We demonstrate that a Drosophila Golgi protein, Gorab, is present not only in the trans-Golgi but also in the centriole cartwheel where, complexed to Sas6, it is required for centriole duplication. In addition to centriole defects, flies lacking Gorab are uncoordinated due to defects in sensory cilia, which lose their nine-fold symmetry. We demonstrate the separation of centriole and Golgi functions of Drosophila Gorab in two ways: first, we have created Gorab variants that are unable to localize to trans-Golgi but can still rescue the centriole and cilia defects of gorab null flies; second, we show that expression of C-terminally tagged Gorab disrupts Golgi functions in cytokinesis of male meiosis, a dominant phenotype overcome by mutations preventing Golgi targeting. Our findings suggest that during animal evolution, a Golgi protein has arisen with a second, apparently independent, role in centriole duplication.
Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
We acknowledge P. Deák (Department of Genetics, University of Szeged, Szeged, Hungary) for his encouragement and support of L.K. and M. Pál (Department of Genetics, University of Szeged, Szeged, Hungary) for injection of CRISPR–Cas9 guide-RNAs for Gorab mutagenesis; and thank S. Chang, K. Oras, and A. Madich (Deparment of Genetics, University of Cambridge, Cambridge, UK) for injection of Gorab transgenes. We also greatly appreciate the advice of M. Richter and A. Fatalska in studies of protein–protein interactions. We thank T. Megraw (Department of Biomedical Sciences, Florida State University, Tallahassee, FL, USA) for GFP-rootletin flies, C. Gonzalez (Institute for Research in Biomedicine, The Barcelona Institute of Science and Technology, Barcelona, Spain) for YFP-centrobin flies, S. Munro (MRC Laboratory of Molecular Biology, Cambridge, UK) for anti-golgin antibodies, K. Raj (Radiation Effects Department, Public Health England, Didcot, UK) for the U2OSp53DD cell line, and J. Debski for advice in mass-spectrometry. D.M.G. is grateful for a Wellcome Investigator Award, which supported this work. The study was initiated with support from Cancer Research UK.
Integrated supplementary information
(a) D-Mel2 cells in metaphase were immunostained to reveal Gorab, dPLP (centrosome), and Golgin245 (trans-Golgi). Arrowheads indicate centrosomes. Experiment repeated 3 times with similar results.Main scale bar, 5 µm; inset scale bar, 0.5 µm.(b) Wing discs from larvae expressing N-terminally GFP-tagged Gorab were immunostained for dPLP (red) and Golgin245 (grey). Dashed lines highlight an interphase (left) and a mitotic cell (right). Arrowheads indicate centrosomes of mitotic cells. Experiement repeated 3 times with similar results.Main scale bar,5 µm.Inset scale bar,0.5µm.(c) Brains from larvae expressing N-terminally GFP-tagged Gorab immunostained for dPLP (red) and Golgin245 (grey). Dashed lines highlight an interphase (left) and a mitotic cell (right). Arrowheads indicate centrosomes of a mitotic neuroblast. Experiment repeated with similar result. Main scale bar,5 µm, Inset scale bar, 1µm. (d) Fluorescent micrographs of control and Gorab depleted D-Mel2 cells stained by DAPI (blue) and anti-Gorab (green) to demonstrate the specificity of Gorab antibody. Cells subjected to 4 times 4 days siRNA treatment with control or Gorab specific siRNAs followed by fixation in methanol and immunostaning. Experiment independently repeated twice with similar results. Scale bar: 10 µm.
Autoradiograms of GST-tagged Gorab’s interaction with the 35S-Methionine-labelled Sas6 protein fragments schematically illustrated in Fig 2b. The length of the Sas6 fragments are indicated. Experiments were repeated twice by different investigators with similar results.
(a)Autoradiogram of GST-Sas6 interaction with 35S-Methionine-labelled Gorab protein fragments, schematically illustrated in Fig 5b. The experiments were repeated once by a different investigator with similar result. (b) Sas6 and Gorab interaction in vivo. D-Mel2 cells were transiently transfected with myc-tagged Sas6 and GFP-tagged wild type or ΔSID Gorab. Samples were subjected to GFP-trap immunoprecipitation and western blot using anti-myc and anti-GFP antibodies. Experiments were repeated once by a different investigator with similar result. Anti-alpha tubulin is loading control.
Ciliated neurons from femoral chordotonal organs of flies expressing GFP-gorabV266P in a gorab1 mutant background immunostained with dPLP to reveal basal bodies and with phalloidin to visualize actin in scolopale rods. Experiment was repeated once with similar result. Arrowheads, GFP-Gorab at basal bodies. Scale bar, 10 µm.
(a) Fertility of wild type, gorab1mutant and rescued females. Wild type C-terminally GFP tagged gorab cDNA was expressed under the control of the ubiqutin promoter in gorab1 to rescue (Ubq-gorabwt-GFP, gorab1). Individual females were mated with wild type males and allowed to lay eggs at 25 °C for 6 days.Data points represent the number of progeny of individual females. Means ± s.e.m are shown for n=15 females per genotype. p values of two tailed unpaired t-tests are shown. p value in blue indicates significant difference (99% confidence interval). Experiment repeated once with similar result.(b) Climbing ability of wild type, gorab1mutant and rescued flies. Wild type C-terminally GFP tagged gorab cDNA was expressed under the control of the ubiqutin promoter in gorab1 to give rescue (Ubq-gorabwt-GFP, gorab1). Flies were raised at 29 °C and subjected to a climbing assay. Each data point represents the percentage of flies from a group of 15 flies. Means ± s.e.m are shown for N=3 independent experiments, n= 15 flies/genotype were investigated in each experiment. p values of two tailed unpaired t-tests are shown. p value in blue indicates significant difference (99% confidence interval).
(a) Centrosomal localization of wild type and A220P mutant GORAB. U-2 OS cells transiently transfected with constructs constitutively expressing N terminally GFP-tagged wild type (left panel) or A220P mutant human GORAB. Pericentrin (PCNT, red) stains centrosomes, Golgin-97 (grey) highlights trans-Golgi. Main scale bar,2 µm. Inset scale bar, 0.2 µm. Experiments repeated twice with similar results.(b)GORAB localization inside the centrioles.3D-SIM micrographs of U-2 OS centrosomes immunostained for Pericentrin (blue), SAS6 (red) and GORAB (green). Centrosomes of top (upper panel) and side view (lower panel) are not identical. Experiments repeated twice with similar results. Asterisk, site of procentriole formation where SAS6 is recruited. Arrowheads indicate SAS6, which is initially recruited to the proximal lumen of the cartwheel-less mother centriole and which moves to the site of procentriole formation45. Scale bar, 200 nm. (c) Synergistic effect of SAS6 and GORAB on loss of centrosomes. U-2OSp53DDcells subjected to 3 times 72h siRNA treatments against the indicated genes. Fixed cells were immunostained with a combination of CENP-J and gamma-Tubulin antibodies to reveal centrioles and centrosomes..Each data point represent the percentage of cells with the given centrosome number from an independent experiment (n=100 cells). Means ± s.e.m are shown for N=4 independent experiments (n=100 cells/experiment)p values of two tailed unpaired t-tests are shown(99% confidence interval). (d)Representative images of U-2OS cells subjected to siRNAs against GFP (control) or GORAB and treated with a combination of 4 µM aphidicolin and 1.5 mM hydroxyurea (A/HU). Anti-gamma-tubulin immunostaning reveals centrosomes (red) and DAPI staining (blue) reveals nuclei. Experiment repeated twice with similar results. Scale bar, 5 µm. (e) Quantification of centrosome numbers in U-2OS cell shown on d. Each data point represents the percentage of cells with a given number from an independent experiment (n=100 cells). Mean ± s.e.m are shown for N=3 independent experiments (n=100 cells/experiment).p values of two tailed unpaired t-tests are shown (99% confidence interval). (f) Anti-Gorab Western blot on GORAB depleted U-2 OSp53DD cell lysates. Cells were subjected to3 siRNA treatments, each for 72h, with control or GORAB specific siRNAs to check the specificity of the GORAB antibody (Atlas, #HPA027250). Subsequent Coomassie staining of the membrane (CBB) proves a loading control. Experiment independently repeated once with similar results.(g) Fluorescent micrographs of control and GORAB-depleted U-2 OSp53DD cells stained with DAPI (blue) and anti-GORAB (green) to demonstrate the specificity of GORAB antibody (Atlas, #HPA027250). Cells were subjected to3 siRNA treatments, each for 72h, with control or GORAB specific siRNAs followed by fixation in 4% formaldehyde and immunostaning. Experiment independently repeated once with similar results. Scale bar,5 µm.
Supplementary Figures 1–6
Supplementary Tables 1–3
Startle response assay of gorab1 mutant. Flies were placed in 3cm diameter Petri dishes and left for 1 h to acclimatize. Movies were recorded using a LifeCam Cinema HD (Microsoft) web camera