Single-cell RNA-sequencing of pediatric ependymoma

Ependymoma is a rare tumor type arising in the brain and the spinal cord of children as well as of adults. However, in children, this cancer is often more aggressive and recurs frequently despite multimodal treatment approach with surgery, radiotherapy and chemotherapy. Consequently, extensive research efforts are currently ongoing to develop strategies by which this tumor type can be targeted more effectively. This research proposal aims to characterize the cellular architecture of childhood ependymoma at a so-far unprecedented resolution in order to obtain new insights about key mechanisms that drive this cancer type and determine its aggressiveness. The ultimate goal of this project is to identify new vulnerabilities of ependymoma that can be targeted by medication. The basis of the proposed project will be a cutting-edge biomedical technology, termed single-cell RNA sequencing (scRNA-seq), which allows precise determination of changes of the genetic code as well as of the inferred behavior of thousands of single-cells isolated directly from patient tumors. Before the implementation of this technology, characterization of tumors has traditionally been performed from whole, minced tumor tissue. This bears the risk that conditions present within defined areas of a tumor or of cancer cells with especially malignant properties could be overlooked, but the detection of which would be very relevant in order to choose an appropriate medication for a patient. This project will be performed by the applicant (Bernhard Englinger, PhD) under the supervision of Dr. Mariella Filbin at Dana-Farber Cancer Institute of Harvard Medical School Boston, whose laboratory offers world- leading expertise in scRNA-seq analysis. Dr. Filbins laboratory has already contributed significantly to the characterization of molecular mechanisms underlying the development of diverse brain tumor types. Now, Bernhard Englinger and Mariella Filbin are poised to move on to the molecular characterization of ependymoma at single-cell level. The project described in this proposal will be performed in frame of a national as well as international collaboration network. This will on the one hand provide access to patient material (including the Department of Pediatrics and Adolescent Medicine at Medical University of Vienna and Boston Childrens Hospital), and on the other hand support the project by cutting-edge computer-assisted data analysis technologies performed at the Broad Institute of the Massachusetts Institute of Technology (MIT). The results gained from scRNA- seq analysis will be the basis for refined therapeutic approaches for ependymoma, which will subsequently be tested in the laboratory using patient-derived tumor cell cultures. Thus, the unprecedented resolution of scRNA-seq will help to determine which genetic changes and cell behaviors account for the growth and aggressiveness of ependymoma and will lay the basis for refined medical management of this disease.

Ependymomas (EPN) are rare brain tumors, which occur both in children and adults. EPN is a heterogeneous group of tumors, and it has long been unknown why some of them harbor good prognosis, while others - especially those that occur in children - are very aggressive. Molecular biological studies conducted in recent years have shown that EPN harbors distinct molecular subtypes. Furthermore, these studies revealed that those subtypes occurring predominantly in children (supratentorial EPN with RELA fusion, posterior fossa EPN group A (PF-A)) are most aggressive. The study of Gojo, Englinger, Jiang, et al., used modern methods including single cell RNA sequencing to analyze gene expression profiles of single tumor cells to describe their molecular biological characteristics. In total, 28 tumors of all molecular subtypes were analyzed. This has shown that EPN exhibits a high intratumoral heterogeneity, i.e. harbors tumor cells with distinct molecular characteristics, which hampers treatment efficacy. Furthermore, we compared EPN gene expression profiles to the developing nervous system, which demonstrated that this tumor appears to originate from specific progenitor cells in the brain. Thus, EPN subpopulations exhibit stem cell characteristics. Interestingly, we also found that heterogeneous EPN cells follow a specific trajectory, by which they aberrantly mirror normal development of stem cell-like cells to more differentiated brain cell types like neurons, astrocytes, ependymal cells. Aggressive EPN tumors harbored high proportions of stem cell-like cells. Furthermore, high expression profiles of stem cell markers were associated with poor survival. In contrast, tumors which harbored predominantly differentiated cells, had a favorable prognosis. This explains, why EPN in children is more aggressive than in adults - in childhood EPN, the proportion of undifferentiated, stem cell-like cells is higher than in EPN subtypes that occur predominantly in adults. These data agree with prior studies from our group, which showed that the stem cell-specific enzyme telomerase is reactivated in aggressive EPN (Gojo et al., Neuro-Oncology 2017). Based on these findings, it appears promising to specifically target the stem cell-like subpopulations in aggressive EPN. Preliminary experiments demonstrated that blockade of the FGFR-, IGF2/IGF1R-, CDK4/6- or HDAC2 signaling pathways are potential molecular treatment targets for such therapeutic interventions. In summary, this study revealed the cellular architecture of EPN, by which we could gain insights on intratumoral differentiation processes. This enabled us to define prognostic and therapeutic signatures for this difficult-to-treat tumor type. This study was a collaboration project between Dana-Farber Cancer Institute and Harvard Medical School Boston, Broad Institute of MIT in Boston (Mariella Filbin), Medical University of Vienna as well as the German Cancer Research Center (DKFZ), and the Hopp Children's Cancer Center in Heidelberg (Marcel Kool).