TMIC-51. NEUROIMMUNE-COMPETENT ORGANOID MODELING OF MICROGLIA/TUMOR INTERACTIONS AND DRUG RESPONSIVENESS

Abstract INTRODUCTION Due to its highly aggressive and infiltrative nature, lack of effective treatment options and localization in the skull base, diffuse intrinsic pontine glioma (DIPG) is universally fatal with less than a year median survival time. Studying how DIPG cells interact with brain cells in its microenvironment, including microglia which is the primary brain immune cell type, may lead to new therapeutic strategies to halt the invasion of DIPG into brain tissue. We developed a clinically relevant human immunocompetent brain organoid-DIPG fusion model for investigating the tumor invasion process at the cellular and molecular levels. METHODS Neuronal progenitor and GFP-labeled myeloid precursor cells were derived from human embryonic stem cells and combined to self-assemble and co-develop over two months into microglia-containing brain organoids (MiCBOs). The development of the MiCBOs as well as the distribution and motility of the microglia within the brain organoid was characterized by confocal and multiphoton microscopy, immunofluorescence staining and Western Blots. Confocal live-cell imaging was used to investigate the invasion of TdTomato-labeled tumor cells into MiCBO. We further performed single-cell RNA-Sequencing (scRNA-Seq) analysis to gain insight into the cell-specific gene expression changes after the confrontation of the MiCBOs with tumor spheroids. RESULTS MiCBOs contain viable, widely distributed and highly motile microglia as well as functionally mature neurons, highly resembling those in the human brain. scRNA-Seq showed distinct gene expression patterns in both microglia and tumor cells after confronting MiCBOs with DIPG spheroids, suggesting candidate signaling pathways between those two cell types. CONCLUSION Our novel MiCBO-DIPG fusion model provides a pathophysiologically-relevant system to elucidate molecular mechanisms underlying cell-cell interactions during tumor invasion in a human brain tissue context. This scalable model can be adapted to drug screening applications to identify therapeutics that modulate neuroimmune/tumor interactions and to establish the therapeutic index for anti-cancer medications.