. 10 Gene Screening via Machine Learning The optimal regularization parameter lambda was selected using 10-fold cross-validation, and the conservative lambda.1se was adopted to reduce the risk of overfitting. Genes with non-zero coefficients were extracted as key features. A multi-gene Cox proportional hazards model was then constructed to validate the prognostic value of the selected genes. Risk scores were calculated, and Kaplan-Meier (KM) curves were plotted for risk groups for visualization. 11 Receiver Operating Characteristic (ROC) Curve Analysis To evaluate the prognostic predictive value of hub genes for HGG patients, ROC curve analysis was performed based on hub gene expression levels and clinical prognostic information of TCGA-GBM patients. ROC curves were plotted using the pROC package in R, and the area under the curve (AUC) was calculated to quantify the ability of hub genes to distinguish the prognostic status of HGG patients [ 15 ] . 12 CIBERSORT Analysis To estimate the relative abundances of tumor-infiltrating immune cells in tumor masses, the reference set of signature gene profiles for 22 immune cell subtypes provided by the online analysis platform CIBERSORT ( https://cibersort.stanford.edu ) was used. Spearman correlation analysis was performed to identify immune cell types significantly correlated with key hub genes. Results 1 ScRNA-seq Profiling of HGG Tumor Tissues scRNA-seq was performed on tumor tissue samples from 3 patients with WHO grade Ⅲ HGG and their matched peritumoral normal tissues, all of which were pathologically confirmed as HGG after surgery (Fig. 1 A-B). A total of 46,181 high-quality cells were identified and clustered into 17 cell clusters via manual verification of classic marker genes combined with the automated annotation tool SingleR (v1.8.30) (Fig. 1 C-D). The FindAllMarkers() function was used to compare gene expression differences between each cell cluster and all other clusters for screening potential specific marker genes. Screening criteria comprehensively considered the average log2 fold change (avg_log2FC), adjusted P -value (p.adj), and difference in expression percentage (diff.pct); the top 3 genes of each cluster were extracted as representative markers in the order of decreasing avg_log2FC and increasing P -value (Fig. 1 E). Based on the expression distribution of known classic marker genes, 16 cell clusters were annotated into 9 major cell types: Oligodendrocytes, T cells, Microglia, natural killer (NK) cells, Neurons, Astrocytes, Proliferating cells, Pericytes, and Endothelial cells (Fig. 1 F). Compared with peritumoral normal tissues, the proportions of Astrocytes, Microglia, and tumor cells were significantly increased in HGG tissues (Fig. 1 G-H). (A) UMAP projection showing the distribution of all single cells, color-coded by biological replicate samples to clearly distinguish cell populations between glioma and normal brain tissue samples. (B) Cell population distribution colored by sample group, intuitively presenting the differences in cell population distribution between HGG and normal tissues. (C) Transcriptomic profile colored by cell cycle phase, showing the distribution characteristics of cells in G1, G2M, and S phases. (D) Unsupervised clustering results by Seurat, labeled with 17 cell clusters (0–16), providing a basis for subsequent cell type annotation. (E) Expression characteristics of key marker genes in each cell type; the size of the dots represents the proportion of cells expressing the gene, and the color gradient represents the average expression level, assisting in the accurate annotation of cell types. (F) Cell population distribution colored by custom cell type classification, clearly annotating major cell types including Astrocytes, Endothelial cells, Microglia, Neurons, NK cells, Oligodendrocytes, Pericytes, Proliferating cells, and T cells. (G) Percentage composition of each cell type in each biological replicate sample, intuitively reflecting the differences in cell population composition among different samples. (H) Comparison of the proportion of each cell type between glioma and normal tissues; the y-axis represents the cell proportion, and box plots show the distribution range, median, and quartiles of each group, revealing the cellular composition characteristics and intergroup differences of the tumor microenvironment. 2 Spatial Transcriptomics Deciphers the HGG Microenvironment: Astrocytes as the Dominant Cell Population We further integrated the above scRNA-seq data with the public spatial transcriptome dataset GSE253080 to perform deconvolution analysis. The cell type proportion of each spot in the spatial transcriptome was obtained after deconvolution, and the cell type with the highest proportion in each spot was defined as the dominant cell type of that spot (Fig. 3 A-B). The results showed that Astrocytes, Microglia, and Proliferating cells were the top three cell types in terms of proportion. Based on these three core cell types, we further performed cell-cell communication analysis and found that there was intensive signal crosstalk between Astrocytes and Microglia (Fig. 3 C-E). In summary, Astrocytes serve as the major cellular component in HGG and can exert a crucial regulatory effect on the tumor immune microenvironment by targeting and modulating Microglia. (A) Cell density heatmap (nCount_Spatial) of spatial transcriptome sections; the color gradient shows the distribution of transcript counts in different regions, reflecting the cell density and transcriptome capture efficiency of the tissue sections. (B) Spatial distribution map of cell types in spatial transcriptome sections; Astrocytes (red), Microglia (green), and Pericytes (blue) were annotated via deconvolution analysis, intuitively presenting the spatial distribution characteristics of the three cell types in the tumor microenvironment. (C) Heatmap of cell-cell communication analysis; the left panel shows the number of interactions among Astrocytes, Microglia, and Pericytes, and the right panel shows the interaction strength; the color gradients correspond to the levels of interaction number and strength, respectively, and marginal bar charts show the total interaction number and strength of each cell type. (D) Cell-cell communication network diagram; nodes represent cell types, and the thickness of edges represents the number of interactions, clearly presenting the communication connection pattern among the three cell types. (E) Cell-cell communication network diagram; nodes represent cell types, and the thickness of edges represents the interaction strength, intuitively showing the differences in the strength of communication between different cell pairs. 3 DEGs of Astrocytes and Their Enrichment Analysis in the HGG Microenvironment KEGG pathway enrichment analysis of Astrocyte DEGs showed that the upregulated genes were mainly enriched in ribosome, antigen processing and presentation, and oxidative phosphorylation as the core pathways; additionally, pathways associated with neurodegenerative diseases such as Huntington's disease, Parkinson's disease, and Alzheimer's disease were also significantly enriched, along with metabolic pathways including non-alcoholic fatty liver disease and mineral absorption (Fig. 4 A). The downregulated genes were primarily enriched in axon guidance, tight junction and other pathways, and cancer-related pathways as well as infectious disease-related pathways were also identified (Fig. 4 B). GO enrichment analysis was performed on the DEGs of Astrocyte subsets separately. The results showed that the upregulated genes (Fig. 4 C) were concentrated in cellular metabolic and biosynthetic processes such as cytoplasmic translation and ribosomal small subunit biogenesis in biological processes; enriched in binding and catalytic activities such as actin binding and calcium ion binding in molecular functions; and localized to intracellular organelles such as mitochondrial matrix and ribosomal subunits in cellular components. The activation of these metabolism- and ribosome-related functions was highly consistent with the pathological characteristics of energy metabolism disorder and abnormal protein synthesis in neurodegenerative diseases. The downregulated genes (Fig. 4 D) focused on protein assembly and cell adhesion processes such as NADH dehydrogenase complex assembly and cadherin binding in biological processes; enriched in extracellular matrix (ECM)-related functions such as ECM structural constituent and proteoglycan binding in molecular functions; and concentrated in extracellular and membrane-associated structures such as ECM and basement membrane in cellular components. The downregulation of such cell adhesion and ECM interaction functions may be involved in the pathological processes of abnormal neuron-glia connection and synaptic structure damage in neurodegenerative diseases. (A) Scatter plot of KEGG pathway enrichment analysis for upregulated genes; the y-axis represents the names of enriched KEGG pathways, the x-axis represents the -log10-transformed adjusted P -value, the size of the dots represents the number of enriched genes, and the color gradient corresponds to the level of the adjusted P -value. (B) Scatter plot of KEGG pathway enrichment analysis for downregulated genes. (C) Scatter plot of GO enrichment analysis for upregulated genes, displayed from three dimensions: biological process, molecular function, and cellular component; the y-axis represents the names of GO terms in each dimension, the x-axis represents the -log10-transformed adjusted P -value, the size of the dots represents the number of enriched genes, and the color gradient corresponds to the level of the adjusted P -value. (D) Scatter plot of GO enrichment analysis for downregulated genes. 4 Screening of Highly Correlated Modules and Genes via WGCNA To screen highly correlated transcriptomic modules, we further utilized TCGA-LGG and TCGA-GBM transcriptomic data, including 5 normal control samples, 523 LGG samples, and 166 GBM samples, for WGCNA. When the soft threshold was set to 6, the scale-free topology model fit index reached above 0.8, and the mean connectivity showed a steady decline with the increase of the soft threshold, indicating that this parameter selection could effectively construct a scale-free co-expression network (Fig. 5 A). Hierarchical clustering was performed on the topological overlap matrix (TOM) to cluster highly correlated genes into different co-expression modules. The hierarchical clustering dendrogram showed that genome-wide genes were successfully divided into multiple modules with specific expression patterns, labeled with different color bands; a total of 27 gene modules were screened out (Fig. 5 B-C). Clustering analysis of module eigengenes further revealed the expression correlation among each module (Fig. 5 D). Finally, as shown in Fig. 6 , two modules (pink and red) were identified with a significant high positive correlation with GBM, containing 709 and 1,085 genes, respectively. (A) WGCNA soft threshold screening plot: the left panel shows the curve of scale-free topology model fit index with the change of soft threshold; the x-axis represents the soft threshold value, the y-axis represents the model fit index R², and the red horizontal line marks the threshold standard of R²=0.8. The right panel shows the curve of mean connectivity with the change of soft threshold; the x-axis represents the soft threshold value, the y-axis represents the mean connectivity value, used to determine the optimal soft threshold for constructing a scale-free network. (B) TOM heatmap and gene clustering tree: the left panel is the gene hierarchical clustering tree based on TOM, showing the clustering relationship of genes; the right panel is the TOM matrix visualization result, the color gradient reflects the level of topological overlap between genes, with red representing high topological overlap and yellow representing moderate topological overlap, providing a basis for module division. (C) Gene module clustering dendrogram: the y-axis represents the clustering height, the x-axis represents gene samples; genes with similar expression patterns were clustered into different modules via hierarchical clustering, and the color bands at the bottom indicate the module to which each gene belongs. (D) Gene co-expression adjacency heatmap and hierarchical clustering tree: the upper panel is the hierarchical clustering tree based on gene expression similarity, showing the genetic relationship among genes; the lower panel is the module eigengene adjacency heatmap, the color gradient represents the level of adjacency coefficient among genes, with red indicating high adjacency and blue indicating low adjacency. Module-trait relationship heatmap showing the correlation between gene co-expression modules identified by WGCNA and glioma pathological phenotypes; the y-axis represents the names of each gene module, the x-axis represents pathological phenotype traits, and the color gradient of the heatmap represents the value of the correlation coefficient. 5 Screening of Astrocyte-Derived Prognostic Genes for HGG Patients via Machine Learning To further explore the Astrocyte DEGs associated with the prognosis of HGG patients, Venn diagram intersection analysis was performed between the module genes screened by WGCNA and the upregulated (SC_Astrocyte.up) and downregulated (SC_Astrocyte.down) genes of Astrocyte subsets (Fig. 7 A), resulting in a total of 73 DEGs identified. These genes were further subjected to machine learning combined with survival information (Fig. 7 B). The results showed that the gene regression coefficients gradually shrank with the increase of the penalty coefficient λ; when λ was set to the optimal value (logλ≈-3) (Fig. 7 B), a total of 5 core genes with prognostic value (LTF, NOX4, MDK, HSP90B1, and HSPB1) were finally screened out. Univariate Cox regression analysis showed that LTF, NOX4, and HSP90B1 had a statistically significant risk association with HGG prognosis (Fig. 7 C). Based on these 3 genes, glioma patients were divided into high-risk and low-risk groups by the median risk score. Kaplan-Meier survival analysis showed that the overall survival rate of patients in the high-risk group was significantly lower than that in the low-risk group ( P < 0.001) (Fig. 7 D), suggesting that this core gene panel can effectively predict the prognostic outcome of glioma patients and provide a potential molecular marker panel for glioma prognosis assessment. (A) Venn diagram of core gene intersection analysis (overlap): the intersection distribution of module genes identified by WGCNA in the TCGA database, upregulated genes (SC_Astrocyte.up) and downregulated genes (SC_Astrocyte.down) of Astrocyte subsets. (B) LASSO Cox regression coefficient profile plot: the x-axis represents the logλ-transformed penalty coefficient, the y-axis represents the gene regression coefficient; the red line represents the change trend of each gene's regression coefficient with the increase of the penalty coefficient, and the color gradient bar chart shows the gene count corresponding to each λ value. (C) Forest plot of hazard ratio (HR) from univariate Cox regression analysis of core genes: the x-axis represents the HR and 95% confidence interval (CI), the y-axis represents the names of core genes; the black horizontal line for each gene represents the 95% CI, the black square represents the HR value, and the P -value is labeled on the right. (D) Kaplan-Meier survival curve of core gene risk scores (KM Plot by Median Risk Group): the x-axis represents survival time (days), the y-axis represents survival probability; the red curve represents the high-risk group, and the blue curve represents the low-risk group (divided by the median risk value). 3.6 Immune Infiltration Analysis Correlation analysis was performed to explore the association between the expression levels of three core genes (NDUFB2, HSP90B1, and LITAF) and immune cell infiltration in the glioma microenvironment. The expression of NDUFB2 was significantly positively correlated with the infiltration of multiple immune cells, among which the positive correlations with γδ T cells, eosinophils, and CD8 + T cells were the most significant; it was significantly negatively correlated with M0 macrophages and resting NK cells (Fig. 8 A). The expression of HSP90B1 was significantly positively correlated with CD4 + memory activated T cells, regulatory T cells (Tregs), and M2 macrophages; it was significantly negatively correlated with eosinophils, activated NK cells, and M1 macrophages (Fig. 8 B). The expression of LITAF was significantly positively correlated with CD4 + activated T cells, resting dendritic cells, and monocytes; it was significantly negatively correlated with follicular helper T cells, M0 macrophages, and resting mast cells (Fig. 8 C). (A) Correlation analysis plot of NDUFB2 gene expression and immune cell infiltration in glioma. (B) Correlation analysis plot of HSP90B1 gene expression and immune cell infiltration in glioma. (C) Correlation analysis plot of LITAF gene expression and immune cell infiltration in glioma. Discussion As one of the most invasive malignant tumors of the CNS, HGG is characterized by a complex TME composition and intercellular interaction network, which are the core causes of treatment resistance and poor prognosis [ 1 , 2 ] . Astrocytes, the most abundant glial cell type in the CNS, not only participate in the maintenance of neural homeostasis but also act as key regulators of the TME through phenotypic transformation and functional remodeling during tumor progression [ 6 , 7 ] . In this study, we integrated scRNA-seq, spatial transcriptome sequencing, and public database analysis to systematically decipher the gene expression characteristics of Astrocytes in HGG, their spatial interaction patterns with other cells, screen prognosis-related core genes, and clarify their regulatory effects on the immune microenvironment. This study provides an important theoretical basis for an in-depth understanding of the pathogenesis of HGG and the development of novel targeted therapeutic strategies. Firstly, scRNA-seq was used to clarify the cellular composition characteristics of HGG tissues, and we found that the proportions of Astrocytes, Microglia, and tumor cells were significantly increased in HGG compared with peritumoral normal tissues, suggesting that these three cell types may jointly participate in the remodeling of the TME. Spatial transcriptome deconvolution analysis further confirmed that Astrocytes are the dominant cell population in the HGG microenvironment and have intensive signal communication with Microglia, which is consistent with the conclusion of previous studies that "Astrocyte-Microglia crosstalk regulates tumor progression" [ 16 , 17 ] . Cell-cell communication network analysis showed that the number and strength of interactions between Astrocytes and Microglia were significantly higher than those of other cell combinations, implying that the two cell types may form a functional complex through paracrine signals and direct intercellular contact to jointly regulate the formation of the tumor immunosuppressive microenvironment, which provides a clear direction for the subsequent exploration of the molecular mechanism of intercellular interaction. DEG enrichment analysis revealed the functional abnormal characteristics of Astrocytes in HGG: the upregulated genes were mainly enriched in metabolism-related pathways such as ribosome and oxidative phosphorylation, and neurodegenerative disease-related pathways such as Huntington's disease and Parkinson's disease were also significantly enriched, suggesting that the metabolic reprogramming of Astrocytes may share common molecular mechanisms with the pathological changes of neurodegenerative diseases [ 18 ] . The downregulated genes focused on processes such as ECM organization and cell adhesion, and the loss of these functions may damage the integrity of neuron-glia connections and the blood-brain barrier, facilitating the invasion and metastasis of tumor cells [ 19 ] . These results indicated that Astrocytes in HGG undergo bidirectional functional remodeling of "metabolic activation and ECM interaction inhibition", creating a favorable microenvironmental condition for tumor progression and providing a new molecular perspective to explain the high invasiveness of HGG. To screen core genes associated with HGG prognosis, WGCNA was performed on the TCGA database to identify pink and red modules with a high positive correlation with HGG. Further intersection analysis with Astrocyte DEGs and LASSO Cox regression screening finally identified 5 core genes (LTF, NOX4, MDK, HSP90B1, and HSPB1). Among them, LTF, NOX4, and HSP90B1 were confirmed as independent risk factors for the prognosis of HGG patients, and their high expression was significantly associated with short overall survival of patients. The risk score model based on these three genes could effectively distinguish high-risk and low-risk patients ( P < 0.001), suggesting that this gene panel has potential value as a molecular marker for HGG prognosis. Previous studies have shown that HSP90B1, a member of the heat shock protein family, is involved in the anti-apoptotic process of tumor cells by regulating protein folding and stability [ 20 ] ; NOX4 promotes tumor angiogenesis and invasion by generating reactive oxygen species (ROS) [ 18 , 21 ] . This study for the first time confirmed that these two genes are closely associated with Astrocyte dysfunction and HGG prognosis, expanding the understanding of their functions in the tumor microenvironment. Immune infiltration analysis further revealed the regulatory effects of core genes on the HGG immune microenvironment: high expression of NDUFB2 was positively correlated with the infiltration of anti-tumor immune cells such as CD8 + T cells and γδ T cells, and negatively correlated with M0 macrophage infiltration, suggesting that it may improve patient prognosis by enhancing anti-tumor immune responses. HSP90B1 was positively correlated with the infiltration of immunosuppressive cells such as Tregs and M2 macrophages, and negatively correlated with M1 macrophage infiltration, which is consistent with the known mechanism that M2 macrophages and Tregs promote tumor immune escape by inhibiting the function of effector T cells[22,23]. High expression of LITAF was positively correlated with the infiltration of CD4 + activated T cells and negatively correlated with follicular helper T cell infiltration, implying that it may participate in the balance of the immune microenvironment by regulating the polarization of T cell subsets. The above results indicated that Astrocyte-derived core genes can shape the HGG immune microenvironment by regulating the infiltration pattern of immune cells, and their abnormal expression may lead to an immunosuppression-dominant microenvironmental phenotype, thereby promoting tumor progression and treatment resistance. Conclusion In this study, we comprehensively investigated the biological roles of astrocytes in the progression and prognosis of HGG by integrating single-cell RNA sequencing, spatial transcriptome analysis, and multi-dimensional bioinformatics mining of public omics datasets. Our findings first characterized the cellular composition of the HGG microenvironment, confirming that astrocytes are a dominant cell population in HGG tissues with significantly increased proportions, and that astrocytes and microglia form a core cell communication pair with intensive intercellular crosstalk, which is a key structural basis for remodeling the HGG tumor microenvironment. Differential gene and enrichment analyses further revealed the distinct functional remodeling features of HGG-associated astrocytes: the activation of metabolic pathways such as ribosome and oxidative phosphorylation, together with the enrichment of neurodegenerative disease-related signaling, and the downregulation of extracellular matrix organization and cell adhesion-related functions. This bidirectional functional alteration not only reflects the functional abnormality of astrocytes in HGG but also creates a favorable microenvironmental condition for tumor cell invasion and immune suppression. Through WGCNA and machine learning strategies, we screened out astrocyte-derived core genes (LTF, NOX4, HSP90B1, MDK, HSPB1) associated with HGG prognosis, among which LTF, NOX4 and HSP90B1 were verified as independent prognostic risk factors for HGG patients. The risk score model constructed by these three genes can effectively distinguish the survival outcome of HGG patients, providing a novel potential molecular marker panel for clinical prognostic assessment of HGG. Immune infiltration analysis further elucidated the regulatory mechanism of astrocyte-derived core genes on the HGG immune microenvironment: NDUFB2, HSP90B1 and LITAF exert distinct regulatory effects on the infiltration and polarization of tumor-infiltrating immune cells (including T cell subsets, macrophages, dendritic cells, etc.), and their abnormal expression drives the formation of an immunosuppression-dominant tumor microenvironment, thereby promoting HGG progression and treatment resistance. These results confirm that astrocyte-derived core genes are key regulators of the HGG immune microenvironment, and clarify the molecular link between astrocyte dysfunction and immune microenvironment remodeling in HGG. Overall, our study identifies the critical regulatory role of astrocytes in HGG, reveals the functional characteristics of astrocytes in HGG and their molecular mechanism of regulating tumor immune microenvironment, and screens out prognostic core genes with potential clinical application value. These findings not only deepen the understanding of the molecular mechanism of HGG progression mediated by stromal cells such as astrocytes, but also provide new candidate targets for the development of astrocyte-targeted therapeutic strategies and the optimization of HGG prognostic evaluation systems. Limitations of this study include the small sample size of clinical specimens for single-cell sequencing and the lack of in vitro and in vivo experimental validation of the core genes' functions. Future research will expand the sample size, and verify the biological functions and regulatory mechanisms of the core genes through cell experiments and animal models, and further explore the specific molecular signaling pathways of astrocyte-microglia crosstalk in the HGG microenvironment, so as to provide more solid experimental basis for translating these findings into clinical practice. Abbreviations HGG High-grade glioma LGG Low-grade glioma GBM Glioblastoma WHO The World Health Organization CNS central nervous system BBB blood-brain barrier scRNA-seq single-cell RNA sequencing ST-seq spatial transcriptomic sequencing GO Three letter acronym KEGG Kyoto Encyclopedia of Genes and Genomes WCGNA weighted gene co-expression network analysis PCA principal component analysis DEG differentially expressed genes BP Biological Process CC Cellular Component MF Molecular Function Declarations Ethics approval and consent to participate This study was approved by the Ethics Committee of the Second Affiliated Hospital of Xinjiang Medical University (approval number: [2025022624]). All procedures performed in studies involving human participants were in accordance with the ethical standards of the national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Written informed consent was obtained from all individual participants included in the study. Clinical trial number: not applicable Consent for publication not applicable Availability of data and materials The datasets generated and/or analysed during the current study are available in the Genome Sequence Archive (GSA) at the National Genomics Data Center (NGDC) repository, https://ngdc.cncb.ac.cn/omix/preview/RO7wSvN1 . Competing interests The authors declare that they have no competing interests Funding This research was funded by Xinjiang Uygur Autonomous Region Collaborative Innovation Special Project, grant number 2022E02060 and Open Project of the State Key Laboratory of High Incidence Diseases Prevention and Treatment in Central Asia, grant number SKL-HIDCA-2022-NKX5. Authors' contributions J.X.X. proofread and wrote the manuscript. C.G. wrote the manuscript and prepared all figures. W.J.Z. and P.C.Z. collected and organized the data. All authors reviewed and approved the final manuscript. Acknowledgements We would like to express our sincere gratitude to Doubao AI for its valuable contributions to the linguistic polishing and translation work of this manuscript, which has significantly improved the accuracy and fluency of the text. Additionally, we thank the participants who donated clinical specimens for this study and the medical staff of the Second Affiliated Hospital of Xinjiang Medical University for their support in sample collection and processing. This work was supported by the Xinjiang Uygur Autonomous Region Collaborative Innovation Special Project (Grant No. 2022E02060) and the Open Project of the State Key Laboratory of High Incidence Diseases Prevention and Treatment in Central Asia (Grant No. SKL-HIDCA-2022-NKX5). References Ostrom Q T, Price M, Neff C, et al. CBTRUS Statistical Report: Primary Brain and Other Central Nervous System Tumors Diagnosed in the United States in 2015-2019[J]. Neuro-Oncology, 2022, 24(Suppl 5): v1-v95. Zhou Y, Xiao D, Jiang X, et al. EREG is the core onco-immunological biomarker of cuproptosis and mediates the cross-talk between VEGF and CD99 signaling in glioblastoma[J]. Journal of Translational Medicine, 2023, 21(1): 28. Louis D N, Perry A, Wesseling P, et al. The 2021 WHO Classification of Tumors of the Central Nervous System: a summary[J]. NEURO-ONCOLOGY, 2021, 23(8): 1231-1251. Liu Y, Wu Z, Feng Y, et al. Integration analysis of single-cell and spatial transcriptomics reveal the cellular heterogeneity landscape in glioblastoma and establish a polygenic risk model[J]. Frontiers in Oncology, 2023, 13: 1109037. Wu L, Wu W, Zhang J, et al. Natural Coevolution of Tumor and Immunoenvironment in Glioblastoma[J]. Cancer Discovery, 2022, 12(12): 2820-2837. Ferris H A, Perry R J, Moreira G V, et al. Loss of astrocyte cholesterol synthesis disrupts neuronal function and alters whole-body metabolism[J]. Proceedings of the National Academy of Sciences of the United States of America, 2017, 114(5): 1189-1194. Allen N J, Eroglu C. Cell Biology of Astrocyte-Synapse Interactions[J]. Neuron, 2017, 96(3): 697-708. Linnerbauer M, Wheeler M A, Quintana F J. Astrocyte Crosstalk in CNS Inflammation[J]. Neuron, 2020, 108(4): 608-622. Liddelow S A, Barres B A. Reactive Astrocytes: Production, Function, and Therapeutic Potential[J]. Immunity, 2017, 46(6): 957-967. Mayo L, Trauger S A, Blain M, et al. Regulation of astrocyte activation by glycolipids drives chronic CNS inflammation[J]. Nature Medicine, 2014, 20(10): 1147-1156. Wheeler M A, Jaronen M, Covacu R, et al. Environmental Control of Astrocyte Pathogenic Activities in CNS Inflammation[J]. Cell, 2019, 176(3): 581-596.e18. Kanehisa M, Goto S. KEGG: kyoto encyclopedia of genes and genomes[J]. Nucleic Acids Research, 2000, 28(1): 27-30. Kanehisa M, Furumichi M, Sato Y, et al. KEGG: biological systems database as a model of the real world[J]. Nucleic Acids Research, 2025, 53(D1): D672-D677. Kanehisa M. Toward understanding the origin and evolution of cellular organisms[J]. Protein Science: A Publication of the Protein Society, 2019, 28(11): 1947-1951. Verhaak R G W, Hoadley K A, Purdom E, et al. Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1[J]. Cancer Cell, 2010, 17(1): 98-110. Saad S H, Kashanchi A, Zadeh M A, et al. Exosome-Mediated Crosstalk Between Cancer Cells and Tumor Microenvironment[J]. Cells, 2025, 14(22): 1750. Luo H, Zhang H, Mao J, et al. Exosome-based nanoimmunotherapy targeting TAMs, a promising strategy for glioma[J]. Cell Death & Disease, 2023, 14(4): 235. Sun M, You H, Hu X, et al. Microglia-Astrocyte Interaction in Neural Development and Neural Pathogenesis[J]. Cells, 2023, 12(15): 1942. Li X, Gou W, Zhang X. Neuroinflammation in Glioblastoma: Progress and Perspectives[J]. Brain Sciences, 2024, 14(7): 687. Catalano M, Limatola C, Trettel F. Non-neoplastic astrocytes: key players for brain tumor progression[J]. Frontiers in Cellular Neuroscience, 2023, 17: 1352130. Lin C, Wang N, Xu C. Glioma-associated microglia/macrophages (GAMs) in glioblastoma: Immune function in the tumor microenvironment and implications for immunotherapy[J]. Frontiers in Immunology, 2023, 14: 1123853. Lin C, Wang N, Xu C. Glioma-associated microglia/macrophages (GAMs) in glioblastoma: Immune function in the tumor microenvironment and implications for immunotherapy[J]. Frontiers in Immunology, 2023, 14: 1123853. Tao J C, Yu D, Shao W, et al. Interactions between microglia and glioma in tumor microenvironment[J]. Frontiers in Oncology, 2023, 13: 1236268. Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9156282","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":625182021,"identity":"45c95c47-f42a-4399-8040-d8b9e91ecdbd","order_by":0,"name":"Chang Ge","email":"","orcid":"","institution":"The Second Affiliated Hospital of Xinjiang Medical University","correspondingAuthor":false,"prefix":"","firstName":"Chang","middleName":"","lastName":"Ge","suffix":""},{"id":625182022,"identity":"c82f7963-0172-4c27-9308-b82cf7e9ce17","order_by":1,"name":"Wenjie Zhang","email":"","orcid":"","institution":"The Second Affiliated Hospital of Xinjiang Medical University","correspondingAuthor":false,"prefix":"","firstName":"Wenjie","middleName":"","lastName":"Zhang","suffix":""},{"id":625182023,"identity":"5d3e826d-b15f-4849-9704-c38e96d5dcda","order_by":2,"name":"Pengcheng Zhang","email":"","orcid":"","institution":"The Second Affiliated Hospital of Xinjiang Medical University","correspondingAuthor":false,"prefix":"","firstName":"Pengcheng","middleName":"","lastName":"Zhang","suffix":""},{"id":625182024,"identity":"dbc9eaae-afbb-4ca9-9478-1671cdc95510","order_by":3,"name":"Jingxuan Xu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA/klEQVRIie3PMUsDMRTA8RwH5/JK1gvod3jgIMKhX+WFg0w3ON4YKZzLoesVQb+CLs4pGVyCroE6XL9Bi0uVDh7t5NC0o2D+y4PwfjzCWCz2NyOzncl1T3UBnOuDSWqxd+pYdObga5kS88YWqCm8h57m9urbyrv7FnNy74DMJItltZuIjshObpXsPhwi1TM4S3UqJi+7Cc8HMmoLqX2FRG4G59pk6ShAsi3J5eNAjGzeAA2FyeYKrAr55JXUsjH7iWj74YpWp8++tIxcCaKbjoN/wdeq/IS1PXnw8uZrVV9ccj6eLpYBwhgQS5pfL4kO7Q8dGcbWe3ZisVjsf/cDsFde1l79/ScAAAAASUVORK5CYII=","orcid":"","institution":"The Second Affiliated Hospital of Xinjiang Medical University","correspondingAuthor":true,"prefix":"","firstName":"Jingxuan","middleName":"","lastName":"Xu","suffix":""}],"badges":[],"createdAt":"2026-03-18 07:55:43","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9156282/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9156282/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":107450163,"identity":"60546f10-17c5-405f-8ac8-84c721b3da47","added_by":"auto","created_at":"2026-04-21 15:11:37","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":133727,"visible":true,"origin":"","legend":"\u003cp\u003eTechnical roadmap of this study\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-9156282/v1/6e898246916e8046892c2327.png"},{"id":107450203,"identity":"feb25b4f-b9c3-4f25-a3d9-5b840434dc13","added_by":"auto","created_at":"2026-04-21 15:11:45","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":210689,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSingle-cell RNA sequencing analysis of tumor and peritumoral normal tissues from 3 HGG patients\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) UMAP projection showing the distribution of all single cells, color-coded by biological replicate samples to clearly distinguish cell populations between glioma and normal brain tissue samples. (B) Cell population distribution colored by sample group, intuitively presenting the differences in cell population distribution between HGG and normal tissues. (C) Transcriptomic profile colored by cell cycle phase, showing the distribution characteristics of cells in G1, G2M, and S phases. (D) Unsupervised clustering results by Seurat, labeled with 17 cell clusters (0–16), providing a basis for subsequent cell type annotation. (E) Expression characteristics of key marker genes in each cell type; the size of the dots represents the proportion of cells expressing the gene, and the color gradient represents the average expression level, assisting in the accurate annotation of cell types. (F) Cell population distribution colored by custom cell type classification, clearly annotating major cell types including Astrocytes, Endothelial cells, Microglia, Neurons, NK cells, Oligodendrocytes, Pericytes, Proliferating cells, and T cells. (G) Percentage composition of each cell type in each biological replicate sample, intuitively reflecting the differences in cell population composition among different samples. (H) Comparison of the proportion of each cell type between glioma and normal tissues; the y-axis represents the cell proportion, and box plots show the distribution range, median, and quartiles of each group, revealing the cellular composition characteristics and intergroup differences of the tumor microenvironment.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-9156282/v1/318aa7fa327564ab7033c607.png"},{"id":107450222,"identity":"6c291458-1e43-418f-a713-26e258390b74","added_by":"auto","created_at":"2026-04-21 15:11:48","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":264181,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSpatial transcriptomics reveals the cell-cell communication network among Astrocytes, Microglia and Pericytes in the glioblastoma microenvironment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Cell density heatmap (nCount_Spatial) of spatial transcriptome sections; the color gradient shows the distribution of transcript counts in different regions, reflecting the cell density and transcriptome capture efficiency of the tissue sections. (B) Spatial distribution map of cell types in spatial transcriptome sections; Astrocytes (red), Microglia (green), and Pericytes (blue) were annotated via deconvolution analysis, intuitively presenting the spatial distribution characteristics of the three cell types in the tumor microenvironment. (C) Heatmap of cell-cell communication analysis; the left panel shows the number of interactions among Astrocytes, Microglia, and Pericytes, and the right panel shows the interaction strength; the color gradients correspond to the levels of interaction number and strength, respectively, and marginal bar charts show the total interaction number and strength of each cell type. (D) Cell-cell communication network diagram; nodes represent cell types, and the thickness of edges represents the number of interactions, clearly presenting the communication connection pattern among the three cell types. (E) Cell-cell communication network diagram; nodes represent cell types, and the thickness of edges represents the interaction strength, intuitively showing the differences in the strength of communication between different cell pairs.\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-9156282/v1/dfa65f8b1fd8e26d742a1b9e.png"},{"id":107450164,"identity":"2ae906bf-dab7-46a7-a849-4333f9858793","added_by":"auto","created_at":"2026-04-21 15:11:37","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":257309,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEnrichment analysis of Astrocyte differentially expressed genes\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Scatter plot of KEGG pathway enrichment analysis for upregulated genes; the y-axis represents the names of enriched KEGG pathways, the x-axis represents the -log10-transformed adjusted \u003cem\u003eP\u003c/em\u003e-value, the size of the dots represents the number of enriched genes, and the color gradient corresponds to the level of the adjusted \u003cem\u003eP\u003c/em\u003e-value. (B) Scatter plot of KEGG pathway enrichment analysis for downregulated genes. (C) Scatter plot of GO enrichment analysis for upregulated genes, displayed from three dimensions: biological process, molecular function, and cellular component; the y-axis represents the names of GO terms in each dimension, the x-axis represents the -log10-transformed adjusted \u003cem\u003eP\u003c/em\u003e-value, the size of the dots represents the number of enriched genes, and the color gradient corresponds to the level of the adjusted \u003cem\u003eP\u003c/em\u003e-value. (D) Scatter plot of GO enrichment analysis for downregulated genes.\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-9156282/v1/c87e69d543c2002ef9562f18.png"},{"id":107450218,"identity":"1ab7d0ab-71cf-4977-b062-f1ac11471786","added_by":"auto","created_at":"2026-04-21 15:11:48","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":307024,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eScreening of core gene co-expression modules via WGCNA\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) WGCNA soft threshold screening plot: the left panel shows the curve of scale-free topology model fit index with the change of soft threshold; the x-axis represents the soft threshold value, the y-axis represents the model fit index R², and the red horizontal line marks the threshold standard of R²=0.8. The right panel shows the curve of mean connectivity with the change of soft threshold; the x-axis represents the soft threshold value, the y-axis represents the mean connectivity value, used to determine the optimal soft threshold for constructing a scale-free network. (B) TOM heatmap and gene clustering tree: the left panel is the gene hierarchical clustering tree based on TOM, showing the clustering relationship of genes; the right panel is the TOM matrix visualization result, the color gradient reflects the level of topological overlap between genes, with red representing high topological overlap and yellow representing moderate topological overlap, providing a basis for module division. (C) Gene module clustering dendrogram: the y-axis represents the clustering height, the x-axis represents gene samples; genes with similar expression patterns were clustered into different modules via hierarchical clustering, and the color bands at the bottom indicate the module to which each gene belongs. (D) Gene co-expression adjacency heatmap and hierarchical clustering tree: the upper panel is the hierarchical clustering tree based on gene expression similarity, showing the genetic relationship among genes; the lower panel is the module eigengene adjacency heatmap, the color gradient represents the level of adjacency coefficient among genes, with red indicating high adjacency and blue indicating low adjacency.\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-9156282/v1/478706c1931a648371de93b7.png"},{"id":107450207,"identity":"145ae6f9-8896-4076-99fc-24ef3ad137d4","added_by":"auto","created_at":"2026-04-21 15:11:46","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":217348,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eModule-trait relationship heatmap of gene modules and glioma clinical phenotypes via WGCNA\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eModule-trait relationship heatmap showing the correlation between gene co-expression modules identified by WGCNA and glioma pathological phenotypes; the y-axis represents the names of each gene module, the x-axis represents pathological phenotype traits, and the color gradient of the heatmap represents the value of the correlation coefficient.\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-9156282/v1/7befe7854443e35d5e660aa4.png"},{"id":107450165,"identity":"b2f59a97-4696-4d57-ae00-d59f5356e9f6","added_by":"auto","created_at":"2026-04-21 15:11:37","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":136492,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eScreening of highly prognosis-related genes for HGG patients via machine learning\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Venn diagram of core gene intersection analysis (overlap): the intersection distribution of module genes identified by WGCNA in the TCGA database, upregulated genes (SC_Astrocyte.up) and downregulated genes (SC_Astrocyte.down) of Astrocyte subsets. (B) LASSO Cox regression coefficient profile plot: the x-axis represents the logλ-transformed penalty coefficient, the y-axis represents the gene regression coefficient; the red line represents the change trend of each gene's regression coefficient with the increase of the penalty coefficient, and the color gradient bar chart shows the gene count corresponding to each λ value. (C) Forest plot of hazard ratio (HR) from univariate Cox regression analysis of core genes: the x-axis represents the HR and 95% confidence interval (CI), the y-axis represents the names of core genes; the black horizontal line for each gene represents the 95% CI, the black square represents the HR value, and the \u003cem\u003eP\u003c/em\u003e-value is labeled on the right. (D) Kaplan-Meier survival curve of core gene risk scores (KM Plot by Median Risk Group): the x-axis represents survival time (days), the y-axis represents survival probability; the red curve represents the high-risk group, and the blue curve represents the low-risk group (divided by the median risk value).\u003c/p\u003e","description":"","filename":"7.png","url":"https://assets-eu.researchsquare.com/files/rs-9156282/v1/0f8a3498c936a074ccc9c7d5.png"},{"id":107450166,"identity":"b4e5e012-7b28-4355-9d96-6057ddbf7fca","added_by":"auto","created_at":"2026-04-21 15:11:37","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":225065,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCIBERSORT analysis of the relationship between key genes and the immune microenvironment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003e(A) Correlation analysis plot of NDUFB2 gene expression and immune cell infiltration in glioma. (B) Correlation analysis plot of HSP90B1 gene expression and immune cell infiltration in glioma. (C) Correlation analysis plot of LITAF gene expression and immune cell infiltration in glioma.\u003c/p\u003e","description":"","filename":"8.png","url":"https://assets-eu.researchsquare.com/files/rs-9156282/v1/e329aae64e495c97cfd97fc1.png"},{"id":108807296,"identity":"0c4c8241-62c0-4e87-be69-29bddbabe8db","added_by":"auto","created_at":"2026-05-08 15:30:21","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1754488,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9156282/v1/9bb73a61-ff7e-4a16-a139-bbc96f8542c0.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Single-Cell and Spatial Transcriptomics Identify Astrocyte Regulation of the Microenvironment and Prognosis in High-Grade Gliomas","fulltext":[{"header":"Background","content":"\u003cp\u003eGlioma is the most common primary malignant tumor of the central nervous system (CNS), characterized by rapid growth, strong invasiveness, high recurrence rate, and poor prognosis\u003csup\u003e[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]\u003c/sup\u003e. According to the World Health Organization (WHO) classification criteria, gliomas are graded from grade Ⅰ to Ⅳ based on malignancy, with WHO grades Ⅲ and Ⅳ categorized as high-grade gliomas (HGG)\u003csup\u003e[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]\u003c/sup\u003e. Currently, the standard treatment regimen for HGG remains the Stupp protocol, which consists of surgical resection combined with radiotherapy and temozolomide-based chemotherapy. However, even with standardized treatment, the overall prognosis of HGG patients remains unsatisfactory, with a limited median survival time and persistently high recurrence rate\u003csup\u003e[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]\u003c/sup\u003e. Therefore, in-depth exploration of the molecular mechanisms underlying HGG initiation and progression, as well as identification of potential key regulatory targets, are crucial for developing novel targeted therapeutic strategies and improving patient outcomes.\u003c/p\u003e \u003cp\u003eAstrocytes are the most abundant glial cell type in the CNS, playing pivotal roles in nervous system development and homeostasis maintenance. Their functions include maintaining the blood-brain barrier (BBB), storing and distributing energy substrates to support neuronal activity, and promoting neural cell survival and synaptogenesis\u003csup\u003e[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]\u003c/sup\u003e. Additionally, astrocytes regulate inflammatory responses and neurodegenerative processes in the CNS through multiple mechanisms, such as releasing neurotoxic factors, modulating microglial activity, and participating in the recruitment and infiltration of inflammatory cells\u003csup\u003e[\u003cspan additionalcitationids=\"CR9 CR10\" citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e]\u003c/sup\u003e. However, systematic studies on the key genes in astrocytes that are closely associated with HGG patient prognosis and their specific roles in reshaping the tumor immune microenvironment are still lacking.\u003c/p\u003e \u003cp\u003eIn this study, we integrated single-cell RNA sequencing (scRNA-seq) technology with transcriptomic data from public databases to systematically decipher the gene expression profiles of astrocytes in HGG tissues. Firstly, scRNA-seq was performed on HGG tumor tissues to identify differentially expressed genes (DEGs) in astrocytes. Subsequently, public transcriptomic datasets were integrated, and weighted gene co-expression network analysis (WGCNA) was employed to screen key modules closely related to HGG initiation and progression. On this basis, combined with clinical prognosis information of patients from public databases, we further identified core astrocyte-derived genes significantly associated with HGG patient prognosis. Finally, immune infiltration analysis was conducted to explore the potential regulatory roles of these key genes in the HGG tumor immune microenvironment, aiming to reveal the molecular mechanisms by which astrocytes contribute to HGG progression. This study provides a theoretical basis for a deeper understanding of the composition of the HGG immune microenvironment, exploration of novel targeted therapeutic strategies, and identification of prognostic biomarkers.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e"},{"header":"Methods","content":"\n\u003ch3\u003e1 Human Tumor Specimen Collection\u003c/h3\u003e\n\u003cp\u003eGlioma samples and peritumoral normal tissues were collected from patients who underwent surgical treatment at the Second Affiliated Hospital of Xinjiang Medical University. All specimens were collected in accordance with national standards. This study was approved by the Ethics Committee of the Second Affiliated Hospital of Xinjiang Medical University, and written informed consent was obtained from all patients whose specimens were included in the study.\u003c/p\u003e\n\u003ch3\u003e2 Single-Cell RNA Sequencing (scRNA-seq) Data Collection\u003c/h3\u003e\n\u003cp\u003eTumor samples were thoroughly rinsed with phosphate-buffered saline (PBS) to remove blood, then dissected with a surgical blade and washed in pre-chilled RPMI 1640 medium. Tissue dissociation was performed using a collagenase cocktail, followed by red blood cell lysis with a red blood cell lysis buffer (BL503A, Biosharp, China). The cells were collected by centrifugation, and cell count and viability were assessed using a fluorescent cell analyzer (Countstar\u0026reg; Rigel S2, ALIT Life Science, China), after which dead cell removal (Miltenyi, 130-090-101, Germany) was performed as needed. Finally, fresh cells were washed twice with RPMI 1640 medium and resuspended in 1\u0026times;PBS containing 0.04% bovine serum albumin (BSA).\u003c/p\u003e \u003cp\u003eLibraries for single-cell/nucleus RNA sequencing were constructed from the successfully dissociated single-cell/nucleus suspension using the DNBelab C Series High-Throughput scRNA Library Prep Kit V3.0. First, water-in-oil emulsions were generated with the DNBelab C-TaiM 4 system, and reverse transcription was performed on qualified emulsions. Subsequently, demulsification was conducted, and the cDNA products were purified. The target cDNA was subjected to fragmentation, end repair, adaptor ligation, amplification and purification to generate the BGI single-cell/nucleus transcriptome library. The library concentration was quantified using a Qubit 4.0 Fluorometer, and the libraries were stored at -80\u0026deg;C. Paired-end sequencing was performed on the constructed libraries using the DNBSEQ-T7 sequencing platform to obtain high-quality transcriptomic data.\u003c/p\u003e\n\u003ch3\u003e3 Quality Control of scRNA-seq Data\u003c/h3\u003e\n\u003cp\u003escRNA-seq data were processed and analyzed following standard protocols. Raw sequencing reads were aligned to the human reference genome GRCh38 using Cell Ranger (v5.0, 10x Genomics), and gene expression was quantified using unique molecular identifiers (UMIs). For data generated by the BGI sequencing platform, preprocessing was performed using the dnbc4tools software with similar parameters. Downstream analyses were conducted on the filtered gene expression matrix using the Seurat software (v4.0.0).\u003c/p\u003e \u003cp\u003eLow-quality cells were filtered out based on the following criteria: number of detected genes\u0026thinsp;\u0026lt;\u0026thinsp;500, mitochondrial gene transcript proportion\u0026thinsp;\u0026gt;\u0026thinsp;20%, or ribosomal gene transcript proportion\u0026thinsp;\u0026gt;\u0026thinsp;50%. Meanwhile, genes detected in fewer than 3 cells were excluded. Potential doublets were identified and removed using the DoubletFinder software (v2.0.3), with the expected doublet rate set to 7.5% based on the Poisson distribution.\u003c/p\u003e\n\u003ch3\u003e4 Normalization and Dimensionality Reduction\u003c/h3\u003e\n\u003cp\u003eUMI counts were normalized using the LogNormalize method (with a scale factor of 10,000) and log-transformed. The top 2,000 highly variable genes were selected for principal component analysis (PCA). Batch effects between different samples or experimental conditions were corrected using the Harmony software (v1.0). The number of principal components (PCs) used for subsequent analyses was determined via an elbow plot, and the first 30 PCs were selected in this study.\u003c/p\u003e \u003cp\u003eCell clustering was performed using the graph-based Louvain algorithm with a resolution parameter of 0.4 (adjustable from 0.1 to 1 according to dataset size). Marker genes for each cell cluster were identified using the FindAllMarkers function, with screening thresholds set as |avg_log2FC| \u0026ge; 0.5 and adjusted P-value\u0026thinsp;\u0026le;\u0026thinsp;0.05.\u003c/p\u003e\n\u003ch3\u003e5 Cell Feature Selection and Annotation\u003c/h3\u003e\n\u003cp\u003eTo distinguish malignant cells from normal cells, genome-wide copy number variations (CNVs) were inferred from gene expression data using the CopyKat software (v0.1.0). Aneuploid cells were classified as tumor cells, while diploid cells were considered normal stromal or immune cells, and this classification was validated by the expression of known lineage-specific marker genes. Cell type annotation was performed by combining automated tools with manual verification based on classic marker genes.\u003c/p\u003e\n\u003ch3\u003e6 Spatial Transcriptome and Cell-Cell Communication Analysis\u003c/h3\u003e\n\u003cp\u003ePublic spatial transcriptome data (GSE253080) were used for deconvolution analysis with scRNA-seq data as the reference on spatial transcriptome spots. The cell type proportion of each spot in the spatial transcriptome was obtained after deconvolution, and the cell type with the highest proportion in each spot was defined as the dominant cell type of that spot. Based on the above cell type annotation results, cell-cell communication analysis was performed on the spatial transcriptome data.\u003c/p\u003e\n\u003ch3\u003e7 Differential Expression Analysis\u003c/h3\u003e\n\u003cp\u003eTo identify differentially expressed genes (DEGs) in astrocytes, the DESeq2 algorithm was applied to analyze the scRNA-seq transcriptomic data. First, preprocessing was performed on the raw gene expression count matrix using the built-in normalization method of DESeq2 following its standard analysis pipeline to eliminate systematic biases caused by factors such as sequencing depth variation. DEGs were then screened via statistical tests with the criteria of adjusted P-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 and |log₂FC| \u0026gt; 1.\u003c/p\u003e\n\u003ch3\u003e8 Enrichment Analysis\u003c/h3\u003e\n\u003cp\u003eTo further elucidate the biological functions of DEGs, Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analyses were performed on the screened DEGs.\u003c/p\u003e \u003cp\u003eGO analysis annotated gene functions from three dimensions: Biological Process (BP), Cellular Component (CC) and Molecular Function (MF). DEGs were mapped to each term in the GO database, and the number of genes in each term was counted. Using genome-wide GO annotations as the background, hypergeometric distribution tests were performed to identify significantly enriched GO terms for DEGs.\u003c/p\u003e \u003cp\u003eKEGG pathway enrichment analysis was used to characterize the major metabolic and signal transduction pathways involved in DEGs [12\u0026ndash;14]. DEGs were mapped to the KEGG database, and the number of genes in each pathway was counted. Significantly enriched KEGG pathways were screened using hypergeometric distribution tests with genome-wide KEGG annotations as the background.\u003c/p\u003e \u003cp\u003eFor data visualization, bubble plots were generated using the ggplot2 package in R to intuitively display the distribution of significantly enriched GO terms and KEGG pathways.\u003c/p\u003e\n\u003ch3\u003e9 Weighted Gene Co-Expression Network Analysis (WGCNA)\u003c/h3\u003e\n\u003cp\u003eOfficially batch-corrected TCGA pan-cancer transcriptomic data were downloaded from the UCSC Toil RNA-seq Recompute database (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://xenabrowser.net/datapages/\u003c/span\u003e\u003cspan address=\"https://xenabrowser.net/datapages/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). Transcriptomic data of low-grade glioma (LGG) and glioblastoma (GBM) \u0026mdash; the main subtype of high-grade glioma (HGG) \u0026mdash; were extracted, including 5 normal control samples, 523 LGG samples and 166 GBM samples. The expression matrix of protein-coding genes was further filtered for subsequent WGCNA analysis.\u003c/p\u003e\n\u003ch3\u003e10 Gene Screening via Machine Learning\u003c/h3\u003e\n\u003cp\u003eThe optimal regularization parameter lambda was selected using 10-fold cross-validation, and the conservative lambda.1se was adopted to reduce the risk of overfitting. Genes with non-zero coefficients were extracted as key features. A multi-gene Cox proportional hazards model was then constructed to validate the prognostic value of the selected genes. Risk scores were calculated, and Kaplan-Meier (KM) curves were plotted for risk groups for visualization.\u003c/p\u003e\n\u003ch3\u003e11 Receiver Operating Characteristic (ROC) Curve Analysis\u003c/h3\u003e\n\u003cp\u003eTo evaluate the prognostic predictive value of hub genes for HGG patients, ROC curve analysis was performed based on hub gene expression levels and clinical prognostic information of TCGA-GBM patients. ROC curves were plotted using the pROC package in R, and the area under the curve (AUC) was calculated to quantify the ability of hub genes to distinguish the prognostic status of HGG patients\u003csup\u003e[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]\u003c/sup\u003e.\u003c/p\u003e\n\u003ch3\u003e12 CIBERSORT Analysis\u003c/h3\u003e\n\u003cp\u003eTo estimate the relative abundances of tumor-infiltrating immune cells in tumor masses, the reference set of signature gene profiles for 22 immune cell subtypes provided by the online analysis platform CIBERSORT (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cibersort.stanford.edu\u003c/span\u003e\u003cspan address=\"https://cibersort.stanford.edu\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) was used. Spearman correlation analysis was performed to identify immune cell types significantly correlated with key hub genes.\u003c/p\u003e"},{"header":"Results","content":"\n\u003ch3\u003e1 ScRNA-seq Profiling of HGG Tumor Tissues\u003c/h3\u003e\n\u003cp\u003escRNA-seq was performed on tumor tissue samples from 3 patients with WHO grade Ⅲ HGG and their matched peritumoral normal tissues, all of which were pathologically confirmed as HGG after surgery (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eA-B). A total of 46,181 high-quality cells were identified and clustered into 17 cell clusters via manual verification of classic marker genes combined with the automated annotation tool SingleR (v1.8.30) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eC-D). The FindAllMarkers() function was used to compare gene expression differences between each cell cluster and all other clusters for screening potential specific marker genes. Screening criteria comprehensively considered the average log2 fold change (avg_log2FC), adjusted \u003cem\u003eP\u003c/em\u003e-value (p.adj), and difference in expression percentage (diff.pct); the top 3 genes of each cluster were extracted as representative markers in the order of decreasing avg_log2FC and increasing \u003cem\u003eP\u003c/em\u003e-value (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eE). Based on the expression distribution of known classic marker genes, 16 cell clusters were annotated into 9 major cell types: Oligodendrocytes, T cells, Microglia, natural killer (NK) cells, Neurons, Astrocytes, Proliferating cells, Pericytes, and Endothelial cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eF). Compared with peritumoral normal tissues, the proportions of Astrocytes, Microglia, and tumor cells were significantly increased in HGG tissues (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e1\u003c/span\u003eG-H).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e(A) UMAP projection showing the distribution of all single cells, color-coded by biological replicate samples to clearly distinguish cell populations between glioma and normal brain tissue samples. (B) Cell population distribution colored by sample group, intuitively presenting the differences in cell population distribution between HGG and normal tissues. (C) Transcriptomic profile colored by cell cycle phase, showing the distribution characteristics of cells in G1, G2M, and S phases. (D) Unsupervised clustering results by Seurat, labeled with 17 cell clusters (0\u0026ndash;16), providing a basis for subsequent cell type annotation. (E) Expression characteristics of key marker genes in each cell type; the size of the dots represents the proportion of cells expressing the gene, and the color gradient represents the average expression level, assisting in the accurate annotation of cell types. (F) Cell population distribution colored by custom cell type classification, clearly annotating major cell types including Astrocytes, Endothelial cells, Microglia, Neurons, NK cells, Oligodendrocytes, Pericytes, Proliferating cells, and T cells. (G) Percentage composition of each cell type in each biological replicate sample, intuitively reflecting the differences in cell population composition among different samples. (H) Comparison of the proportion of each cell type between glioma and normal tissues; the y-axis represents the cell proportion, and box plots show the distribution range, median, and quartiles of each group, revealing the cellular composition characteristics and intergroup differences of the tumor microenvironment.\u003c/p\u003e\n\u003ch3\u003e2 Spatial Transcriptomics Deciphers the HGG Microenvironment: Astrocytes as the Dominant Cell Population\u003c/h3\u003e\n\u003cp\u003eWe further integrated the above scRNA-seq data with the public spatial transcriptome dataset GSE253080 to perform deconvolution analysis. The cell type proportion of each spot in the spatial transcriptome was obtained after deconvolution, and the cell type with the highest proportion in each spot was defined as the dominant cell type of that spot (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA-B). The results showed that Astrocytes, Microglia, and Proliferating cells were the top three cell types in terms of proportion. Based on these three core cell types, we further performed cell-cell communication analysis and found that there was intensive signal crosstalk between Astrocytes and Microglia (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC-E).\u003c/p\u003e \u003cp\u003eIn summary, Astrocytes serve as the major cellular component in HGG and can exert a crucial regulatory effect on the tumor immune microenvironment by targeting and modulating Microglia.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e(A) Cell density heatmap (nCount_Spatial) of spatial transcriptome sections; the color gradient shows the distribution of transcript counts in different regions, reflecting the cell density and transcriptome capture efficiency of the tissue sections. (B) Spatial distribution map of cell types in spatial transcriptome sections; Astrocytes (red), Microglia (green), and Pericytes (blue) were annotated via deconvolution analysis, intuitively presenting the spatial distribution characteristics of the three cell types in the tumor microenvironment. (C) Heatmap of cell-cell communication analysis; the left panel shows the number of interactions among Astrocytes, Microglia, and Pericytes, and the right panel shows the interaction strength; the color gradients correspond to the levels of interaction number and strength, respectively, and marginal bar charts show the total interaction number and strength of each cell type. (D) Cell-cell communication network diagram; nodes represent cell types, and the thickness of edges represents the number of interactions, clearly presenting the communication connection pattern among the three cell types. (E) Cell-cell communication network diagram; nodes represent cell types, and the thickness of edges represents the interaction strength, intuitively showing the differences in the strength of communication between different cell pairs.\u003c/p\u003e\n\u003ch3\u003e3 DEGs of Astrocytes and Their Enrichment Analysis in the HGG Microenvironment\u003c/h3\u003e\n\u003cp\u003eKEGG pathway enrichment analysis of Astrocyte DEGs showed that the upregulated genes were mainly enriched in ribosome, antigen processing and presentation, and oxidative phosphorylation as the core pathways; additionally, pathways associated with neurodegenerative diseases such as Huntington's disease, Parkinson's disease, and Alzheimer's disease were also significantly enriched, along with metabolic pathways including non-alcoholic fatty liver disease and mineral absorption (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA). The downregulated genes were primarily enriched in axon guidance, tight junction and other pathways, and cancer-related pathways as well as infectious disease-related pathways were also identified (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB).\u003c/p\u003e \u003cp\u003eGO enrichment analysis was performed on the DEGs of Astrocyte subsets separately. The results showed that the upregulated genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC) were concentrated in cellular metabolic and biosynthetic processes such as cytoplasmic translation and ribosomal small subunit biogenesis in biological processes; enriched in binding and catalytic activities such as actin binding and calcium ion binding in molecular functions; and localized to intracellular organelles such as mitochondrial matrix and ribosomal subunits in cellular components. The activation of these metabolism- and ribosome-related functions was highly consistent with the pathological characteristics of energy metabolism disorder and abnormal protein synthesis in neurodegenerative diseases. The downregulated genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD) focused on protein assembly and cell adhesion processes such as NADH dehydrogenase complex assembly and cadherin binding in biological processes; enriched in extracellular matrix (ECM)-related functions such as ECM structural constituent and proteoglycan binding in molecular functions; and concentrated in extracellular and membrane-associated structures such as ECM and basement membrane in cellular components. The downregulation of such cell adhesion and ECM interaction functions may be involved in the pathological processes of abnormal neuron-glia connection and synaptic structure damage in neurodegenerative diseases.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e(A) Scatter plot of KEGG pathway enrichment analysis for upregulated genes; the y-axis represents the names of enriched KEGG pathways, the x-axis represents the -log10-transformed adjusted \u003cem\u003eP\u003c/em\u003e-value, the size of the dots represents the number of enriched genes, and the color gradient corresponds to the level of the adjusted \u003cem\u003eP\u003c/em\u003e-value. (B) Scatter plot of KEGG pathway enrichment analysis for downregulated genes. (C) Scatter plot of GO enrichment analysis for upregulated genes, displayed from three dimensions: biological process, molecular function, and cellular component; the y-axis represents the names of GO terms in each dimension, the x-axis represents the -log10-transformed adjusted \u003cem\u003eP\u003c/em\u003e-value, the size of the dots represents the number of enriched genes, and the color gradient corresponds to the level of the adjusted \u003cem\u003eP\u003c/em\u003e-value. (D) Scatter plot of GO enrichment analysis for downregulated genes.\u003c/p\u003e\n\u003ch3\u003e4 Screening of Highly Correlated Modules and Genes via WGCNA\u003c/h3\u003e\n\u003cp\u003eTo screen highly correlated transcriptomic modules, we further utilized TCGA-LGG and TCGA-GBM transcriptomic data, including 5 normal control samples, 523 LGG samples, and 166 GBM samples, for WGCNA. When the soft threshold was set to 6, the scale-free topology model fit index reached above 0.8, and the mean connectivity showed a steady decline with the increase of the soft threshold, indicating that this parameter selection could effectively construct a scale-free co-expression network (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). Hierarchical clustering was performed on the topological overlap matrix (TOM) to cluster highly correlated genes into different co-expression modules. The hierarchical clustering dendrogram showed that genome-wide genes were successfully divided into multiple modules with specific expression patterns, labeled with different color bands; a total of 27 gene modules were screened out (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB-C). Clustering analysis of module eigengenes further revealed the expression correlation among each module (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD). Finally, as shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e, two modules (pink and red) were identified with a significant high positive correlation with GBM, containing 709 and 1,085 genes, respectively.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e(A) WGCNA soft threshold screening plot: the left panel shows the curve of scale-free topology model fit index with the change of soft threshold; the x-axis represents the soft threshold value, the y-axis represents the model fit index R\u0026sup2;, and the red horizontal line marks the threshold standard of R\u0026sup2;=0.8. The right panel shows the curve of mean connectivity with the change of soft threshold; the x-axis represents the soft threshold value, the y-axis represents the mean connectivity value, used to determine the optimal soft threshold for constructing a scale-free network. (B) TOM heatmap and gene clustering tree: the left panel is the gene hierarchical clustering tree based on TOM, showing the clustering relationship of genes; the right panel is the TOM matrix visualization result, the color gradient reflects the level of topological overlap between genes, with red representing high topological overlap and yellow representing moderate topological overlap, providing a basis for module division. (C) Gene module clustering dendrogram: the y-axis represents the clustering height, the x-axis represents gene samples; genes with similar expression patterns were clustered into different modules via hierarchical clustering, and the color bands at the bottom indicate the module to which each gene belongs. (D) Gene co-expression adjacency heatmap and hierarchical clustering tree: the upper panel is the hierarchical clustering tree based on gene expression similarity, showing the genetic relationship among genes; the lower panel is the module eigengene adjacency heatmap, the color gradient represents the level of adjacency coefficient among genes, with red indicating high adjacency and blue indicating low adjacency.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eModule-trait relationship heatmap showing the correlation between gene co-expression modules identified by WGCNA and glioma pathological phenotypes; the y-axis represents the names of each gene module, the x-axis represents pathological phenotype traits, and the color gradient of the heatmap represents the value of the correlation coefficient.\u003c/p\u003e\n\u003ch3\u003e5 Screening of Astrocyte-Derived Prognostic Genes for HGG Patients via Machine Learning\u003c/h3\u003e\n\u003cp\u003eTo further explore the Astrocyte DEGs associated with the prognosis of HGG patients, Venn diagram intersection analysis was performed between the module genes screened by WGCNA and the upregulated (SC_Astrocyte.up) and downregulated (SC_Astrocyte.down) genes of Astrocyte subsets (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eA), resulting in a total of 73 DEGs identified. These genes were further subjected to machine learning combined with survival information (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eB). The results showed that the gene regression coefficients gradually shrank with the increase of the penalty coefficient λ; when λ was set to the optimal value (logλ\u0026asymp;-3) (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eB), a total of 5 core genes with prognostic value (LTF, NOX4, MDK, HSP90B1, and HSPB1) were finally screened out. Univariate Cox regression analysis showed that LTF, NOX4, and HSP90B1 had a statistically significant risk association with HGG prognosis (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eC). Based on these 3 genes, glioma patients were divided into high-risk and low-risk groups by the median risk score. Kaplan-Meier survival analysis showed that the overall survival rate of patients in the high-risk group was significantly lower than that in the low-risk group (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001) (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eD), suggesting that this core gene panel can effectively predict the prognostic outcome of glioma patients and provide a potential molecular marker panel for glioma prognosis assessment.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e(A) Venn diagram of core gene intersection analysis (overlap): the intersection distribution of module genes identified by WGCNA in the TCGA database, upregulated genes (SC_Astrocyte.up) and downregulated genes (SC_Astrocyte.down) of Astrocyte subsets. (B) LASSO Cox regression coefficient profile plot: the x-axis represents the logλ-transformed penalty coefficient, the y-axis represents the gene regression coefficient; the red line represents the change trend of each gene's regression coefficient with the increase of the penalty coefficient, and the color gradient bar chart shows the gene count corresponding to each λ value. (C) Forest plot of hazard ratio (HR) from univariate Cox regression analysis of core genes: the x-axis represents the HR and 95% confidence interval (CI), the y-axis represents the names of core genes; the black horizontal line for each gene represents the 95% CI, the black square represents the HR value, and the \u003cem\u003eP\u003c/em\u003e-value is labeled on the right. (D) Kaplan-Meier survival curve of core gene risk scores (KM Plot by Median Risk Group): the x-axis represents survival time (days), the y-axis represents survival probability; the red curve represents the high-risk group, and the blue curve represents the low-risk group (divided by the median risk value).\u003c/p\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003e3.6 Immune Infiltration Analysis\u003c/h2\u003e \u003cp\u003eCorrelation analysis was performed to explore the association between the expression levels of three core genes (NDUFB2, HSP90B1, and LITAF) and immune cell infiltration in the glioma microenvironment. The expression of NDUFB2 was significantly positively correlated with the infiltration of multiple immune cells, among which the positive correlations with γδ T cells, eosinophils, and CD8\u0026thinsp;+\u0026thinsp;T cells were the most significant; it was significantly negatively correlated with M0 macrophages and resting NK cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eA). The expression of HSP90B1 was significantly positively correlated with CD4\u0026thinsp;+\u0026thinsp;memory activated T cells, regulatory T cells (Tregs), and M2 macrophages; it was significantly negatively correlated with eosinophils, activated NK cells, and M1 macrophages (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eB). The expression of LITAF was significantly positively correlated with CD4\u0026thinsp;+\u0026thinsp;activated T cells, resting dendritic cells, and monocytes; it was significantly negatively correlated with follicular helper T cells, M0 macrophages, and resting mast cells (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eC).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e(A) Correlation analysis plot of NDUFB2 gene expression and immune cell infiltration in glioma. (B) Correlation analysis plot of HSP90B1 gene expression and immune cell infiltration in glioma. (C) Correlation analysis plot of LITAF gene expression and immune cell infiltration in glioma.\u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eAs one of the most invasive malignant tumors of the CNS, HGG is characterized by a complex TME composition and intercellular interaction network, which are the core causes of treatment resistance and poor prognosis\u003csup\u003e[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e, \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]\u003c/sup\u003e. Astrocytes, the most abundant glial cell type in the CNS, not only participate in the maintenance of neural homeostasis but also act as key regulators of the TME through phenotypic transformation and functional remodeling during tumor progression\u003csup\u003e[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]\u003c/sup\u003e. In this study, we integrated scRNA-seq, spatial transcriptome sequencing, and public database analysis to systematically decipher the gene expression characteristics of Astrocytes in HGG, their spatial interaction patterns with other cells, screen prognosis-related core genes, and clarify their regulatory effects on the immune microenvironment. This study provides an important theoretical basis for an in-depth understanding of the pathogenesis of HGG and the development of novel targeted therapeutic strategies.\u003c/p\u003e \u003cp\u003eFirstly, scRNA-seq was used to clarify the cellular composition characteristics of HGG tissues, and we found that the proportions of Astrocytes, Microglia, and tumor cells were significantly increased in HGG compared with peritumoral normal tissues, suggesting that these three cell types may jointly participate in the remodeling of the TME. Spatial transcriptome deconvolution analysis further confirmed that Astrocytes are the dominant cell population in the HGG microenvironment and have intensive signal communication with Microglia, which is consistent with the conclusion of previous studies that \"Astrocyte-Microglia crosstalk regulates tumor progression\"\u003csup\u003e[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]\u003c/sup\u003e. Cell-cell communication network analysis showed that the number and strength of interactions between Astrocytes and Microglia were significantly higher than those of other cell combinations, implying that the two cell types may form a functional complex through paracrine signals and direct intercellular contact to jointly regulate the formation of the tumor immunosuppressive microenvironment, which provides a clear direction for the subsequent exploration of the molecular mechanism of intercellular interaction.\u003c/p\u003e \u003cp\u003eDEG enrichment analysis revealed the functional abnormal characteristics of Astrocytes in HGG: the upregulated genes were mainly enriched in metabolism-related pathways such as ribosome and oxidative phosphorylation, and neurodegenerative disease-related pathways such as Huntington's disease and Parkinson's disease were also significantly enriched, suggesting that the metabolic reprogramming of Astrocytes may share common molecular mechanisms with the pathological changes of neurodegenerative diseases\u003csup\u003e[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]\u003c/sup\u003e. The downregulated genes focused on processes such as ECM organization and cell adhesion, and the loss of these functions may damage the integrity of neuron-glia connections and the blood-brain barrier, facilitating the invasion and metastasis of tumor cells\u003csup\u003e[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]\u003c/sup\u003e. These results indicated that Astrocytes in HGG undergo bidirectional functional remodeling of \"metabolic activation and ECM interaction inhibition\", creating a favorable microenvironmental condition for tumor progression and providing a new molecular perspective to explain the high invasiveness of HGG.\u003c/p\u003e \u003cp\u003eTo screen core genes associated with HGG prognosis, WGCNA was performed on the TCGA database to identify pink and red modules with a high positive correlation with HGG. Further intersection analysis with Astrocyte DEGs and LASSO Cox regression screening finally identified 5 core genes (LTF, NOX4, MDK, HSP90B1, and HSPB1). Among them, LTF, NOX4, and HSP90B1 were confirmed as independent risk factors for the prognosis of HGG patients, and their high expression was significantly associated with short overall survival of patients. The risk score model based on these three genes could effectively distinguish high-risk and low-risk patients (\u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.001), suggesting that this gene panel has potential value as a molecular marker for HGG prognosis. Previous studies have shown that HSP90B1, a member of the heat shock protein family, is involved in the anti-apoptotic process of tumor cells by regulating protein folding and stability\u003csup\u003e[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]\u003c/sup\u003e; NOX4 promotes tumor angiogenesis and invasion by generating reactive oxygen species (ROS)\u003csup\u003e[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]\u003c/sup\u003e. This study for the first time confirmed that these two genes are closely associated with Astrocyte dysfunction and HGG prognosis, expanding the understanding of their functions in the tumor microenvironment.\u003c/p\u003e \u003cp\u003eImmune infiltration analysis further revealed the regulatory effects of core genes on the HGG immune microenvironment: high expression of NDUFB2 was positively correlated with the infiltration of anti-tumor immune cells such as CD8\u0026thinsp;+\u0026thinsp;T cells and γδ T cells, and negatively correlated with M0 macrophage infiltration, suggesting that it may improve patient prognosis by enhancing anti-tumor immune responses. HSP90B1 was positively correlated with the infiltration of immunosuppressive cells such as Tregs and M2 macrophages, and negatively correlated with M1 macrophage infiltration, which is consistent with the known mechanism that M2 macrophages and Tregs promote tumor immune escape by inhibiting the function of effector T cells[22,23]. High expression of LITAF was positively correlated with the infiltration of CD4\u0026thinsp;+\u0026thinsp;activated T cells and negatively correlated with follicular helper T cell infiltration, implying that it may participate in the balance of the immune microenvironment by regulating the polarization of T cell subsets. The above results indicated that Astrocyte-derived core genes can shape the HGG immune microenvironment by regulating the infiltration pattern of immune cells, and their abnormal expression may lead to an immunosuppression-dominant microenvironmental phenotype, thereby promoting tumor progression and treatment resistance.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eIn this study, we comprehensively investigated the biological roles of astrocytes in the progression and prognosis of HGG by integrating single-cell RNA sequencing, spatial transcriptome analysis, and multi-dimensional bioinformatics mining of public omics datasets. Our findings first characterized the cellular composition of the HGG microenvironment, confirming that astrocytes are a dominant cell population in HGG tissues with significantly increased proportions, and that astrocytes and microglia form a core cell communication pair with intensive intercellular crosstalk, which is a key structural basis for remodeling the HGG tumor microenvironment.\u003c/p\u003e \u003cp\u003eDifferential gene and enrichment analyses further revealed the distinct functional remodeling features of HGG-associated astrocytes: the activation of metabolic pathways such as ribosome and oxidative phosphorylation, together with the enrichment of neurodegenerative disease-related signaling, and the downregulation of extracellular matrix organization and cell adhesion-related functions. This bidirectional functional alteration not only reflects the functional abnormality of astrocytes in HGG but also creates a favorable microenvironmental condition for tumor cell invasion and immune suppression. Through WGCNA and machine learning strategies, we screened out astrocyte-derived core genes (LTF, NOX4, HSP90B1, MDK, HSPB1) associated with HGG prognosis, among which LTF, NOX4 and HSP90B1 were verified as independent prognostic risk factors for HGG patients. The risk score model constructed by these three genes can effectively distinguish the survival outcome of HGG patients, providing a novel potential molecular marker panel for clinical prognostic assessment of HGG.\u003c/p\u003e \u003cp\u003eImmune infiltration analysis further elucidated the regulatory mechanism of astrocyte-derived core genes on the HGG immune microenvironment: NDUFB2, HSP90B1 and LITAF exert distinct regulatory effects on the infiltration and polarization of tumor-infiltrating immune cells (including T cell subsets, macrophages, dendritic cells, etc.), and their abnormal expression drives the formation of an immunosuppression-dominant tumor microenvironment, thereby promoting HGG progression and treatment resistance. These results confirm that astrocyte-derived core genes are key regulators of the HGG immune microenvironment, and clarify the molecular link between astrocyte dysfunction and immune microenvironment remodeling in HGG.\u003c/p\u003e \u003cp\u003eOverall, our study identifies the critical regulatory role of astrocytes in HGG, reveals the functional characteristics of astrocytes in HGG and their molecular mechanism of regulating tumor immune microenvironment, and screens out prognostic core genes with potential clinical application value. These findings not only deepen the understanding of the molecular mechanism of HGG progression mediated by stromal cells such as astrocytes, but also provide new candidate targets for the development of astrocyte-targeted therapeutic strategies and the optimization of HGG prognostic evaluation systems. Limitations of this study include the small sample size of clinical specimens for single-cell sequencing and the lack of in vitro and in vivo experimental validation of the core genes' functions. Future research will expand the sample size, and verify the biological functions and regulatory mechanisms of the core genes through cell experiments and animal models, and further explore the specific molecular signaling pathways of astrocyte-microglia crosstalk in the HGG microenvironment, so as to provide more solid experimental basis for translating these findings into clinical practice.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003cdiv class=\"DefinitionList\"\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eHGG\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eHigh-grade glioma\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eLGG\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eLow-grade glioma\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eGBM\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eGlioblastoma\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eWHO\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eThe World Health Organization\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCNS\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003ecentral nervous system\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eBBB\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eblood-brain barrier\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003escRNA-seq\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003esingle-cell RNA sequencing\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eST-seq\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003espatial transcriptomic sequencing\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eGO\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eThree letter acronym\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eKEGG\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eKyoto Encyclopedia of Genes and Genomes\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eWCGNA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eweighted gene co-expression network analysis\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003ePCA\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eprincipal component analysis\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eDEG\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003edifferentially expressed genes\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eBP\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eBiological Process\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eCC\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eCellular Component\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv class=\"DefinitionListEntry\"\u003e \u003cdiv class=\"Term\"\u003eMF\u003c/div\u003e \u003cdiv class=\"Description\"\u003e \u003cp\u003eMolecular Function\u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003c/div\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was approved by the Ethics Committee of the Second Affiliated Hospital of Xinjiang Medical University (approval number: [2025022624]). All procedures performed in studies involving human participants were in accordance with the ethical standards of the national research committee and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Written informed consent was obtained from all individual participants included in the study.\u003c/p\u003e\n\u003cp\u003eClinical trial number: not applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003enot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAvailability of data and materials\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated and/or analysed during the current study are available in the Genome Sequence Archive (GSA) at the National Genomics Data Center (NGDC) repository,\u0026nbsp;\u003ca href=\"https://ngdc.cncb.ac.cn/omix/preview/RO7wSvN1\"\u003ehttps://ngdc.cncb.ac.cn/omix/preview/RO7wSvN1\u003c/a\u003e.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research was funded by Xinjiang Uygur Autonomous Region Collaborative Innovation Special Project, grant number 2022E02060 and Open Project of the State Key Laboratory of High Incidence Diseases Prevention and Treatment in Central Asia, grant number SKL-HIDCA-2022-NKX5.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors' contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eJ.X.X. proofread and wrote the manuscript. C.G. wrote the manuscript and prepared all figures. W.J.Z. and P.C.Z. collected and organized the data. All authors reviewed and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe would like to express our sincere gratitude to Doubao AI for its valuable contributions to the linguistic polishing and translation work of this manuscript, which has significantly improved the accuracy and fluency of the text.\u003c/p\u003e\n\u003cp\u003eAdditionally, we thank the participants who donated clinical specimens for this study and the medical staff of the Second Affiliated Hospital of Xinjiang Medical University for their support in sample collection and processing. This work was supported by the Xinjiang Uygur Autonomous Region Collaborative Innovation Special Project (Grant No. 2022E02060) and the Open Project of the State Key Laboratory of High Incidence Diseases Prevention and Treatment in Central Asia (Grant No. SKL-HIDCA-2022-NKX5).\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eOstrom Q T, Price M, Neff C, et al. CBTRUS Statistical Report: Primary Brain and Other Central Nervous System Tumors Diagnosed in the United States in 2015-2019[J]. Neuro-Oncology, 2022, 24(Suppl 5): v1-v95.\u003c/li\u003e\n \u003cli\u003eZhou Y, Xiao D, Jiang X, et al. EREG is the core onco-immunological biomarker of cuproptosis and mediates the cross-talk between VEGF and CD99 signaling in glioblastoma[J]. Journal of Translational Medicine, 2023, 21(1): 28.\u003c/li\u003e\n \u003cli\u003eLouis D N, Perry A, Wesseling P, et al. The 2021 WHO Classification of Tumors of the Central Nervous System: a summary[J]. NEURO-ONCOLOGY, 2021, 23(8): 1231-1251.\u003c/li\u003e\n \u003cli\u003eLiu Y, Wu Z, Feng Y, et al. Integration analysis of single-cell and spatial transcriptomics reveal the cellular heterogeneity landscape in glioblastoma and establish a polygenic risk model[J]. Frontiers in Oncology, 2023, 13: 1109037.\u003c/li\u003e\n \u003cli\u003eWu L, Wu W, Zhang J, et al. Natural Coevolution of Tumor and Immunoenvironment in Glioblastoma[J]. Cancer Discovery, 2022, 12(12): 2820-2837.\u003c/li\u003e\n \u003cli\u003eFerris H A, Perry R J, Moreira G V, et al. Loss of astrocyte cholesterol synthesis disrupts neuronal function and alters whole-body metabolism[J]. Proceedings of the National Academy of Sciences of the United States of America, 2017, 114(5): 1189-1194.\u003c/li\u003e\n \u003cli\u003eAllen N J, Eroglu C. Cell Biology of Astrocyte-Synapse Interactions[J]. Neuron, 2017, 96(3): 697-708.\u003c/li\u003e\n \u003cli\u003eLinnerbauer M, Wheeler M A, Quintana F J. Astrocyte Crosstalk in CNS Inflammation[J]. Neuron, 2020, 108(4): 608-622.\u003c/li\u003e\n \u003cli\u003eLiddelow S A, Barres B A. Reactive Astrocytes: Production, Function, and Therapeutic Potential[J]. Immunity, 2017, 46(6): 957-967.\u003c/li\u003e\n \u003cli\u003eMayo L, Trauger S A, Blain M, et al. Regulation of astrocyte activation by glycolipids drives chronic CNS inflammation[J]. Nature Medicine, 2014, 20(10): 1147-1156.\u003c/li\u003e\n \u003cli\u003eWheeler M A, Jaronen M, Covacu R, et al. Environmental Control of Astrocyte Pathogenic Activities in CNS Inflammation[J]. Cell, 2019, 176(3): 581-596.e18.\u003c/li\u003e\n \u003cli\u003eKanehisa M, Goto S. KEGG: kyoto encyclopedia of genes and genomes[J]. Nucleic Acids Research, 2000, 28(1): 27-30.\u003c/li\u003e\n \u003cli\u003eKanehisa M, Furumichi M, Sato Y, et al. KEGG: biological systems database as a model of the real world[J]. Nucleic Acids Research, 2025, 53(D1): D672-D677.\u003c/li\u003e\n \u003cli\u003eKanehisa M. Toward understanding the origin and evolution of cellular organisms[J]. Protein Science: A Publication of the Protein Society, 2019, 28(11): 1947-1951.\u003c/li\u003e\n \u003cli\u003eVerhaak R G W, Hoadley K A, Purdom E, et al. Integrated genomic analysis identifies clinically relevant subtypes of glioblastoma characterized by abnormalities in PDGFRA, IDH1, EGFR, and NF1[J]. Cancer Cell, 2010, 17(1): 98-110.\u003c/li\u003e\n \u003cli\u003eSaad S H, Kashanchi A, Zadeh M A, et al. Exosome-Mediated Crosstalk Between Cancer Cells and Tumor Microenvironment[J]. Cells, 2025, 14(22): 1750.\u003c/li\u003e\n \u003cli\u003eLuo H, Zhang H, Mao J, et al. Exosome-based nanoimmunotherapy targeting TAMs, a promising strategy for glioma[J]. Cell Death \u0026amp; Disease, 2023, 14(4): 235.\u003c/li\u003e\n \u003cli\u003eSun M, You H, Hu X, et al. Microglia-Astrocyte Interaction in Neural Development and Neural Pathogenesis[J]. Cells, 2023, 12(15): 1942.\u003c/li\u003e\n \u003cli\u003eLi X, Gou W, Zhang X. Neuroinflammation in Glioblastoma: Progress and Perspectives[J]. Brain Sciences, 2024, 14(7): 687.\u003c/li\u003e\n \u003cli\u003eCatalano M, Limatola C, Trettel F. Non-neoplastic astrocytes: key players for brain tumor progression[J]. Frontiers in Cellular Neuroscience, 2023, 17: 1352130.\u003c/li\u003e\n \u003cli\u003eLin C, Wang N, Xu C. Glioma-associated microglia/macrophages (GAMs) in glioblastoma: Immune function in the tumor microenvironment and implications for immunotherapy[J]. Frontiers in Immunology, 2023, 14: 1123853.\u003c/li\u003e\n \u003cli\u003eLin C, Wang N, Xu C. Glioma-associated microglia/macrophages (GAMs) in glioblastoma: Immune function in the tumor microenvironment and implications for immunotherapy[J]. Frontiers in Immunology, 2023, 14: 1123853.\u003c/li\u003e\n \u003cli\u003eTao J C, Yu D, Shao W, et al. Interactions between microglia and glioma in tumor microenvironment[J]. Frontiers in Oncology, 2023, 13: 1236268.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"High-grade glioma, Astrocytes, Tumor microenvironment, Prognostic marker, Immune infiltration, Cell-cell communication","lastPublishedDoi":"10.21203/rs.3.rs-9156282/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9156282/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground:\u003c/strong\u003e High-grade glioma is a highly malignant primary tumor of the central nervous system characterized by a poor clinical prognosis. The regulatory mechanisms of astrocytes within its TME are not fully understood. This study aimed to investigate the gene expression profiles, functional remodeling, prognostic significance, and regulatory role of astrocytes in the immune microenvironment of HGG, thereby identifying novel molecular markers and therapeutic targets for the management of HGG.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods: \u003c/strong\u003eScRNA-seq was conducted on tumor and peritumoral normal tissues from three patients diagnosed with World Health Organization grade III HGG. To characterize astrocyte properties in HGG, spatial transcriptome deconvolution, cell-cell communication analysis, and differential gene enrichment analysis were performed. WGCNA was applied to data from The Cancer Genome Atlas, while Venn intersection combined with LASSO Cox regression was utilized to identify prognostic core genes derived from astrocytes. To assess the prognostic value and immune regulatory function of the identified core genes, Kaplan-Meier survival analysis, receiver operating characteristic curve analysis, and CIBERSORT immune infiltration analysis were executed.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResults: \u003c/strong\u003eIn HGG, the proportions of astrocytes, microglia, and tumor cells were significantly elevated, with astrocytes emerging as the predominant cell population within the TME and engaging in extensive signaling interactions with microglia. Astrocytes demonstrated a bidirectional functional remodeling characterized by metabolic activation and the inhibition of matrix interactions. A total of five prognostic core genes derived from astrocytes were identified, among which LTF, NOX4, and HSP90B1 served as independent prognostic risk factors. A risk score model based on these genes effectively differentiated survival outcomes for HGG patients. Furthermore, core genes NDUFB2, HSP90B1, and LITAF were found to regulate the infiltration and polarization of CD8+ T cells and M2 macrophages, thereby contributing to an immunosuppressive TME that facilitates the malignant progression of HGG.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConclusions: \u003c/strong\u003eAstrocytes are crucial regulators of high-grade glioma HGG TME remodeling, exhibiting distinct functional changes and molecular mechanisms. The identified prognostic core genes derived from astrocytes may serve as potential molecular markers for assessing HGG prognosis and as candidate targets for astrocyte-targeted antitumor therapies, thereby offering new insights for precision treatment of HGG.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eTrial registration: \u003c/strong\u003eNot applicable.\u003c/p\u003e","manuscriptTitle":"Single-Cell and Spatial Transcriptomics Identify Astrocyte Regulation of the Microenvironment and Prognosis in High-Grade Gliomas","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-21 15:10:20","doi":"10.21203/rs.3.rs-9156282/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"61e7061f-08f4-42d1-9a31-80636673efda","owner":[],"postedDate":"April 21st, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Rejected","date":"2026-05-08T00:15:26+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-07T05:22:03+00:00","index":33,"fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-07T04:40:23+00:00","index":32,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-05-08T00:24:19+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-21 15:10:20","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9156282","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9156282","identity":"rs-9156282","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}