Fig 2: RBBP4 is highly expressed in many malignant pediatric and adult brain cancers. RBBP4 Affymetrix data from 2284 human tumor samples (German Cancer Research Center DKFZ). Y axis represents MAS5.0 normalized expression values. atrt, atypical teratoid/rhabdoid tumor; cns, central nervous system; cpc, choroid plexus carcinoma; cph, choroid plexus hyperplasia; cpp, choroid plexus papilloma; dipg, diffuse intrinsic pontine glioma; etmr, embryonal tumor with multilayered rosettes; ews, Ewing's sarcoma; mb, medulloblastoma; ptpr, papillary tumor of the pineal

Fig 3: Morpholino knockdown of tp53 suppresses apoptosis in rbbp4 mutant neural precursors. (A) Inhibition of Tp53 activity by antisense morpholino injection into embryos from rbbp4 Δ4 /+ adult incross. Image of 1 cell embryo from Almeida et al. 32 (B–U) 2 dpf and 3 dpf wild type and rbbp4 Δ4/Δ4 larval midbrain sections after tp53 knockdown. Activated Caspase‐3 (green), HuC/D (red) and DAPI (blue) labeling at 2 dpf (B–E) and 3 dpf (F–I). (J, K) Quantification of Caspase‐3 positive cells in the midbrain and retina at 2 dpf (un‐injected +/+ n = 4; un‐injected rbbp4 Δ4/Δ4 n = 2; injected +/+ n = 4; injected rbbp4 Δ4/Δ4 n = 4) and 3 dpf (un‐injected +/+ n = 3; un‐injected rbbp4 Δ4/Δ4 n = 4; injected +/+ n = 3; injected rbbp4 Δ4/Δ4 n = 6). (J) 2 dpf midbrain +/+ un‐injected vs injected (P = .0170*), rbbp4 Δ4/Δ4 un‐injected vs injected (*P = .0112). The 3 dpf midbrain +/+ un‐injected vs injected (P = .3349 n.s.), rbbp4 Δ4/Δ4 un‐injected vs injected (P = .1736 n.s.). (K) 2 dpf retina +/+ un‐injected and injected quantification values were zero, rbbp4 Δ4/Δ4 un‐injected vs injected (*P = .0032). The 3 dpf +/+ un‐injected vs injected (*P = .0285), rbbp4 Δ4/Δ4 un‐injected vs injected (*P = .0037). (L–S) γ‐H2AX labeling (green), and nuclear stain DAPI (blue) at 2 dpf (L–O) and 3 dpf (P–S). (T) Comparison of γ‐H2AX positive cells in the midbrain between un‐injected and tp53 MO injected wild type at 2 dpf (un‐injected n = 4; tp53 MO injected n = 3; P = .2344), rbbp4 Δ4/Δ4 at 2 dpf (un‐injected n = 2; tp53 MO injected n = 4; P = .0024), wildtype at 3 dpf (un‐injected n = 3; tp53 MO injected n = 5; P = .1101), and rbbp4 Δ4/Δ4 at 3 dpf (un‐injected n = 6; tp53 MO injected n = 10; P = .0657). (U) Comparison of γ‐H2AX positive cells in the retina between un‐injected and tp53 MO injected wildtype at 2 dpf (un‐injected n = 4; tp53 MO injected n = 3; P‐value = 0), rbbp4 Δ4/Δ4 at 2 dpf (un‐injected n = 2; tp53 MO injected n = 4; P = .0001), wildtype at 3 dpf (un‐injected n = 3; tp53 MO injected n = 5; P = .5784), and rbbp4 Δ4/Δ4 at 3 dpf (un‐injected n = 6; tp53 MO injected n = 10; P = .0128). Statistical analysis was performed with one‐tailed Student's t‐test. Plots show mean ± s.e.m. Scale bars: 100 μm

Fig 4: Ciliary marginal zone neural stem cells continue to divide but neural precursors fail to survive in rbbp4 Δ4/Δ4 homozygous mutant retina. (A–F) BrdU pulse‐chase labeling to examine the fate of newborn neurons in the rbbp4 Δ4/Δ4 mutant retina. (A, B) 2 dpf zebrafish embryos treated with a 2.5 hour BrdU pulse and immediately sacrificed. Transverse sections of retina from wild type (n = 4) and rbbp4 Δ4/Δ4 mutants (n = 5) were labeled with antibodies to BrdU and mitotic marker phospho‐Histone H3 (pH 3). BrdU labels proliferating cells at the retina ciliary marginal zone and cells scattered throughout the laminating retina. (C, D) Pulse‐chased larvae at 3 dpf. In wild‐type retina (n = 8), BrdU‐labeled cells are located in the region adjacent to the cmz where neural precursors and newly differentiated neurons reside (C outline). A small section of BrdU‐labeled cells remains at the ciliary marginal zone in the rbbp4 Δ4/Δ4 mutant retina (n = 6) (D outline). (E, F) Pulse‐chased larvae at 5 dpf. In wild type (n = 7), the BrdU‐labeled cells are now more centrally located in an older region of the growing retina (E outline). In the rbbp4 Δ4/Δ4 mutant retina (n = 7) mature retinal tissue is absent and BrdU‐negative stem cells persist at the ciliary marginal zone (F arrow). CMZ, ciliary marginal zone. Scale bars 50 μm

Fig 5: Neural precursors fail to initiate quiescence in the rbbp4 Δ4/Δ4 homozygous mutant retina. (A) Diagram of pattern of gene expression marking stages of neurogenesis in the developing zebrafish retina. (B–I) Transverse sections of 3 dpf zebrafish retina labeled by in situ hybridization to examine expression of the stem cell marker mz98 (B, F arrows), proliferating progenitor marker ccnD1 (C, G brackets), committed progenitor marker atoh7 (D, H brackets) and differentiating precursor marker cdkn1c (E, I arrows). (B–E) Wild‐type retina (n = 5) shows location of progressively committed precursor cell populations. (F–I) In the rbbp4 Δ4/Δ4 homozygous mutant retina (n = 3) the proliferating and committed neural progenitor cell populations are expanded compared to wild type. cdkn1c expression is absent in the rbbp4 Δ4/Δ4 mutant retina (I arrow). Scale bars: 50 μm