mTORC1 signaling and primary cilia are required for brain ventricle morphogenesis

Abstract

Radial glial cells (RCGs) are self-renewing progenitor cells that give rise to neurons and glia during embryonic development. Throughout neurogenesis, these cells contact the cerebral ventricles and bear a primary cilium. Although the role of the primary cilium in embryonic patterning has been studied, its role in brain ventricular morphogenesis is poorly characterized. Using conditional mutants, we show that the primary cilia of radial glia determine the size of the surface of their ventricular apical domain through regulation of the mTORC1 pathway. In cilium-less mutants, the orientation of the mitotic spindle in radial glia is also significantly perturbed and associated with an increased number of basal progenitors. The enlarged apical domain of RGCs leads to dilatation of the brain ventricles during late embryonic stages (ventriculomegaly), which initiates hydrocephalus during postnatal stages. These phenotypes can all be significantly rescued by treatment with the mTORC1 inhibitor rapamycin. These results suggest that primary cilia regulate ventricle morphogenesis by acting as a brake on the mTORC1 pathway. This opens new avenues for the diagnosis and treatment of hydrocephalus.

Keywords: Cilia; Hydrocephalus; Ventricular system; mTORC1 pathway.

Fig. 1.

Ventricular enlargement in ciliary mutants. (A) Representative coronal sections of control and Nestin-K3A^cKO mutant forebrains at E14.5, E18.5 and P2. (B,C) Concomitant increases in the area of the lateral ventricle and decreases in cortical thickness at each embryonic and postnatal stage show the progression of embryonic ventriculomegaly and postnatal hydrocephalus in the ciliary mutant (n=4). (D) Representative Ctip2 immunostaining on coronal sections of control and ciliary mutant forebrains at P2. The boundaries between cortical layers V and VI are indicated by dashed lines. Blue boxes show the area quantified in E. (E) Quantification of the number of Ctip2⁺ cells in the 220-µm-wide area outlined in D (n=6). Scale bars: 0.5 mm (A,D). ns, not significant.

Fig. 2.

Cilia abrogation leads to the progressive enlargement of RGC apical domains. (A) Schematic representation of embryonic forebrain dissection for whole mount preparations of cortical ventricular surfaces. (B) Cortical surfaces immunostained with the ZO-1 antibody shown in Fig. S2A were skeletonized to obtain segmented images of representative cortical surfaces at E12.5, E14.5 and E16.5 of controls and Kif3A ciliary mutants. The surface areas per cell are color coded from white (less than 10 µm²) to dark purple (more than 40 µm²). (C) Quantification of the surface area of the apical domains in controls (white) and ciliary mutants (blue) at E12.5, E14.5 and E16.5. (D,E,F) Sample distribution of apical domain areas in control (white) and Kif3a^cKO (blue) embryos at E12.5 (D), E14.5 (E) and E16.5 (F) (2 µm² bins). Numbers of apical domains measured in B at E12.5 (11805 for controls and 13329 for K3A^cKO, n=3), at E14.5 (12731 for controls and 9173 for K3A^cKO, n=3) and E16.5 (8895 for controls and 7476 for K3A^cKO, n=3). Apical domains measured in C: 100 per genotype blind to the conditions. Scale bar: 5 µm.

Fig. 3.

Apical domain enlargement leads to corticogenesis defects. (A) Triple immunostaining with ZO1 (cell junctions in cyan), GTU88 (pericentriolar marker in magenta) and 4A4 (mitotic cells in yellow) antibodies on E14.5 whole mount cortical surfaces allowed us to classify the cells in three categories: cells in interphase (GTU88⁺/4A4), cells in mitosis (GTU88⁻/4A4⁺) and other cells (GTU88⁻/4A4⁻). (B) Quantification in controls and Nestin-K3A^cKO mutants of the percentage of 4A4⁺ (mitotic) ventricular cells: the total number of mitotic cells contacting the ventricle (number of apical domains in mitosis) was related to the number of GTU88⁻/4A4⁺ cells. (C) Segmented images of the immunostaining shown in Fig. S3A,B, in which mitotic 4A4⁺ cells are shown in yellow. (D) Quantification of the apical domain areas of control (white) and mutant mice (blue) in each cell category. Data are the mean±s.e.m. n=3 per genotype per experimental condition in B and D. Numbers of apical domains analyzed in D in interphase (20279 for controls and 12843 for Nestin-K3A^cKO), mitotic (432 for controls and 245 for Nestin-K3A^cKO) and others (604 for controls and 668 for Nestin-K3A^cKO) cells. (E) En face view and Z-projection of a cilium-less apically positioned PH3⁺ mutant cell double-immunostained with the GTU88 antibody illustrating the methodology for measuring anaphase spindle orientation. The ventricular surface is delineated by the positions of GTU88⁺ dots in neighboring PH3⁻ cells (angle between dashed lines). (F) Quantification of the angle formed by a line passing through both centrosomes of cilium-less and control PH3⁺ cells and the ventricular surface. Anaphase angles measured in L: 324 per genotype. (G) Representative images of PH3⁺ mitotic cells and Tbr2⁺ intermediate progenitors in coronal sections of the somatosensory cortex of Nestin-K3A^cKO mutant mice and controls at E14.5. (H,I) Quantification of PH3⁺ mitotic cells in the SVZ on a 220-µm-wide area from Nestin-K3A^cKO and Nestin-Ift88^cKO mice shows a significant increase in the mutants compared to the controls. (J,K) Quantification of Tbr2⁺ intermediate progenitor cells and DAPI⁺ cells in 220-µm-wide areas of Nestin-K3A^cKO and Nestin-Ift88^cKO mice and their respective controls. Data are the mean±s.e.m. in B,D,H-K or median in F (n=3 per genotype). Scale bars: 5 µm (A,C); 1 µm (E); 50 µm (G).

Fig. 4.

Cilia abrogation leads to an increase in the mTORC1 pathway. (A) Western blot analysis of E14.5 control and cilium-less cortical lysates shows increased phosphorylation of mTOR targets in ciliary mutants compared with controls. (B) Quantification of bands from three independent experiments (A) shows that the increase in mTOR pathway activity in ciliary mutants compared with controls is significant. (C) Double immunostaining of an E14.5 Centrin-2-GFP control cortical ventricular surface with Arl13b (cilia, cyan), and phosphorylated S6RP (p-S6RP, magenta) antibodies showing p-S6RP staining at the mother centriole of the centrosome in ciliated cells. (D) Double immunostaining of E14.5 cortical surfaces from control and Nestin-K3A^cKO ciliary mutants with antibodies against β-catenin (red) and p-S6RP (green). (E) Quantification of the ratio between the intensity of p-S6RP fluorescence and the surface area of the apical domain in controls and Nestin-K3A^cKO ciliary mutants at E14.5 shows an increase in the level of p-S6RP in the mutants. Apical domains measured in E: 100 per genotype blind to the condition. Scale bars: 1 µm (C) and 5 µm (D).

Fig. 5.

Apical domain enlargement and corticogenesis defects at E14.5 are rescued by a single rapamycin injection at E12.5. (A) Cortical surfaces immunostained with ZO-1 antibody shown in Fig. S5B were skeletonized to obtain segmented images of representative cortical surfaces at E14.5 from controls and Nestin-K3A^cKO ciliary mutants injected with rapamycin or vehicle solution. Color code is as described in Fig. 2. (B,C) Quantification of apical domain areas at E14.5 shows a significant decrease in rapamycin-injected (red bar in C) compared with vehicle-injected ciliary mutants (blue bar in C); no significant difference is observed between rapamycin- and vehicle-injected controls (B). (D) Distribution of apical domain surfaces in cortices from E14.5 mutants injected with vehicle (blue) or rapamycin (red). (E) Quantification of the relative number of PH3⁺ cells in 220-µm-wide areas on coronal sections at E14.5 in rapamycin- and vehicle-injected controls and ciliary mutants shows significant rescue of the number of basally positioned PH3⁺ cells, which return to normal levels in rapamycin-injected ciliary mutants but not in controls. (F) Quantification of anaphase angles of PH3⁺ apical cells in control and ciliary mutants injected with rapamycin or vehicle show significant rescue of the mitotic spindle misorientation in rapamycin-injected ciliary mutants. Data are the mean±s.e.m. in B-E or the median in F (n=3 per genotype per experimental condition). Apical domain analyzed in B-D for vehicle (14040 for controls, 14001 for Nestin-K3A^cKO) and rapamycin treatment (14900 for controls, 14399 for Nestin-K3A^cKO). Anaphase angles measured in F: 180 per condition. Scale bar: 5 µm.

Fig. 6.

Ventriculomegaly phenotype at E19 is rescued by repeated rapamycin injections from E14.5 to E17.5. (A) Representative coronal sections of control and Nestin-K3A^cKO mutant forebrains at E19. (B) Quantification of the area of the lateral ventricle in control and ciliary mutants injected with rapamycin or vehicle solution from E14.5 to E17.5 show the rescue of the ventriculomegaly phenotype after successive injections of rapamycin. Scale bar: 500 μm.