GoKawiilExperimental model and participant details
Human participants
All of the deidentified samples used in this study were obtained after informed consent of patients and/or their legal representatives who did not receive compensation. The study was approved by the institutional review board in agreement with local institutional ethics guidelines DFCI 10-417 (Boston Children’s Hospital and Dana-Farber Cancer Institute), S-531/2020 (Heidelberg University) and EK no. 1244/2016 (Medical University of Vienna). Cohort characteristics are provided in Supplementary Table 1.
ZFTA-RELA patient-derived cell lines
EP1NS, BT165 and VBT242 cell lines were generated as previously described47,48 and provided by collaboration partners. Cells were grown in complete neurobasal medium, comprising Neurobasal-A Medium (Thermo Fisher Scientific, 10888022) supplemented with 1× of antibiotic–antimycotic 100× (Thermo Fisher Scientific, 15240062), 1× GlutaMAX (Thermo Fisher Scientific, 35050061), 1 mg ml−1 heparin solution (StemCell Technologies, 7980), 1× B-27 (Thermo Fisher Scientific, 12587010), 20 ng ml−1 recombinant human FGF-basic (Shenandoah Biotech, 100-146) and 20 ng ml−1 recombinant human EGF (Shenandoah Biotech, 100-26). For further passaging, floating cells were centrifuged at 300g for 5 min, dissociated with Accutase (Innovative Cell Technologies, AT104-500) for 5 min at 37 °C and washed with PBS (Gibco, 10010023). All of the cell lines used were authenticated by single-nucleotide polymorphism analysis and cells were regularly tested for mycoplasma contamination.
ZFTA-RELA PDXs
EP1NS, BT165 and VBT242 cells (500,000 cells in 3 μl PBS per mouse) were injected stereotactically into the cortex of 6-week-old female NSG mice (NOD.Cg-Prkdcscid Il2rgtm1Wjl/SzJ, The Jackson Laboratory, 005557), which were treated with 0.05 mg per kg buprenorphine and anaesthetized with 2% to 3% isoflurane. The skull of the mouse was exposed through a small skin incision, and a small burr hole was made using a 25-gauge needle at the selected stereotactic coordinates: 1.0 mm x, 1 mm y and −1.5 mm z. After injection, mice were checked daily for signs of distress, including seizures, weight loss or tremors. Tumour size was monitored monthly by small animal MRI and BLI starting 4 weeks after injection. Mice were euthanized as they developed neurologic symptoms, including head tilt, seizures, sudden weight loss, loss of balance and/or ataxia according to IACUC defined endpoints for central nervous system tumours. No mice exceeded these endpoints. Sample sizes were determined with reference to previous studies to provide sufficient statistical power for detecting biologically meaningful differences. Mice were randomized to experimental groups at the start of each experiment when possible. Blinding was not performed during data collection, but analyses were conducted according to predefined criteria to reduce bias. All animal studies were performed according to protocols approved by the Dana Farber Cancer Institute Institutional Animal Care and Use Committee (IACUC; 13-053).
Molecular classification of tumours
Molecular classification through DNA-methylation profiling was performed using the Heidelberg Brain Tumour Methylation Classifier v.12.4 (https://app.epignostix.com/) as previously described49. Genome-wide DNA methylation profiling was performed using the Infinium HumanMethylation 450k or EPIC Kit according to the manufacturer’s instructions (Illumina). Raw signal intensities from the resulting .idat files were calculated using the minfi Bioconductor package v.1.24.0. The normalization and batch effect correction were performed using the limma package v.3.34.5. Afterwards, β values were calculated before selecting CpG probes for downstream analysis. Unsupervised nonlinear dimensionality reduction was performed using the 1 − variance-weighted Pearson correlation between the samples to construct the distant matrix, which was subsequently used as an input for a t-distributed stochastic neighbour embedding visualization. Classification was performed using the established random-forest algorithm of the Heidelberg Brain Tumour Methylation Classifier v.12.4, which included all molecular EPN groups, including the newly established ST-EPN clusters derived in previous studies13,14.
Detection of fusion genes in tumours
... continue reading