Single-Cell and Bulk Transcriptomic Integration Reveals a Stemness-Related Astrocyte Subpopulation for Prognostic Risk Stratification in Glioblastoma

作者：Jiachong Wang, Jing Li, Chunyuan Zhang, Chunhai Tang, Changfeng Miao, Zigui Chen, Qisheng Luo · 发表于：Current Medicinal Chemistry · 年份：2026 · DOI：10.2174/0109298673451085260213110913 · 被引用次数：2 · 研究领域：Glioma Diagnosis and Treatment、Single-cell and spatial transcriptomics、Ferroptosis and cancer prognosis

Introduction: Glioblastoma (GBM) is an aggressive brain tumor with pronounced heterogeneity. Stemness-related cell subpopulations are crucial for progression and therapy resistance, but their prognostic role remains unclear. Methods: We integrated single-cell RNA sequencing (scRNA-seq) and bulk RNA-seq data to identify a stemness-high astrocyte subpopulation. Key genes were selected to construct a prognostic risk model using Least Absolute Shrinkage and Selection Operator (LASSO) Cox regression, which was validated in The Cancer Genome Atlas (TCGA) and the Chinese Glioma Genome Atlas (CGGA) cohorts. Tumor microenvironment and pathway analyses were performed. in vitro functional assays were conducted on GBM cell lines. results: We identified an astrocyte subpopulation with high stemness potential and four core genes—ALDOA, FABP5, TIMP1, and MT1M—for model construction. The risk model demonstrated robust survival prediction in both TCGA and CGGA cohorts, with high-risk patients exhibiting markedly shorter overall survival (OS). Functional analyses revealed that high-risk tumors exhibited enhanced immune and stromal scores, increased infiltration of immunosuppressive cells, and activation of stemness-related and oncogenic signaling pathways, including EGFR/MAPK, NF-κB, and VEGF-mediated angiogenesis. Results: A four-gene signature (ALDOA, FABP5, TIMP1, and MT1M) was established. The risk model robustly stratified patients into high- and low-risk groups with distinct overall surv...