Exploratory multi-omics links CCL2/TIMP1 axis to immunosuppressive TME in glioblastoma.

Abstract

Glioblastoma (GBM) is defined by extreme lethality and transcriptomic plasticity, but the signatures driving the most aggressive tumors remain incompletely defined. In this exploratory in silico study, TCGA-GBM patients were stratified using a strict 1-year overall survival threshold. We integrated differential expression analysis, WGCNA, single-cell RNA-seq, spatial transcriptomics, and virtual knockout simulations. A high-risk signature centered on CCL2 and TIMP1 was identified. Single-cell and spatial mapping linked these genes to an inflammatory, macrophage-enriched microenvironment. The signature inversely correlated with neuronal synapse mimicry scores, suggesting that extreme aggressiveness involves a macroscopic shift from differentiated neuronal states toward an undifferentiated inflammatory phenotype. Virtual perturbation modeling confirmed CCL2 and TIMP1 as highly interconnected network hubs. Despite limitations inherent to computational and retrospective cohorts, our rigorous multi-omics validation identifies the CCL2/TIMP1 axis as a driver of potential prognostic indicator. These findings generate the hypothesis that these mediators reflect a critical inflammatory, mesenchymal-like tumor microenvironment shift, warranting independent cohort validation and experimental investigation.