Characterisation of the metabolomes of epigenetically distinct subgroups of paediatric ependymoma

Abstract

Abstract Paediatric ependymoma relapses in up to half of patients, with a five-year survival rate of only 25%. Comprehensive characterisation of the genetic/epigenetic mutation and transcriptome landscape has not yet translated to improved treatments. Metabolomics is a functional genomic approach, offering an opportunity to elucidate aberrant metabolic pathways and new therapeutic avenues. Metabolomics thus far has concentrated on 1H High-resolution Magic Angle Spinning, Magnetic Resonance Spectroscopy and Raman Spectroscopy, where low numbers of metabolites were identified compared to that feasible by Liquid Chromatography tandem Mass Spectrometry (LC-MS/MS). Here we present broader metabolome coverage using LC-MS/MS to separately analyse ependymoma polar and non-polar metabolites, comparing two subgroups from distinct neuro-anatomical compartments with different genetic and epigenetic drivers. We homogenised surgically resected ependymoma tissue from two epigenetic subgroups, posterior fossa A (n=10) and supratentorial RELA fusion (n=5), and extracted polar metabolites and lipids using methanol/water/chloroform 1:1:3. LC-MS/MS using a quadrupole-Orbitrap revealed 167 putative metabolites and 400 putative lipids significantly altered in relative abundance between the two subgroups. The metabolites predominantly mapped onto the taurine and hypotaurine pathway. Grade II and III PF-A tumours could be distinguished by the abundances of 53 metabolites, with three metabolites in the protein-lysine degradation pathway increased in abundance. The study presents a first-in-kind description of the paediatric ependymoma metabolome revealing PF-A and ST-RELA subgroups are metabolically distinct brain tumours and therefore warrant consideration of distinct anti-metabolite therapies. Paediatric ependymoma intra-tumour regions have been collected from 6 surgical resections and ongoing intra-tumour metabolomics/lipidomics integrated with transcriptomics will be discussed.