bioRxiv2025
Glioblastoma (GBM) recurrence is driven by tumor cells that infiltrate surrounding brain tissue and evade surgical and therapeutic eradication. Although dysregulated iron handling is a recognized feature of the necrotic and hemorrhagic GBM tumor core, its relationship to invasive tumor cell states remains incompletely defined. Here, we investigated how iron-associated microenvironments relate to transcriptional programs of invasion in human GBM. Using multi-regional single-nucleus RNA sequencing of human GBM specimens encompassing tumor core, invasive front, and infiltrated cortex, we found that malignant cells from the tumor core exhibit coordinated upregulation of iron uptake and storage pathways together with invasion-associated gene programs. At the single-cell level, iron metabolism and invasion signatures were strongly correlated, defining a distinct core-enriched malignant subpopulation with mesenchymal-like transcriptional identity, stress-adaptive features, and angiogenic signaling. These iron-high/invasion-high cells aligned with mesenchymal-and astrocyte-like GBM states and were associated with unfavorable patient survival. Despite elevated oxidative stress and ferroptosis-associated transcriptional pressure, this population concurrently expressed anti-apoptotic and anti-ferroptotic regulators, consistent with an iron-tolerant invasive state. To assess functional consequences of iron exposure, we modeled iron-rich conditions using non-cytotoxic particulate iron in patient-derived GBM cell lines and human organotypic cortical slice cultures. Iron exposure induced intracellular iron accumulation, oxidative stress responses, increased tumor cell motility in vitro, and enhanced invasion within intact human brain tissue. Collectively, these findings demonstrate that iron-rich tumor core niches are closely associated with mesenchymal plasticity and invasive behavior in GBM and support a role for iron-associated microenvironmental pressure in shaping invasive tumor cell states. Key PointsO_LITranscriptional profiling reveals that the GBM tumor core harbors malignant cells that strictly couple active iron metabolism with invasive programs. C_LIO_LIThis iron-accumulating, Mesenchymal-like subpopulation is distinctively characterized by angiogenesis and stress resistance. C_LIO_LIIron supplementation in vitro and human ex vivo models are sufficient to drive mesenchymal transition and significantly enhance tumor migration and tissue invasion. C_LI Importance of the StudyGlioblastoma (GBM) recurrence is driven by highly invasive tumor cells that evade resection and resist therapy. Yet the microenvironmental pressures that push GBM cells into an invasive, therapy-resistant state remain poorly defined. Although iron dysregulation has been implicated across cancers, its role as a microenvironmental determinant of GBM invasion has never been demonstrated in physiologically relevant human systems. By integrating multi-regional single-nucleus RNA sequencing with functional validation in patient-derived GBM lines and human organotypic cortical slice cultures, we uncover a core-enriched, iron-associated mesenchymal program that co-segregates with invasion, stress adaptation, and angiogenic signaling. We further show that physiologically relevant iron exposure is sufficient to induce mesenchymal transition, enhance motility, and accelerate tissue invasion within human cortical architecture. These findings position iron as a selective driver that links the hemorrhagic, necrotic GBM core to the emergence of invasive subpopulations that seed recurrence. The data identify iron handling and stress-response pathways as actionable therapeutic vulnerabilities, providing a foundation for strategies that target metabolic resilience and iron-dependent invasive states in GBM. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=157 SRC="FIGDIR/small/695920v1_ufig1.gif" ALT="Figure 1"> View larger version (35K): org.highwire.dtl.DTLVardef@141cc11org.highwire.dtl.DTLVardef@bcb055org.highwire.dtl.DTLVardef@1dcd332org.highwire.dtl.DTLVardef@ad7cad_HPS_FORMAT_FIGEXP M_FIG C_FIG