Innovative In Vivo Imaging and Single Cell Expression from Tumor Bulk and Corpus Callosum Reveal Glioma Stem Cells with Unique Regulatory Programs.

Background/Objectives: High-grade gliomas (HGGs), including glioblastomas, are among the most aggressive brain tumors due to their high intratumoral heterogeneity and extensive infiltration. Glioma stem-like cells (GSCs) frequently invade along white matter tracts such as the corpus callosum, but the molecular programs driving this region-specific invasion remain poorly defined. The aim of this study was to identify transcriptional signatures associated with GSC infiltration into the corpus callosum. Methods: We established an orthotopic xenograft model by implanting fluorescently labeled human GSCs into nude mouse brains. Tumor growth and invasion patterns were assessed using tissue clearing, light-sheet fluorescence microscopy, and histological analyses. To characterize region-specific molecular profiles, we performed microfluidic-based single-cell RNA expression analysis of 48 invasion- and stemness-related genes in cells isolated from the tumor bulk (TB) and corpus callosum (CC). Results: By six weeks post-implantation, GSCs displayed marked tropism for the corpus callosum, with distinct infiltration patterns captured by three-dimensional imaging. Single-cell gene expression profiling revealed significant differences in 7 of the 48 genes (14.6%) between TB- and CC-derived GSCs. These genes-NES, CCND1, GUSB, NOTCH1, E2F1, EGFR, and TGFB1-collectively defined a "corpus callosum invasion signature" (CC-Iv). CC-derived cells showed a unimodal, high-expression profile of CC-Iv genes, whereas TB cells exhibited bimodal distributions, suggesting heterogeneous transcriptional states. Importantly, higher CC-Iv expression correlated with worse survival in patients with low-grade gliomas. Conclusions: This multimodal approach identified a corpus callosum-specific invasion signature in glioma stem-like cells, revealing how local microenvironmental cues shape transcriptional reprogramming during infiltration. These findings provide new insights into the spatial heterogeneity of gliomas and highlight potential molecular targets for therapies designed to limit tumor spread through white matter tracts.