Fig 2: Astroblastoma-like tumor subtypes exhibit characteristic histologies and genetic alterations.a AB tumor histology is characterized by astroblastic pseudorosettes, and oRG-like morphology in Group C tumors. a–i H&E-stained AB-like tumor cases. Top row MN1-BEND2 cases; Middle row BRAFV600E-mutant MAPK-A cases; Bottom row Group C methylation class tumor cases. Subpial accumulations of tumor cells in C22 (g) demonstrated long basal processes extending to the pial surface and inconspicuous apical processes, morphologically resembling outer radial glial10. Similar cells extend basal processes to blood vessels in C27 and C19 (h–i). b Astroblastoma tumors cells within astroblastic pseudorosettes from Bailey and Bucy’s1 original cohort. a, b Tumor cells with long basal processes resembling those in Group C tumors (Plate III, Figs. 1. and 2). c Tumor cells from a case more similar in appearance to those in MAPK-AB tumors (Plate IV, Fig. 1). Mallory-Davidoff stain (a, b 850× and c, 600×)1. Reproduced with the publisher’s permission. c MN1-BEND2 and MAPK-ABC AB-like tumors show characteristic GFAP immunoreactivity. a–c Only focal staining for GFAP is observed in MN1-BEND2 tumors. d–f MAPK-ABC tumors are strongly and diffusely GFAP immunoreactive. d MN1 diffusely highlights tumor nuclei in all MN1-BEND2 tumors tested (a–c). Significant nuclear MN1 immunoreactivity is absent in non-MN1-BEND2 AB cases (d–f). e MN1-BEND2 tumors showed strong granular cytoplasmic IGF2 immunostaining (a–c), not seen in non-MN1-BEND2 AB-like tumors (d–f). Immunohistochemical staining for all antibodies and samples was performed at least twice with equivalent results. f, g Mutations in AB-like tumor related genes occur in a patient age dependent manner. ABCC1 mutations and fusions all occurred in patients 16 years old or less. Its highest expression in the Allen Human Developmental Transcriptome (AHDT) was before 25 pcw. VEZT and VEZF mutations and CTD-2152M20.2-GOLPH3 fusions were all found in tumors from patients aged 25 and younger (Supplementary Data 2). The highest GOLPH3, VEZT and VEZF1 expression in the AHDT was under 25 pcw. MEG8 fusions occurred only in females 30 years of age and younger, most below 16 years, and only in MN1-BEND2 and MAPK-A tumors. GRIA2 mutations were present in MAPK-ABC cases. ANK3, PARP8 and PTEN alterations were mostly present in non-BRAFV600E MAPK-AB patients. A PARP8 fusion was present in a BRAFV600E tumor. Conversely, high GRIA2 expression spans the fetal and postnatal AHDT, but not below approximately 12 pcw. GRIA2 mutations are absent in MN1-rearranged tumors. Some genetic lesions were relatively specific to AB-like tumor genetic types: MN1-BEND2 fusions and ABCC1 mutation or fusion in MN1-rearranged tumors, and TRIO and KALRN fusions in BRAFV600E tumors. SON mutations span the gamut of patient ages, as does its expression in the AHDT. These presumed somatic mutations were found in tumors clinically presenting at the indicated ages, however this data only indicates that the mutation was acquired in a patient sometime prior to that age, and some could have indeed been germline. Colors are arbitrary in f. In g MN1-BEND2 tumor mutated genes are in red and MAPK-ABC tumor mutated genes are in blue. Source data: Supplementary Data 1 and 2.

Fig 3: Igf2 overexpression does not initiate adrenal tumourigenesis. A- β-catenin is not activated in AdIgf2 adrenals.β-catenin expression was analysed by immunohistochemistry in wild-type (WT, a), ΔCat (positive control, b) and AdIgf2 (c–d) adrenals. Section in c was counterstained with haemotxylin. Nucleo-cytoplasmic staining of β-catenin (white arrowheads) is restricted to zona glomerulosa in WT and AdIgf2 adrenals but spreads throughout the cortex and inside the medulla in ΔCat adrenals. Black arrows show infiltrating mesenchymal subcapsular cells. B -Wnt pathway is not activated in AdIgf2 adrenals. Expression levels of Axin2 and Lef-1, two canonical Wnt pathway targets, were determined by RTqPCR with cDNAs from wild-type, AdIgf2 and ΔCat (positive control) adrenals. Bars represent the mean relative quantification (Rq Tg/WT) of gene expression for each gene in at least 6 adrenals per genotype ± standard deviation. P-value was calculated using Student's t-test. C- Analysis of Ki67 expression. Ki67 was detected by immunohistochemistry on wild-type (WT, a) and AdIgf2 adrenals (b). D- Proliferation is not increased in AdIgf2 adrenals. Numbers of Ki67-positive cells in a whole adrenal section were counted separately in the cortex (Co) and medulla (M). For each zone, the number of cells was corrected for surface and is expressed as a percentage of Ki67-positive cells in one of the control individuals. Bars represent the mean of at least 7 individual counts in wild-type and AdIgf2 adrenals ± standard deviation. P-value was calculated using Student's t-test. E- Cyclin D1 expression is not increased in AdIgf2 adrenals. Expression of Cyclin D1 was analysed by RTqPCR with cDNAs from Wild-type and AdIgf2 adrenals. Bars represent the mean relative quantification (Rq AdIgf2/WT) of gene expression in at least 5 adrenals per genotype ± standard deviation. P-value was calculated using Student's t-test. F- Igf2 overexpression does not induce adrenal tumour formation. The phenotype of AdIgf2 adrenals was followed over a 14 months time course by haematoxylin & eosin staining (b–c) and immunohistochemistry for Ki67 (e–f). A 10 month-old wild-type adrenal was included as a reference (WT, a, d). Some twelve month-old AdIgf2 mice were treated with ACTH for two months in order to increase Igf2 expression. Adrenal histology and proliferation were then assessed by haematoxylin & eosin staining (g) and Ki67 immunohistochemistry (h). Arrows show mesenchymal subcapsular cells. M, medulla; F, fasciculata; G, glomerulosa; Co, cortex. Scale bar is 80 µm.

Fig 4: Metastasis-associated macrophages and fibroblasts express IGF-1 and IGF-2 in metastatic lungs. a Left, identification of metastatic tumor lesions in the lung by bioluminescent imaging technique of orthotopically implanted PY230luc breast cancer cells. Right, images show H&E staining of metastatic foci in the lungs. Arrows indicate metastatic foci, scale bars 200 μm, 100μm, and 50 μm. b Immunofluorescent images of lung metastatic foci stained for F4/80 (green), αSMA (green), Ki67 (red), and nuclei (blue). Scale bar 50 μm. c Quantification of Igf-1 mRNA expression levels in metastatic tumor cells, metastasis-associated non-immune stromal cells and metastasis-associated macrophages isolated from pulmonary metastasis. Error bars represent s.e. (n = 3), * p-value ≤ 0.05, *** p-value ≤ 0.0001 using one-way ANOVA and Bonferroni post hoc test. d) Quantification of Igf-2 mRNA expression levels in metastatic tumor cells, metastasis-associated non-immune stromal cells and metastasis-associated macrophages isolated from pulmonary metastasis. Error bars represent s.e. (n = 3), *** p-value ≤ 0.05 using one-way ANOVA and Bonferroni post hoc test

Fig 5: Fetal/neonatal activation of tumor-associated genes and their products. (a) Quantitative RT-PCR analyses of mRNA expression of tumor-associated genes during fetal/neonatal periods. Each value is expressed as the mean ± SEM. The number of samples in each group was 3, 5, 7, 10, 4, 4, 4, and 7 for E13.5, E16.5, P0 (immediately after birth), P1 (1 day after birth), P3 (3 days after birth), P6 (6 days after birth), 1 m (1 month old), and 5 m (5 months old), respectively. *P < 0.05, **P < 0.01, ***P < 0.001 versus control (5 m); one-way factorial anova. (b) In situ detection of insulin-like growth factor 2 (IGF2), H19 mRNA, trefoil factor 3 (TFF3), and α-fetoprotein (AFP) in developing livers. Immunohistochemistry for IGF2, TFF3, and AFP and in situ hybridization for H19 mRNA in fetal (E16.5) and neonatal (P0) livers. Scale bar = 20 μm.