Table S1. Modifier screen to identify genes interacting with Tao-1^a

Fig. S1. The evolutionary relatedness of various Tao-1 homologues. (A) Graphic alignment of the amino acid sequence of Drosophila Tao-1 compared with its homologues in various other species including human, using intensity of greyscale to indicate similarity (black, amino acid identity, whereas lighter shades indicate decreasing degrees of amino acid similarity). A consensus identity histogram is shown on the top of each row with green peaks showing the regions of strongest identity. A phylogenetic tree (B) shows how Drosophila Tao-1 is related to homologues in various other species from worm to human.

Fig. S2. Tao-1 localises to microtubules. (A) Drosophila S2R⁺ cells were fixed and stained for microtubules (MT) as well as for a polyclonal antibodies against Tao-1 and examined by confocal microscopy 5 days after treatment with Tao-1 or control (lacZ) dsRNAs; settings were held constant to allow enable a direct comparison of fluorescence intensity in each case. Following Tao1 silencing, staining with Tao-1 antibody was significantly reduced. (B) Cells overexpressing RFP-Tao-1 were fixed and stained with polyclonal antibodies against Tao-1 (α-Tao, top right) and phosphorylated Tao-1 (α-P-Tao, bottom right) and examined as in A. A strong signal was seen in cells expressing the fusion protein (above that of the endogenous protein seen in the surrounding RFP-negative cells), confirming the specificity of the antibodies. (C) When GFP-Tao was expressed at low levels it could clearly be seen colocalised with a subset of microtubules (D). Endogenous Tao-1 was seen associated with the spindle during mitosis, whereas phosphorylated Tao was primarily peri-centrosomal. (E) GFP-Tao was also seen associated with the mitotic spindle.

Fig. S3. Tao-1 overexpression compromises microtubule regrowth. (A) RFP-Tao-1 overexpression impairs microtubule growth variably. In S2R⁺ cells, the phenotype is slightly milder than in S2 cells: transfected cultures were scored in triplicate (n=100) for reduced microtubule filaments compared to control cells, for each cell type. (B) Control S2 cells and S2 cells expressing RFP-Tao-1 are shown indicating that, in this insect cell type, over-expressed full-length RFP-Tao-1 causes drastic dismantling of the microtubule network and concomitant rounding compared with neighbouring untransfected controls. (C) To assess the effects of ectopic Tao-1 on microtubule re-growth, S2 cells stably expressing EB1-GFP and transiently overexpressing RFP-Tao-1 were used in a microtubule inhibitor wash-out experiment (microtubules are shown in green, RFP-Tao-1 in red). EB1-GFP comets were visualized (green) in adjacent control cells and cells overexpressing RFP-Tao-1 (red) in the presence of 0.5 M colcemid, and at 4 and 12 minutes after drug wash-out.

Fig. S4. Tao-1-knockdown-induced microtubule bundling precedes any visual effect on cell shape or the actin cytoskeleton. S2R⁺ cells treated with dsRNA targeting Tao1, Scar or lacZ for 2, 3 and 4 days were replated onto ConA-coated coverslips, and fixed and stained to visualize F-actin, microtubules and DNA. Both Scar and Tao-1 exhibited the usual spiky cell phenotype by day 4 (see text for details). (A) At days 2 and 3, however, Tao1 RNAi cells often exhibited microtubule bundles before the onset of visible changes to cell shape or the actin cortex (see arrow, Day 3 for an example). In the Scar RNAi experiment, the development of changes in microtubules and cell shape paralleled the loss of lamellipodial actin. (B) The percentage of cells with both a macroscopically normal actin cytoskeleton and clear evidence of a disturbed microtubule cytoskeleton was quantified in this time-course experiment. Error bars indicate s.d. based upon three independent measurements.

Fig. S5. The actin cortex is dispensable for Tao-1 to influence microtubule-based protrusion formation, and the plus-end protein EB1 genetically interacts with Tao-1. (A) GFP-EB1 S2 cells were transfected with RFP-Δ-423-900. 48 hours later, cells were replated onto concanavalin-A-coated glass coverslips. After allowing 30 minutes for the cells to adhere and spread, cells were live-imaged by microscopy with or without drug treatment. Images show a representative cell at various times after the addition of latrunculin B (0.5 M final) immediately following drug addition and 2 and 4 minutes later. The top image sequence shows a control cell developing microtubule spikes; the bottom panel shows an equivalent RFP-Δ-423-900-positive cell over time. (B) EB1 silencing alone has no effect on the shape of S2R⁺ cells, but leads to a significant rescue of the spiky phenotype in the Tao1 RNAi microtubule phenotype. Representative images are shown on the right of the quantitative graph.

Movie 1. Representative cell from Fig. 6A. Microtubule comet behaviour in EB1-GFP S2 control cell. See Fig. 6 legend for details.

Movie 2. Representative cell from Fig. 6B. Microtubule comet behaviour in a cell expressing a dominant-negative Tao1-null mutant.

Movie 3. Representative cell from Fig. 6C. Microtubule comet behaviour in a Tao1-knockdown cell.