SPATIAL, SINGLE-CELL, AND BULK RNA SEQUENCING IDENTIFY MOLECULAR PROGRAMS UNDERLYING MENINGIOMA BRAIN INVASION

Abstract Brain invasion is a negative prognostic factor for meningiomas, but the molecular programs underlying meningioma brain invasion are poorly understood. To address this, we performed transcriptomic analyses of 3 independent datasets: Bulk RNA sequencing from 199 meningiomas (n=33 with brain invasion), single-cell RNA sequencing of 2 regionally distinct samples from the brain tumor interface (BTI) versus tumor core (TC) (n=22,118 single-cell transcriptomes), and spatial RNA sequencing from 6 samples with (n=4) versus without (n=2) brain invasion (n=18,471 spatial transcriptomes). Single-cell and spatial datasets were integrated with Harmony, cell interactions were defined using CellChat, and cross-platform comparisons were performed with the Seurat data integration pipeline. Bulk RNA sequencing revealed enrichment of 1835 genes in meningiomas with brain invasion compared to those without brain invasion, including glial markers (GFAP, NCAN) and markers of extracellular matrix remodeling (PDGFRA, COL9A1). Single-cell RNA sequencing comparison of BTI versus TC meningioma cells identified 2793 genes that were enriched in BTI cells, 268 of which were conserved across bulk and single-cell transcriptomes from brain-invasive meningiomas. Gene ontology enrichment analysis showed molecular programs underlying cell migration and regulation of neurogenesis at the BTI, while glycolysis and antigen presentation programs were enriched in the TC. Cholesterol, CSPG4, and LIF signaling were active in the TC, while SLIT signaling was enriched at the BTI. Spatial RNA sequencing showed enrichment of 439 genes at the BTI, and a conserved set of 12 genes were enriched in BTI meningioma cells across bulk, single-cell, and spatial datasets (TGM2, APOC1, IGFBP6, PLEKHO1, GFPT2, ZYX, IQGAP3, LUM, S100A11, ARPC1B, CDK1, CYCS). Comparison of tumor microenvironment monocytes at the BTI versus TC demonstrated enrichment of 1571 genes that were associated with macrophage migration and activation. These data elucidate molecular programs underlying meningioma brain invasion and provide insights into meningioma intratumor heterogeneity.