Ependymoma subpopulation lineages underlie clinical classification and outcome (scRNAseq data set)

ABSTRACT: Ependymoma (EPN) is a brain tumor commonly presenting in childhood that remains fatal in the majority of children. Intra-tumoral cellular heterogeneity in bulk tumor samples significantly confounds our understanding of EPN biology, impeding development of effective therapy. We used single-cell RNA sequencing to identify four neoplastic cell subpopulations in posterior fossa EPN patient samples that underlie traditional subgroup classification and harbor divergent lineage trajectories and clinical outcomes. Neoplastic subpopulations expressed gene signatures associated with normal ependyma, ependymoma, and mesenchymal phenotypes, consequently named ciliated (CEC), transportive (TEC), undifferentiated (UEC), and mesenchymal ependymoma cells (MEC). The abundance of EPN subpopulations as estimated by deconvolution of EPN bulk tumor transcriptomes showed that the ratio of MEC and CEC cell fractions dictated assignment to the two main classification subgroups in PFA. Longitudinal analysis revealed evolution of subpopulation fractions, notably MEC and CEC, which changed between presentation and recurrence. Outcome analyses demonstrated that a higher proportion of UEC or MEC subpopulations was associated with shorter survival, whereas CEC was associated with longer survival. Unsupervised pseudotime analysis revealed divergent subpopulation lineages. The first lineage conformed to a classic cancer stem cell trajectory where UEC progenitors differentiate sequentially to TECs and then CECs. In the second trajectory MECs arise from UECs via TECs, potentially in response to hypoxia-mediated cellular stress, a process that we recreated in vitro in a hypoxia treated EPN cell line. Our study reveals the existence of significant, unappreciated cellular heterogeneity in EPN and provides important resolution of EPN tumor biology.

ORGANISM(S): Homo sapiens

PROVIDER: GSE125969 | GEO | 2020/07/31

REPOSITORIES: GEO