SUPPLEMENTARY INFORMATION

Transcription

1 DOI: /ncb2880 Supplementary Figure 1 Sequence alignment of Deup1 and Cep63. The protein sequence alignment was generated by the Clustal X 2.0 multiple sequence alignment program using default parameters. The protein sequences were from GenBank accessions numbers NM_ (human DEUP1), KC (mouse Deup1), NM_ (human CEP63), and KC (mouse Cep63). The histogram depicts conservation between the two proteins. Related to Figure

2 Supplementary Figure 2 The aggregates formed by GFP-Cep63 or GFP- Cep152 do not support de novo procentriole formation. (a) Expression of GFP-tagged Deup1, Cep63, or Cep152 in U2OS cells. (b,c) Analyses for procentriole formation. The arrowheads and arrows point to typical mother centrioles and GFP-positive aggregates, respectively. The bright SAS-6 foci, which mark the cartwheels, are used as procentriole markers in the quantification analyses. Only interphase cells containing MCD procentrioles, thus in the S or G2 phase, were scored (n=30 cells from triplicates). GFP- Deup1-expressing cells served as the positive control for de dovo procentriole biogenesis. (d) Endogenous CEP152 exhibits punctate distributions in the GFP-Cep63 aggregates. Such a pattern is in sharp contrast to the ring-shaped distributions of CEP152 in the deuterosomes (arrow in the GFP-Deup1-positive cell) and MCD cradles (arrowheads) (also see Fig. 1h), suggesting that the GFP-Cep63 aggregates are disorganised. Apparently, the punctate CEP152 is unable to induce the cartwheel assembly, not to mention de novo centriole formation. Related to Figure

3 Supplementary Figure 3 Immunoblotting results for domain mapping. (a) Mapping of the Deup1- and CEP63-interactiing domains of Cep152. The numbers in the diagrams indicate amino acid positions. Exogenous Cep152, its mutants, or firefly luciferase (Luc) was expressed alone or together with Flag-Deup1 in HEK293T cells, as indicated. Coimmunoprecipitation was then performed using anti-flag resin. Luciferase served as the negative control. The asterisks indicate the positions of full-length fusion proteins. (b) Mapping of the CEP152-interacting domain of Deup1. Flag-tagged Deup1 or mutants were expressed in HEK293T cells. Coimmunoprecipitation was then performed. (c) Mapping of the CEP152-interacting domain of Cep63. Flag-tagged Cep63 or mutants were expressed in HEK293T cells. Coimmunoprecipitation was then performed. These results are summarised in Figure 1j. 3

4 Supplementary Figure 4 Effect of Deup1 or Cep63 RNAi in MTECs at ALI d 3. (a) Cep152 failed to exhibit the ring-shaped deuterosome localisation upon Deup1 RNAi. The arrows indicate MCD procentriole. The insets are magnified 2 to show typical de novo procentrioles (arrowheads). (b-d) Ablation of Cep63 alone in MTECs does not block deuterosome formation and centriole amplification. MTECs were infected with lentivirus to express control (Ctrl-i) or Cep63-specific (63-i1 or 63-i2) shrna together with GFP-Centrin1. The GFP-positive cells at ALI d 3 were sorted out using FACS and subjected to immunoblotting to show RNAi efficiency. a-tubulin served as the loading control. The insets are magnified to show MCD procentrioles. The statistical results for centriole numbers are presented in Figure 6d. 4

5 Supplementary Figure 5 Validation of the specificity of Dp-MO1 and Dp- MO2 using a GFP reporter. (a) Diagrams of in vitro-transcribed mrnas. The red bars indicate the locations of the MO-targeting sites. (b) Dp-MO1 and Dp-MO2 specifically blocked the translation of the xdpatg-gfp mrna. Xenopus embryos at the 2- or 4-cell stage were microinjected with a mixture of 10 ng MO, 100 pg GFP or xdpatg-gfp mrna, and 50 pg RFP-F mrna (as an injection marker). The autofluorescence of GFP or RFP was visualised at approximately the developmental stage 26. Related to Figure 8d. 5

6 Supplementary Figure 6. Full scans of original blots. The boxed regions indicate the areas shown in the figures. 6

7 Supplementary Video Legends Supplementary Video 1 3D model of a part of a cell in stage II immunostained for Deup1 (green), Cep152 (blue), and Centrin (red). The top and side views Supplementary Video 2 3D model of a part of a cell in stage III immunostained for Deup1 (green), Cep152 (blue), and Centrin (red). The top and side views Supplementary Video 3 3D model of a part of a cell in stage IV immunostained for Deup1 (green), Cep152 (blue), and Centrin (red). The top and side views Supplementary Video 4 3D model of a part of a cell in stage V immunostained for Deup1 (green), Cep152 (blue), and Centrin (red). The top and side views Supplementary Video 5 3D model of a part of a cell in stage VI immunostained for Deup1 (green), Cep152 (blue), and Centrin (red). The top and side views Supplementary Video 6 EM images from four consecutive sections of a control (Ctrl-i) and a Deup1-knockdown (Dp-i1) MTEC at ALI d3. The thickness of each section is 70 nm. Deuterosomes (d), mother centrioles (m), and daughter procentrioles around mother (p) are numbered according to the sequence of appearance. The overlaid images are shown in Figure 7a. Supplementary Video 7 3D model for deuterosome-like structures formed by His-Deup1 (green) in E. coli. The protein expression was induced with 10 mm IPTG. See Figure 7d for the original 3D-SIM image. Supplementary Video 8 3D model for deuterosome-like structures formed by His-Deup1 (green) in E. coli. The protein expression was induced with 50 mm IPTG. See Figure 7d for the original 3D-SIM image. Legends for Supplementary Tables Supplementary Table 1 Sequences of sirnas, shrnas, MO, and primers. Supplementary Table 2 List of antibodies. Supplementary Table 3 Statistics source data. 7