Deletion of CPAP causes centriole and cilium loss in neurosphere-derived progenitors. (A) Cell lysates of neurospheres from the indicated mice were analyzed by western blotting using antibodies against CPAP and α-tubulin. α-tubulin was used as a loading control. (B) Immunofluorescence staining of CPAP, the centrosomal protein, γ-tubulin (γTUB) and DAPI in neurosphere-derived progenitors of the indicated mice. Quantitative data are shown at the bottom. Results represent means±s.d. calculated from pooled cells (n) in four independent experiments. ***P<0.001; n.s., not significant. Scale bars: 5 μm (upper panel); 0.5 μm (lower panel). (C) Immunofluorescence staining of the ciliary marker protein ARL13b, the PCM protein PCNT, and DAPI in neurosphere-derived progenitors of the indicated mice. Scale bar: 2 μm. Quantitative data are shown at the bottom. Results represent means±s.d. calculated from pooled cells (n) in five independent experiments. ***P<0.001; n.s., not significant.

Centriole loss induces mitotic abnormalities and p53-dependent apoptosis in neurosphere-derived progenitors. (A) Immunofluorescence staining of the nuclear-mitotic apparatus protein (NuMA) and pericentrin (PCNT) in neurosphere-derived progenitors. Abnormal mitotic spindles, particularly monopolar spindles, are significantly increased in dKO neurosphere progenitors. Quantitative data are shown on the right. Scale bar: 2 μm. (B,C) Centriole loss caused by centrinone (CN) treatment induces p53-dependent apoptosis. Neurosphere-derived progenitors from CTL and p53 KO mice were treated with DMSO or CN and analyzed by immunofluorescence staining with anti-cleaved caspase 3 and TUNEL at 1 and 4 days in vitro (DIV). Scale bars: 50 μm (B); 20 μm (C). (D) The effects of CN-induced centriole loss on apoptosis are correlated with increased p53 expression (left panel) and decreased centriole numbers, as detected by centrin 3 staining (right panel). Scale bar: 20 μm. Results represent means±s.d. from pooled cells from more than three independent experiments. *P<0.05; **P<0.01; ***P<0.001; n.s., not significant.

Cpap deletion leads to RGP delocalization and heterotopia formation in dKO embryonic brains. (A) Representative images of E16.5 CTL, p53 KO, cKO, and dKO cortices stained with antibodies against the RGP markers PAX6 (top) and SOX2 (bottom), CTIP2 (a lower-layer neuron marker), TUJ1 (a neuron-specific β-tubulin) and DAPI (blue). The RGPs are reduced in cKO brains due to p53-dependent apoptosis. MZ, marginal zone; CP, cortical plate; SP, subplate; IZ/SVZ, intermediate zone/subventricular zone. Scale bars: 100 μm. (B) A significant portion of PAX6+ cells are localized outside of the VZ (extra VZ), resulting in a thinner VZ layer in dKO cortices (top). Quantification of the number of means±s.d. PAX6+ cells per unit area (200-μm radial column) in the VZ (green) and extra VZ (red) (CTL, n=3; p53 KO, n=3; dKO, n=3). *P<0.05; n.s., not significant for the differences in the numbers of VZ PAX6+ cells (blue asterisk) and extra VZ PAX6+ cells (red asterisk) are labeled. Scale bar: 50 μm. (C) Representative images of P0 CTL and dKO cortices stained with antibodies against CTIP2 (a lower-layer neuron marker), CUX1 (an upper-layer neuron marker) and DAPI. In P0 dKO mouse brains, the formation of bilateral neuronal heterotopic cortex (HC, blue) was associated with reductions in CTIP2+ and CUX1+ cells, compared to normotopic cortex (NC, red). Quantification data of CTIP2+ and CUX1+ means±s.d. cells per unit area (300-μm radial column) in the NC (red) and HC (blue) (CTL, n=3; dKO, n=3). Scale bar: 100 μm.

Cpap deletion impairs the junctional integrity of RGPs. (A) Representative images of E16.5 CTL, p53 KO, cKO and dKO cortices stained for the polarity protein PARD3 (green), pericentrin (PCNT, red) and DAPI (blue) (left). To clearly represent the relationship between centrosomes and the apical polarity protein PARD3, and to enable quantitation analysis, the merged images of double-labeled PARD3 and PCNT at the end-foot of RGPs at the VZ of E16.5 cortices are shown on the right. Severely impaired PARD3-labeling of the end-foot of RGPs at the ventricular surface was observed in cKO and dKO cortices. (B) Representative images of E16.5 cortices from the indicated mice stained for the adherens junction protein N-cadherin (green), pericentrin (PCNT, red) and DAPI (blue) (left). Representative images of sections double labeled for N-cadherin and PCNT are shown on the right. Scale bars: 10 μm. (C) Quantification data of PARD3 integrity at the apical end-foot at the ventricular surface (120 µm) of E16.5 cortices (CTL, 323 end-feet, n=3; p53 KO, 438 end-feet, n=3; cKO, 96 end-feet, n=3; dKO, 353 end-feet, n=5). An incompletely PARD3-labeled end-foot was counted as an impaired end-foot. Results represent means±s.d. *P<0.05; **P<0.01; n.s., not significant.

Increased neuronal production is observed in neurosphere-derived Cpap knockout progenitors by the in vitro pair-cell assay. (A) Representative images of E12.5 neurosphere-derived precursors of indicated mice immunostained with antibodies against PAX6 (a neural progenitor marker) and TUJ1 (a premature neuronal marker) and counterstained with DAPI. The percentages of P-P (progenitor–progenitor), P-N (progenitor–neuron) and N-N (neuron–neuron) pairs for each cell division were analyzed. Scale bar: 10 μm. (B) Quantification data show a significant increase of P-N and N-N divisions in dKO progenitors. Results represent means±s.d. from pools of the indicated neurosphere-derived progenitors (n) in five independent experiments. *P<0.05; **P<0.01; n.s., not significant.

During postnatal development, P14 dKO brains show severe cerebellar hypoplasia and deteriorating hydrocephalus. (A) Hematoxylin and eosin (H&E) staining of P14 whole brains (sagittal sections) of control, p53 KO and dKO mice. Severe hypoplasia of the cerebellum (CRB) with no foliation was observed in the dKO brain at P14. Heterotopia formation and an enlarged ventricle were also observed in the dKO cortex. Scale bar: 2 mm. (B) Whole-brain H&E staining (coronal sections) shows severe hydrocephalus in the dKO brain at P14. Some tissue tearing was introduced during processing of the fluid-filled brain samples. Section images were obtained using a Pannoramic 250 slide scanner (3DHISTECH). Scale bars: 1 mm.