Fig 2: TBK1 directly binds to and phosphorylates CEP170 and NuMA.(a) Detection of co-localization of TBK1 and CEP170 as well as TBK1 and NuMA using proximity ligation assay (PLA). Green fluorescent foci represent co-localization of TBK1 and CEP170 or TBK1 and NuMA. TBK1 antibody alone or TBK1 antibody along with IgG were used as negative controls. (b) Immunoprecipitation–western blot analysis showing the interaction of endogenous TBK1 with endogenous CEP170 and NuMA in lysates of asynchronously growing H1650 cells. (c) Immunoprecipitation–western blot analysis of H460 cell lysates showing the binding of TBK1 with NuMA and CEP170 when released from G1/S block. This interaction was inhibited when cells were released from double-thymidine block in the presence of BX795. (d,e) In vitro binding of full-length TBK1, the kinase domain of TBK1 (1–294) or the TBK1 C terminus (TBK1-284-729) with CEP170 (d) or NuMA (e). The data presented are representative of three independent experiments. Lysate/ input lane represents 20% of the input 35S-methionine-labelled TBK1 used in the pull-down assay. (f) Immunoprecipitation–western blot analysis showing inhibition in serine and threonine phosphorylation of CEP170 and NuMA when TBK1 was depleted following transduction of cells with lentivirus expressing shRNA targeting TBK1. (g) In vitro kinase assays showing the phosphorylation of NuMA and CEP170 by TBK1. Coomassie brilliant blue (CBB) staining shows the amount of protein in each reaction.

Fig 3: Phospho-TBK1 co-localizes with CEP170 and TBK1 inhibition provokes supernumerary centrosomes.(a) Confocal microscopy images of control and BX795-treated H460 cells at metaphase. Cells were stained for phospho-TBK1 (green), alpha tubulin (red) and DNA (DAPI, blue). Scale bar, 10 µm. (b) Confocal microscopy images of control and BX795-treated AALE cells at metaphase. Cells were stained for phospho-TBK1 (green), CEP170 (red) and DNA (DAPI, blue). Scale bar, 10 µm. (c) H460 cells stained for phospho-TBK1 (green), CEP170 (red) and DNA (DAPI, blue) in control and TBK1 inhibitor-treated cells. CEP170 fails to co-localize to centrosomes in BX795-treated cells. (d,e) TBK1 depletion by shRNA resulted in multiple centrosomes in H1650 (d) and H460 (e) cells. Cells were stained for phospho-TBK1 (green), CEP170 (red) and DNA (DAPI, blue). (f) Inhibition of TBK1 by 2 µM MRT affected spindle pole localization of NuMA. H460 and HeLa cells were stained for NuMA (green), Alpha Tubulin (red), DAPI (Blue). Scale bar, 10 µm. (g,h) Immunofluorescence staining for Dynein intermediate chain (red) and NuMA (green) in shcontrol and shTBK1-transfected H1650 (g) and A549 (h) cells. Scale bar, 10 µm.

Fig 4: Increased NPM1 phosphorylation at T199 leads to centrosome duplication in 14-3-3?knockdown cells.(a) Protein extracts from vector control or 14-3-3? knockdown cells were resolved on SDS-PAGE followed by immuno-blotting with the indicated antibodies. Actin served as loading control. (b,c) HCT116 cells were transfected with mCherry-a-tubulin and co-transfected with either the control plasmid (pCDNA3) or plasmids encoding wild type cdk1 (cdk1) or constitutively active cdk1 (cdk1AF). Post-transfection the cells were stained with antibodies to Cep170 (green) and DAPI (blue) (b). The percentage of mitotic cells with >2 centrosomes was determined in three independent experiments (c). (d) Protein extracts from the vector control and 14-3-3? knockdown cells were resolved by SDS-PAGE, followed by Western blotting with the indicated antibodies. Note that NPM1 is phosphorylated on T199 to a greater degree in the 14-3-3? knockdown cells. Actin served as loading control. (e) The 14-3-3? -knockdown and vector control cells were transfected with either the vector control or FLAG-epitope tagged versions of WT NPM1 or the NPM1 mutants (T199A and T199D). Post transfections, protein extracts prepared from these cells were resolved by SDS-PAGE followed by Western blotting with the indicated antibodies. Western blots for actin serves as loading controls. (f,g) 14-3-3?-knockdown and vector-control cells, transfected with mCherry-a-tubulin and FLAG-NPM1 constructs, were stained with antibodies to Cep-170, co-stained with DAPI and followed by confocal microscopy (f) to determine the percentage of cells containing >2 centrosomes. The mean and standard deviation of three independent cells is plotted (g) p values were determined using a Student’s t-test (2 sample unequal variance) with p < 0.05. Scale bars indicate 5 µm. All the Western blots were run under the same experimental conditions and the full length blots are in Supplementary Fig. 6.

Fig 5: Loss of 14-3-3? causes centrosome amplification in human cells.(a–b) The protein (a) and mRNA (b) levels of the indicated gene products in 14-3-3?-knockdown and vector-control cells were determined as described. Actin and GAPDH served as loading controls. (c–d) Centrosomes of 14-3-3?-knockdown and vector-control cells were stained with antibodies specific to ?-Tubulin, Ninein and Cep170, or the cells were transfected with GFP-Centrin as indicated (c). Cells were co-stained with DAPI to visualize the nuclei and analyzed by confocal microscopy. Representative images are shown (c). The percentage of mitotic cells with more than 2 centrosomes or more than four centrioles was determined in three independent experiments (d). (e–f) HEK293 or U2OS cells were transfected with plasmids encoding EGFP-f (control) or EGFP-f and shRNA-14-3-3?. Cells were stained with anti-Cep-170 antibodies, to determine centrosome number. (g–i) Multiple spindle poles associated with supernumerary centrosomes were observed by transfecting the cells with mCherry-a-tubulin and GFP-Centrin, or transfection of mCherry-a-tubulin followed by immuno-staining for ?-Tubulin (g). The number of spindle poles was determined as indicated in the materials and methods (h–i). Note that all the centrosomes anchor the mitotic spindle. (j–l) The indicated cells were stained with antibodies to Cep-170 or transfected with GFP-Centrin followed by co-staining with DAPI and the number of centrosomes determined. The images shown are merged with DIC image. (k) The number of cells with more than 1 centrosome (for Cep-170) and (l) more than 2 centrioles (for Centrin) were determined in three independent experiments. (m) Osmium tetroxide stained 14-3-3?-knockdown and vector-control cells were visualized at 25000x magnification under scanning Electron Microscope. Centrioles are indicated by arrows. All the Western blots were run under the same experimental conditions and the full length blots are in Supplementary Fig. 6. In all the experiments the mean and standard error from at least three independent experiments were plotted, error bars denote standard error of mean and p-values are obtained using Student’s t test (2 sample unequal variance) and the asterisk (*) represents a p value <0.05. Original magnification 630X with 2X optical zoom. Scale bar indicate 10 µm, unless mentioned.