Abstract
Various types of brain tumors occur in both children and adults. These tumors manifest with different characteristics such as malignancy, cellular lineage, location of origin, and genomic profile. Recently, immunotherapy, which manipulates immune cells in the tumor microenvironment (TME) to kill tumor cells, has attracted attention as a treatment strategy for tumors. Here, we analyzed the transcriptomic architecture of the brain tumor microenvironment to provide potential guidelines to overcome the therapeutic vulnerabilities to brain tumors. We decomposed the cellular populations of six brain tumor types (meningioma, pilocytic astrocytoma, ependymoma, medulloblastoma, glioblastoma, and lower-grade glioma) using publicly available microarray data and single-cell RNA sequencing (scRNA-seq) data. Interestingly, transcriptome-based immune cell profiling revealed that infiltrating immune cell types in the brain TME, particularly M2 macrophages, CD8+ T cells, and CD4+ T cells, could be distinguished by tumor type, malignancy, and location. scRNA-seq revealed differences in the proportions of dendritic and mural cells. Unsupervised clustering using immune-related genes divided all samples into two distinct clusters with different characteristics. In addition, immune subpopulations showed disparate reactions after anti-PD-1 therapy for glioblastoma. Our results unveiled the distinct TME across brain tumor types and provided a transcriptomic landscape. Our findings may contribute to realizing future precision medicine, providing a basic rationale for the therapeutics of brain tumors.