ID #157 Vascular architecture and blood-brain barrier function in ZFTA-RELA ependymomas

Abstract

The blood–brain barrier (BBB) normally functions to maintain brain homeostasis through the tight regulation of nutrient and waste exchange, while limiting the passage of most xenobiotics into the central nervous system. There is limited understanding of vascular and BBB heterogeneity across different brain tumor entities, including ZFTA-RELA ependymomas (EPN). Defining the vascular architecture and BBB function of distinct brain tumor entities will be important to optimize drug delivery and therapeutic strategies.

To investigate the vasculature of EPN we utilized an established mouse model of supratentorial ZFTA-RELA EPN created by in utero electroporation (IUE). We first assessed overall vascular content and architecture by immunostaining tumor sections for the pan-endothelial marker CD31, finding increased vascular density and branching. Blood vessels maintained expression of the glucose transporter Glut1 and were negative for the fenestrated endothelial marker Plvap. Vessel integrity was intact, with normal Collagen IV wrapping. Tumor associated vessels were also positive for the tight junction protein Claudin-5, and transmission electron microscopy imaging confirmed the presence of tumor endothelial tight junctions. Moreover, tumors displayed little to no extravascular leakage after intravenous administration of a 10kDa dextran tracer. Additional studies using single cell transcriptional profiling of endothelial cells from both ZFTA-RELA EPN mouse models and primary human tumors are underway to determine similarities and differences compared to normal brain endothelial cells.

While additional studies in primary human tumor samples and patient derived xenografts are needed to validate findings in mouse models, our data thus far shows that ZFTA-RELA EPN are highly vascularized and appear to maintain BBB function.

Future studies on the vascular architecture and BBB function in primary and patient derived xenograft models are required to validate our mouse model findings.