The phosphatase PP4c controls spindle orientation to maintain proliferative symmetric divisions in the developing neocortex

Abstract

In the developing neocortex, progenitor cells expand through symmetric division before they generate cortical neurons through multiple rounds of asymmetric cell division. Here, we show that the orientation of the mitotic spindle plays a crucial role in regulating the transition between those two division modes. We demonstrate that the protein phosphatase PP4c regulates spindle orientation in early cortical progenitor cells. Upon removing PP4c, mitotic spindles fail to orient in parallel to the neuroepithelial surface and progenitors divide with random orientation. As a result, their divisions become asymmetric and neurogenesis starts prematurely. Biochemical and genetic experiments show that PP4c acts by dephosphorylating the microtubule binding protein Ndel1, thereby enabling complex formation with Lis1 to form a functional spindle orientation complex. Our results identify a key regulator of cortical development and demonstrate that changes in the orientation of progenitor division are responsible for the transition between symmetric and asymmetric cell division.

Figure 1

PP4c Localizes to Centrosomes and Is Essential for the Cortical Lamination in the Developing Mouse Brain (A) Western blot of cortical lysates from various developmental stages probed for PP4c. The relative level of PP4c is quantified using Image J software and the level of PP4c at E11.5 is normalized to 1. (B) Immunostaining of E14.5 brain sections shows higher PP4c expression in the ventricular zone (VZ). (C) Coronal sections stained for PP4c (green) and the centrosomal marker γ-Tubulin (red). (D–D″) Higher magnification of box marked in (C). Arrows in (D″) indicate colocalization of PP4c (green) with the centrosomal marker γ-Tubulin (red). (E–J) Confocal images of E18.5 coronal sections from PP4c^fl/+;Emx1Cre brains (Ctr) (E–G) and PP4c^fl/fl;Emx1Cre brains (H–J) stained with layer-specific markers Brn2 (layer II-III), Ctip2 (layer V), and Tbr1 (layer VI) in green. Sections were counterstained with DAPI (blue). Arrows in (H) indicate mislocalized upper layer neurons in deep layers. Scale bars represent 20 μm in (B) and (C), 10 μm in (D)–(D″), and 50 μm in (E)–(J). See also Figures S1 and S2.

Figure 2

PP4c Is Required for Neural Progenitor Maintenance during Early Cortical Development (A–L) Confocal images of coronal sections from PP4c^fl/+;Emx1Cre brains (Ctr) (A–C and G–I) and mutants (D–F and J–L) at indicated developmental stages stained for Tuj1 (A–F, green) or PH3 (G–L, green) and DAPI (G–L, blue). (M) Quantification of PH3-positive cells per 100 μm VZ surface at different developmental stages in controls (blue bar) and mutants (red bar) (at least three brains were analyzed for each genotype). (N–Q) RGPs and BPs are identified by staining for Pax6 (N and P) and Tbr2 (O and Q), respectively, in controls (N and O) and PP4c mutants (P and Q). (R and S) Quantification of Pax6-positive RGPs (R) and Tbr2-positive BPs (S). (T and U) Mitotic exit rates in control brains (T) and PP4c-deficient brains (U): neural progenitors double labeled for EdU (for 24 hr) and Ki67. Cells that exited the cell cycle during the past 24 hr are EdU positive but Ki67 negative. (V) Number of cells exiting the cell cycle was increased in PP4c^fl/fl;Emx1Cre brains compared to controls (n = 3 for each genotype). Data are presented as mean ± SEM; ns, not significant; ^∗p < 0.05, ^∗∗p < 005, ^∗∗∗p < 0.001; Student’s t test for two-tailed distribution with unequal variance. Scale bars represent 50 μm in (A)–(Q) and 20 μm in (T) and (U).

Figure 3

Depletion of PP4c Leads to Misorientation of the Mitotic Spindle of Neural Progenitors (A) Samples of 3D-reconstructed mitotic progenitors from three different classes of divisions. Cell outline and centrosomes (asterisks) are marked by N-cadherin (red) and γ-Tubulin (red), respectively. PH3 (green) marks mitotic DNA (DAPI, blue). (B and C) Distribution of spindle orientation in RGCs from PP4c^fl/+;Emx1Cre (Ctr) and mutant brains (B) and statistic analysis of the distribution of spindle orientation (C). Data are presented as Circular Standard Deviation. ^∗∗∗p < 0.001, Wilcoxon rank-sum test. (D–I) TUNEL (green, D and E) and caspase-3 (red, F) staining in PP4c^fl/fl;Emx1Cre brains. (D′–F′) Enlargements of the areas indicated in (D)–(F) show that Pax6-positive RGPs (red) and Tbr2-positive BPs (red) are negative for TUNEL staining. Newborn neurons labeled with Tuj1 (green), however, are positive for caspase-3, whereas either TUNEL or caspase-3 is barely detectable in control brains (G–I). Scale bars represent 20 μm. See also Figure S3.

Figure 4

Cortical Integrity Is Maintained upon the Depletion of PP4c at a Later Developmental Stage (A and B) Orientation of the mitotic spindle is randomized in RGPs of PP4c^fl/fl;NesCre (mut) brains, while the majority of spindles of RGPs are orientated parallel to the ventricular surface in PP4c^fl/+;NesCre (Ctr) brains at E12.5. Distribution (A) and statistic analysis (B) of spindle orientation are presented. Data are presented as Circular Standard Deviation. ^∗∗∗p < 0.001, Wilcoxon rank-sum test. (C–H) Cortical layers are formed correctly at postnatal day 3 (P3) in PP4c^fl/+;NesCre (Ctr) and PP4c^fl/fl;NesCre brains as revealed by the staining of layer-specific markers Brn2 (C, C′, F, and F′), Ctip2 (D, D′, G, and G′), and Tbr1 (E, E′, H, and H′). DNA (DAPI) is blue. Scale bars represent 50 μm. See also Figure S4.

Figure 5

PP4c Acts through Ndel1-Lis1 to Regulate Spindle Orientation (A) Binding of Ndel1-Lis1 was tested in E15.5 brains by coimmunoprecipitation. Ndel1-Lis1 binding was reduced in PP4c^fl/fl;NesCre brains (Mut) compared to PP4c^fl/+;NesCre brains (Ctr). Data are presented as mean ± SEM. (B and C) In utero electroporation was performed at E14.5 using various constructs indicated. Distribution (B) and statistic analysis (C) of spindle orientation are presented. Spindle angles were analyzed at E17.5. Data are presented as Circular Standard Deviation. ^∗∗p < 0.005, ns, not significant, Wilcoxon rank-sum test. See also Figures S5 and S6.

Figure 6

Notch Activity Is Reduced in Progenitors upon PP4c RNAi (A–H) Notch activity was analyzed 2 days after in utero electroporation performed at E13.5 using a Notch reporter construct (CBFRE-GFP, green) together with constructs expressing Scramble (A), shPP4c (B), shPP4c and an RNAi resistant form of PP4c (Res) (C), shPP4c and Ndel1(SA) (D), shPP4c and Ndel1(SE) (E), Cre under a CAG promoter in PP4c^fl/fl background (F), Ndel1(SA) (G), or Ndel1(SE) (H). (I) Quantification of the percentage of GFP^high cells in the neocortex (the number of cells quantified for each condition is indicated in the bars; at least three animals were analyzed for each condition). ^∗∗∗p < 0.001, data are shown as mean ± SEM, Student’s t test for two-tailed distribution with unequal variance. (J) Decreased Notch target Hes1 mRNA in PP4c^fl/fl;Emx1Cre (Mut.) brains compared to PP4c^fl/fl (Contr.) brains (n = 3 for each condition). ^∗p < 0.05, data are shown as mean ± SEM, Student’s t test for two-tailed distribution with unequal variance.

Figure 7

A Mode that Illustrates Developmental Roles of Spindle Orientation in the Mouse Brain During early neuroepithelial stages before neurogenesis (A), planar spindle orientation is essential for the survival of neuroepithelial progenitors. Misoriented spindles resulting from disruption of Lis1 function lead to dramatic apoptosis in progenitor cells (Yingling et al., 2008). During a critical time window at the onset of neurogenesis (B), planar spindle orientation is required for expansion of the RGP pool and to prevent premature neuronal differentiation. Loss of PP4c mediated by Emx1Cre leads to dramatically increased neuronal differentiation accompanied by depletion of the progenitor pool. At the peak of neurogenesis (C), oblique spindle orientation promotes indirect neurogensis, as observed upon overexpression of mInsc (Postiglione et al., 2011).