LGG-13. GANGLIOGLIOMA DEEP TRANSCRIPTOMICS REVEALS A CD34+ PRIMITIVE NEUROECTODERM NEURAL PRECURSOR-LIKE POPULATION

Gangliogliomas are glioneuronal brain tumors that typically present in childhood or early adulthood. Though most often low-grade, new insights are needed to refine glioneuronal tumor classification and to identify therapeutic approaches, particularly for unresectable, high-grade, and/or recurrent disease. Gangliogliomas often possess a rare population of immature astrocyte-appearing CD34+ cells. CD34 is expressed in neuroectoderm neural precursor cells during embryogenesis. Neural cell CD34 expression is usually lost prior to birth. We hypothesized that CD34+ ganglioglioma cells represent tumor precursor/stem cells and closely resemble ganglioglioma cells of origin. To test this, we evaluated five gangliogliomas with single nucleus RNA-seq, cellular indexing of transcriptomes and epitopes by sequencing, and/or spatially-resolved RNA-seq. Developmental trajectory analyses uncovered a neoplastic cellular hierarchy, with CD34+PAX6+SOX2+MSI1+PECAM1-VWF-DCN-COL1A2- cells being the most primordial and precursor to neuron-like and macroglia-like neoplastic cell states. Gene regulatory network inference of CD34+ ganglioglioma cells indicated TCF7L2/MEIS1-PAX6 and SOX2-mediated transcriptional programming, similar to that found during neuroectodermal/neural development in utero. Spatially-resolved transcriptomics revealed a neuroectoderm neural precursor-like tumor cell niche relatively devoid of vascular and immune cells. We used these high-resolution results to deconvolute clinically-annotated transcriptomic data, confirming that CD34+ neural precursor-like cell-associated gene programs associate with gangliogliomas compared to other brain tumors. Together, these deep transcriptomic approaches characterized a ganglioglioma cellular hierarchy - identifying CD34+ neuroectoderm neural precursor-like tumor-initiating cells, corresponding regulatory transcriptional programs, cell signaling pathways, and associated immune cell states. These findings may facilitate improved ganglioglioma tumor classification, diagnosis, and therapeutic investigations.