Transgenic animals

Both hGFAP-cre::CrebbpFl/Fl [8, 20, 27] and hGFAP-cre::lsl-MYCN [19, 20, 28] mouse lines have previously been described. These mice were crossed to generate hGFAP-cre::CrebbpFl/Fl::lsl-MYCN mice. The genotype was confirmed by PCR using genomic DNA from tail or ear biopsies utilizing the following primers: Cre: TCCGGGCTGCCACGACCAA, GGCGCGGCAACACCATTTT, Crebbp: CCTCTGAAGGAGAAACAAGCA, ACCATCATTCATCAGTGGACT; MYCN: ACCACAAGGCCCTCAGTACC, TGGGACGCACAGTGATGG as previously described [8, 28]. Mice were maintained on a C57BL/6J background. Both male and female mice were examined. All experiments were performed according to applicable animal protection laws and were approved by the Government of Hamburg, Germany (Reference TVA N099/2019).

Cell culture experiments

For primary culture of tumor cells, OB cells of 3- to 4-week-old hGFAP-cre::CrebbpFl/Fl::lsl-MYCN mice were isolated and dissociated mechanically and enzymatically with Accutase (#A1110501, Thermo Fisher) for 10 min at RT. The cell suspension was filtered by a 40 µm cell strainer (#352340, Corning) and cells were seeded in a T25 flask (#83.3919.502, Sarstedt). They were cultured in serum-free medium with growth factors (NSC medium: DMEM/F12+Glutamax supplement (#31331093, Thermo Fisher), 2% B27 supplement (#17504001, Thermo Fisher), 4% HEPES buffer (#15630056, Life Technologies), 0.1% MEM non-essential amino acids (#M7145-100ML, Sigma), 0.1% penicillin/streptomycin (#15140122, Life Technologies), 0.02% epidermal growth factor (EGF, #AF-100-15, Peprotech) and fibroblast growth factor (FGF, #100-18c, Peprotech)) for up to 6 weeks. After freezing and thawing, the newly established tumor cell lines were dissociated by using Accutase and used for further experiments. For experiments with secondary NSCs, the V-SVZ of two to four 5-day-old hGFAP-cre::CrebbpFl/Fl::lsl-MYCN mice and two to four CrebbpFl/Fl::lsl-MYCN control littermates was dissociated and pooled within mutant and control group, respectively. In order to remove non-NSC cells, cells were cultured in NSC medium. After DNA isolation with a NucleoSpin Tissue Kit (#740952.50, Machery-Nagel), successful recombination in the V-SVZ cells was confirmed via PCR utilizing the following primers: Crebbp recombined: ACCATCATTCATCAGTGGAC, ATGTAAGAACAGCCCCAAAC; MYCN recombined: GCCCGCGGTGATGCCTTTGAGG, CGGGGACTGGGCGGTGGAAC as previously described [20, 28]. After 7 days in culture, the neurospheres were split into single cells using Accutase, and further experiments were performed.

Histology and immunohistochemistry

For mouse brain analysis, brain tissue was prepared for staining according to standardized procedures by fixation in 4% formaldehyde (#70002-4-10, Grimm med. Logistik) (minimum 12 h), dehydration and paraffin embedding. Sections of 4 μm were cut according to standard laboratory protocols. For histological analysis of tumor cells, cell pellets were fixed in 4% paraformaldehyde (PFA) (#CP10.2, Roth) for 30 min, embedded in agarose (#840004, Biozym) and dehydrated. Sections of 4 μm were cut. Haematoxylin/eosin (H&E) staining followed a standard protocol. Immunohistochemical staining was performed on a staining machine (Ventana BenchMark XT System, Roche), or using a staining kit (Novolink Polymer Detection System, #RE7140-CE, Leica) according to manufacturer’s instructions. Primary antibodies included: Ki67 (#ab15580, Abcam, 1:100), MYCN (#51705, Cell Signaling, 1:100), CREBBP (C-1) (#sc-7300, Santa Cruz, 1:50), SOX2 (#ab97959, Abcam, 1:100), OLIG2 (#ab9610, Millipore, 1:200), MAP2 (#M4403, Sigma, 1:3000), GFAP (#M0761, Dako, 1:200), S100A1 (#Z0311, Dako, 1:100), OTX2 (Thermo Scientific, #1H12C4B5, 1:2000). Chromogenic detection was performed by secondary antibodies and 3,3′-Diaminobenzidine (DAB). Stained slides were analyzed on a microscope (BX43F, Olympus), photographed (OLYMPUS cellSens Entry 1.15 software) and measured using ImageJ.

Immunofluorescence staining

For immunofluorescence staining cells were blocked in 10% normal goat serum (NGS) (#S26-100ML, Merck Millipore) and visualized using the following primary antibodies: anti-GFAP (#14-9892, Invitrogen, mouse IgG1, 1:500), anti-TUBB3 (#T2200, Sigma, rabbit IgG, 1:200), anti-Ki67 (#ab16667, Abcam, rabbit, 1:200). Fluorescent detection was achieved by species-specific fluorophore-linked secondary antibodies: Alexa 555 α-mouse (#CST4409, Cell Signaling, 1:500 and Alexa fluor 488 α-rabbit (#CST4412, Cell Signaling, 1:500). DAPI (#28718–90-3, Roth, 1:1000) was used to stain cell nuclei. A microscope (Eclipse Ti2-E, Nikon) was used for analyses.

Viability assay

To assess cell viability, cells were seeded in a density of 5 × 104 cells in 100 μl NSC medium per well in a 96-well plate (#136101, Thermo Scientific). A CellTiter-Glo® Luminescent Cell Viability Assay (#G7571, Promega) was conducted according to the manufacturer’s instructions. The assay is read by luminescence, which is proportional to the amount of ATP present. The latter is proportional to the count of metabolically active cells. Luminescence was measured at 0, 24, 48, and 72 h post-seeding using a microplate reader (infinite M200, Tecan).

In vitro differentiation assay

To facilitate cell adhesion, coverslips were coated with poly-L-ornithine (#P3655-100MG, Sigma) overnight in a freezer, washed with PBS and coated with laminin (#L2020-1MG, Sigma Aldrich) for 1 h at 37 °C beforehand. Subsequently, cells were cultured in medium containing 10% fetal calf serum (FCS) (#10270106, Life Technol.) to induce differentiation. After 7 days in culture, cells were fixed by PFA (10 min). Immunofluorescence staining was performed, and 10 representative pictures of every coverslip were taken and analyzed using NIS-Elements AR software (5.11.03).

Tumor cell transplantation

For tumor cell injection, tumor cells were dissociated with Accutase, washed and resuspended in a solution of NSC Medium and Matrigel (#356234, Corning). Under anesthesia, the skull bone of recipient mice (8-week-old, CD1 nu/nu mice) was punctured and 1 × 106 cells were stereotactically injected as previously described [29]. Remaining tumor cells were used as a growth control to confirm that cells survived the procedure. Animals were observed for signs of tumor development and sacrificed after 8 months. For details of mouse brain analysis see histology and immunohistochemistry.

Single-cell RNA sequencing and data analysis

OB tumors of a 90-day-old male and a 100-day-old female hGFAP-cre::CrebbpFl/Fl::lsl-MYCN mouse were isolated and minced with scalpels on ice. For enzymatic dissociation, 5 ml Papain (#LK003178, Worthington-Biochem) in DMEM/F12 (#21331020, Life Technologies) solution, supplemented with 1000 U DNase (#10104159001, Roche), was added to each sample, followed by incubation at 37 °C in 5% CO2 for 30 min. The cell solution was then passed through a 40 µm cell strainer. Red blood cells were lysed with ACK lysis buffer (#A1049201, Thermo Fisher) (5 min at 4 °C). The resulting single-cell suspensions were stained with 7-aminoactinomycin D (7-AAD) (#699350, Invitrogen). Non-viable, 7-AAD-positive cells were removed by fluorescence-activated cell sorting. Approximately 10,000 vital cells were used as input for scRNA-seq. Single-index libraries were generated with Chromium Single Cell 3′ v3.1 technology (10x Genomics) and sequenced using a NextSeq 2000 sequencing instrument (high-throughput kit, 100 cycles) at the Genomics Core Facility (University Hospital Münster, Germany) after quality control using a Tapestation 2200 (Agilent Technologies). The samples were analyzed with the 10x Genomics CellRanger v6.0.2 pipeline [30] and Seurat R package v4.0.5 (ref. [31]). Raw data were converted to fastq format with the CellRanger mkfastq function and then aligned against the murine reference transcriptome mm10 v2020-A with CellRanger count and default values. Seurat objects were generated for both samples based on the following filter criteria: at least three cells, a minimum feature count of 200, and cells with < 25% of mitochondrial genes. Outlier cells with a high nCount_RNA value were classified as doublets and removed (threshold: 30,000-35,000). The filtered data were then normalized, integrated, and clustered with Seurat, using a resolution parameter of 0.5. Feature plots, UMAPs and heatmap visualizations were created with Seurat functions; a cluster-based cell type annotation was conducted based on the expression of characteristic marker genes per cell type. Finally, a logistic regression analysis based on the approach of Young et al. [32] was used to calculate similarity scores between the original murine clusters and a reference dataset by Mizrak et al. [24].

Comparison of human ONBs and mouse tumors

Global DNA methylation data of mouse and human tumors generated by Illumina array sequencing (450k or mouse methylation bead chip array, respectively) was used for creating copy number variation (CNV) profiles. Human data originates from Capper and Engel et al. [33]. CNV data was generated with the conumee package in R.

Mouse methylation profiles were generated with the in-house established pipeline for human material and the mouse methylation bead chip array MM285 (Illumina) was employed. Mouse CNV profiles were generated similar to human CNVs with an in-house adapted code and a custom reference set.

Statistical analysis

Statistical analysis was conducted with GraphPad Prism software (RRID:SCR_002798). All observations were made in at least three independent animals or experiments. For cell culture experiments and animal studies, all analyses were permored in a blinded fashion without knowlodge of the genotype. No further randomization was needed. If not stated otherwise, all data presented are mean ± SD, with n = 3 for each group. Each data point represents an individual animal or an independent experiment. If not stated otherwise, the unpaired t test (two-tailed) was applied to compare the means of two groups. P-values < 0.05 were considered significant (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001).