Multi-omics characterization of the blood-brain barrier in molecular groups of ependymoma

Ependymoma (EPN) represents the third most common pediatric CNS tumor. While recurrence rates remain high, systemic therapies have so far failed to lead to clinical benefit. A better understanding of pathophysiological blood-brain barrier (BBB) characteristics represents an important component in developing effective (pre-)clinical trials.

Our study seeks to increase knowledge of molecular EPN group-specific BBB compositions as a proof-of-concept for other brain tumor entities. Furthermore, we explore the correlation between BBB characteristics and their functional impact to adapt an established in silico model that currently predicts drug penetration over the healthy BBB.

T-distributed stochastic neighbor embedding (tSNE)-based clustering analyses using the most relevant tight junction and transporter gene sets revealed distinct molecular EPN group-specific expression patterns. While PDX models (n=20) showed high similarity with patient tumor samples, IUE mouse models (n=2) did not fully recapitulate these BBB characteristics. Single-cell analyses and spatial mapping of protein abundance allowed dissection of BBB gene expression patterns in endothelial cells (e.g. Claudin5). Functional validation on protein level showed that coherence of RNA and protein is BBB gene-dependent.

The differences in BBB markers between molecular EPN groups may partly explain drug resistance of aggressive EPN as especially ZFTA fusion-positive tumors are characterized by high tight junction expression suggestive of low BBB permeability. Our multi-omics approach is intended to develop a score that further complements our established in silico prediction tool for BBB drug penetration. These findings will be validated in preclinical studies while molecular BBB characterization will be further expanded to other brain tumor entities.

Publikationsverlauf

Artikel online veröffentlicht:
09. November 2023