Volume 93, Issue 3, Pages 542-551.e4 (February 2017) Differential Routing of Mindbomb1 via Centriolar Satellites Regulates Asymmetric Divisions of Neural Progenitors  Samuel Tozer, Chooyoung Baek, Evelyne Fischer, Rosette Goiame, Xavier Morin  Neuron  Volume 93, Issue 3, Pages 542-551.e4 (February 2017) DOI: 10.1016/j.neuron.2016.12.042 Copyright © 2017 Elsevier Inc. Terms and Conditions

Figure 1 Distribution of Centriolar Satellites and Mib1 in Neuroepithelial Progenitors in Interphase and Mitosis (A) Apical view of the neural tube at embryonic day 3 (E3) showing the localization of Mib1 (red) in relation with the two centrioles (green). (B) Apical views showing the localization of Mib1-GFP (green) in relation with either AZI1-Flag or PCM1 (red) and the centrioles (blue). (C) Apical views at E3 showing the localization (in red) of Mib1 (i), AZI1-Flag (ii), PCM1 (iii), Mib1-RFP (iv), BBS4-GFP (v), and OFD1-GFP (vi) in relation with the cilium (blue) and the centrioles (green). The daughter centriole is identified as the one not carrying the cilium and is framed with a dotted line. The juxtaposed diagrams indicate the average intensity of each marker in the vicinity of the mother (black) or the daughter centriole (gray). Data represent means ± SEM, ∗∗∗p < 0.001 (Student’s t test); n = 49, 28, 31, 24, 20, and 28 cells analyzed for Mib1, AZI1-Flag, PCM1, Mib1-RFP, BBS4-GFP, and OFD1-GFP, respectively. The measurement method is schematized in vii and detailed in the appropriate section. In (A)–(C), centrioles are labeled with Centrin2-GFP (A; i and iii), pan-centrin antibody (B, right; iv), FOP antibody (B, left; ii, v, and vi). The cilium is labeled with Arl13b-GFP (ii and iv–vi) or Arl13b antibody (i and iii). (D) Top: average intensities of Mib1-Myc staining at the centrosome in control versus AZI1-Nter-transfected embryos. Data represent means ± SEM, ∗∗p < 0.01 (Student’s t test), n = 70 and 57 cells analyzed for control and AZI-Nter conditions, respectively. Bottom: apical views showing the localization of Mib1-Myc (red) in relation with the centrioles (blue) and tight junctions (labeled with ZO1-GFP in green) in the two situations. (E–G) The diagrams indicate the symmetry index for PCM1 (E), AZI1-Flag (F), and Mib1 (G) at the centrosome in comparison with the centrosome markers γTubulin and FOP in mitotic cells (each mark corresponds to a single cell; horizontal bars correspond to medians; the pink dotted line indicates the value of the mean centrosomal marker index minus its SD (for all mitotic phases), providing a threshold above which distribution is essentially symmetric). In the case of PCM1, cells were divided into two populations, displaying low versus medium and high PCM1 centrosomal enrichment (see Method Details). The lower panels show representative examples of the localization of PCM1, AZI-Flag, and Mib1 (red) in relation with the centrosomes (blue) and chromosomes (green), labeled with the indicated markers and at the indicated mitotic stages. Neuron 2017 93, 542-551.e4DOI: (10.1016/j.neuron.2016.12.042) Copyright © 2017 Elsevier Inc. Terms and Conditions

Figure 2 A GA-Associated Pool of Mib1 Can Compensate for Mib1 Centrosomal Asymmetry during Mitosis (A) Time series (en-face imaging) of dividing neuroepithelial cells expressing Mib1-GFP (green) and the centrosome reporter PACT-mKO1 (PACT domain of pericentrin fused to Kusabira-Orange, shown here in red). The dotted line schematizes the outline of the cell. Top: in cells exiting mitosis with symmetric Mib1-GFP distribution, Mib1-GFP is initially asymmetric in prophase and observed on only one centrosome (white arrowhead). A pool of non-centrosomal scattered Mib1-GFP (blue arrowheads) appears in prophase, coming from the basal end of the cell (see transverse view), and progressively aggregates near the second centrosome (black arrowhead), resulting in equal centrosomal accumulation of Mib1-GFP at telophase or shortly after. Bottom: conversely, in cells exiting mitosis with asymmetric Mib1-GFP, only the centrosomal pool of Mib1-GFP is observed. A total of 39 cells (4 embryos from 4 independent experiments) with either symmetric (18) or asymmetric (21) Mib1 inheritance were monitored. Symmetry indices for the non-centrosomal “cytoplasmic” and centrosomal pools of Mib1 are plotted in Figure S2C. (B) Transverse sections of the chick neural tube (thoracic level) at E2 (HH st12), E3 (HH st18), and E4 (HH st22) stained with Mib1 antibody (red) and either Giantin (Golgi apparatus [GA] marker) or γ-Tubulin (centrosome marker) (green). Mib1 co-localizes with Giantin at E2 and E3 and with γ-Tubulin at E3 and E4. (C) Transverse sections of the neural tube at E3 showing individual cells transfected with Mib1-RFP (yellow), GalT-CFP (GA reporter, cyan), and membrane-GFP (magenta). (D) 3D reconstruction from en-face imaging of a cell transfected with Mib1-GFP and the GA marker GalT-RFP 8 hr post-transfection at E2. Left: schematic showing the sub-cellular localization of the markers. The framed area indicates the region of interest. Right: 3D reconstructed time series of a cell co-transfected with Mib1-GFP (green) and the GA reporter GalT-RFP (red). Time points 0′–24′: before mitosis, the cell displays an apical (white arrowhead) and a GA-associated (pink arrowhead) pool of Mib1-GFP. Time points 28′–52′: the cell enters mitosis and the GA-associated Mib1-GFP detaches from the GA as the latter fragments. The dotted line schematizes the outline of the cell. 11 cells (3 embryos from 3 independent experiments) with a similar behavior were monitored. Three other examples are displayed in Figure S2. Neuron 2017 93, 542-551.e4DOI: (10.1016/j.neuron.2016.12.042) Copyright © 2017 Elsevier Inc. Terms and Conditions

Figure 3 Mib1 Unequal Inheritance Correlates with Asymmetric Fates of the Daughter Cells (A and B) Time-lapse series (en-face imaging) of dividing neural progenitors showing symmetric (A) or asymmetric (B) Mib1-GFP localization in relation with the centrosome labeled with PACT-mKO1. Top and bottom panels show apical and transverse views, respectively, while the middle panels schematize the position of Mib1 (green) and centrosomes (red) in interphase (black dotted line) and mitosis (gray dotted line). Arrowheads point to centrosomes and the asterisks indicates the loss of centrosomal staining from the observation field. The right panel is a schematized transverse view of each situation using the same color code. In all cells followed, Mib1-GFP was detected in the first cell division. However, Mib1-GFP localization at the centrosome was very variable over time in each cell, due either to Mib1 dynamics or to photobleaching. Thus, Mib1-GFP distribution was only considered for the first division and was used to categorize daughter cells as having or not inherited Mib1-GFP. n = 27 clones analyzed from 7 embryos. (C) Diagram obtained from the time-lapse series presented in (A) and (B) showing the percentage of each cell lineage situation (P-P, P-N, N-N) in the case of symmetric or asymmetric Mib1-GFP inheritance. P-P, P-N, and N-N stand for divisions producing two progenitors, one progenitor and one neuron, or two neurons, respectively. (D) Time-lapse series (transverse section) of a dividing progenitor showing Mib1-RFP inheritance by the daughter cell losing the basal attachment (representative example of n = 11 cells from 10 slices from 6 embryos). Neuron 2017 93, 542-551.e4DOI: (10.1016/j.neuron.2016.12.042) Copyright © 2017 Elsevier Inc. Terms and Conditions

Figure 4 Altering Mib1 Centrosomal Localization Affects the Mode of Division of Neural Progenitors (A) Left: schematic of the neural tube showing the ventricular and the mantle zone identified by the expression of Sox2 and HuC/D, respectively. The boxed area indicates the region of interest. Middle: transverse sections of the neural tube 40 hr after transfection of empty or AZI1-Nter vectors. Sox2 and HuC/D immunostainings label progenitors and neurons, respectively. The brackets indicate groups of ectopic Sox2 progenitors intermingled with HuC/D neurons. Right: differentiation rate (number of HuC/D+ cells on total) in transfected cells 40 hr after transfection in control versus AZI1-Nter transfected spinal cords. Data represent mean ± SEM, ∗∗∗p < 0.001 (Student’s t test), n = 14 sections (4 embryos) and 18 sections (8 embryos) analyzed for control and AZI1-Nter conditions, respectively. (B) Left: schematic of the neural tube. The mantle zone is identified by the expression of HuC/D. The boxed area indicates the region of interest. Middle: transverse sections showing a large view (left) and specific two cell clone examples (right) of the neural tube 40 hr after transfection of Cytobow, followed by immunostaining with HuC/D. Right: diagram indicating the percentage of P-P, P-N, and N-N clones for control and AZI1-Nter transfected embryos. The distribution of P-P and P-N clones between control and AZI1-Nter was compared using a Chi2 test, ∗p < 0.05. (C) Left: schematics indicating the promoter and the downstream coding sequences of the vectors transfected in the two conditions displayed in the middle panel and quantified in the right panel. Middle: sister cells (indicated by the white arrowheads) 20 hr after low-voltage (17 V) electroporation of Inscuteable (Insc) alone or with AZI1-Nter and co-transfected with the Notch reporter Hes5-VNP. Right: diagram showing the ratio of Hes5 activity between sister cells. Each mark represents a clone, and horizontal bars correspond to medians; ∗∗∗p < 0.001 (Mann-Whitney test). Neuron 2017 93, 542-551.e4DOI: (10.1016/j.neuron.2016.12.042) Copyright © 2017 Elsevier Inc. Terms and Conditions