Fig. 2. Bub1 localization at the kinetochores in wild-type and klp5 mutants. (A and B) Wild-type (A) or klp5 mutants (B), in which the endogenous bub1⁺ gene was tagged with GFP, were grown and Bub1 localization during the cell cycle was examined (in A: 1, interphase; 2 and 3, early mitosis; 4, anaphase; in B: 1 and 2, early mitosis; 3–5, prometaphase/metaphase). Nuclear DNA (DAPI), Bub1–GFP and merged images (DAPI, red; Bub1–GFP, green) are shown. The bar indicates 10 µm.

Fig. 3. Time-lapse images of Bub1–GFP during mitosis. (A and B) Live analysis of Bub1–GFP during mitosis. Time-lapse imaging of Bub1– GFP was performed using a conventional fluorescence microscope with 30 s intervals in wild-type (A) or klp5 mutants (B). The period during which Bub1–GFP at the kinetochores was visible is marked with vertical lines, 2.5 min in wild-type and 5 min in klp5 mutants. Multiple Bub1–GFP dots in klp5 mutants are shown with arrowheads. The numbers in the top right corner in each image represent minutes after the start of live imaging. The bar indicates 10 µm.

Fig. 8. A role for Klp5/Klp6 and Alp14/Dis1 in the formation of bipolar mitotic spindles. Alp14 and Dis1 (shown as Alp14 in the figure) localize to both the mitotic spindles [shown by filaments consisting of tubulin dimers (white and red circles)] and the mitotic kinetochores (closed black circles). Klp5 and Klp6 (shown as Klp) also localize to the mitotic kinetochores. In their absence, the major defect is failure in the attachment of the kinetochores to which Bub1 and Mad2 localize. Furthermore, at least Klp5 and Klp6 have an additional role in generation of tension at the kinetochores upon attachment. In the absence of Klp5 and Klp6, the kinetochores fail to produce tension, as the spindles tend only to polymerize without the poleward force (depicted by wavy spindles). These tension-less kinetochores recruit Bub1, but not Mad2.

Fig. 4. Localization of Bub1 and Mad2 in klp5 mutants. (A) Classification of Bub1 and Mad2 localization patterns in nda3-311 or mitotic klp5 mutants. nda-311 mutants incubated at 18°C for 10 h (upper two rows) or exponentially growing klp5 mutants (lower three rows), in which both strains contained Bub1-HA and Mad2–YFP, were fixed with formaldehyde and processed for immunofluorescence microscopy. Cells which displayed clear nuclear dots for Bub1 and/or Mad2 were examined, and localization patterns were scored (type I, single dot and co-localization; type II, multiple dots and co-localization; type III multiple Bub1 dots and a single Mad2 dot, which co-localized with one of the Bub1 dots). The bar indicates 10 µm. (B) Percentage of three types of mitotic cells. One hundred mitotic cells which displayed Bub1 and/or Mad2 dots were examined in klp5 (closed columns) or nda3-311 mutant cells (open columns) and scored for individual patterns.

Fig. 6. Enhanced stability of microtubules in the absence of Klp5 or Klp6. (A) Experimental designs. Wild-type, klp5 and klp6 mutant cells containing GFP-atb2⁺ (encoding α2-tubulin), were grown at 28°C, incubated on ice for 20 min then shifted up to 28°C and incubated for another 20 min. Samples were taken at the indicated times, and microtubule structures were observed by fluorescence microscopy. Time points at which representative images are shown in (B) and (C) are marked with an arrow and arrowhead, respectively. (B) Depolymerization of microtubules by cold shock. The percentage of cells that contain long microtubules (>1 µm in length) is plotted at each time point. Microtubule structures in wild-type (blue), klp5 (red) and klp6 (yellow) mutants after 5 min on ice are shown below. (C) Recovery of microtubule structures upon temperature shift-up. Images of microtubules after 1 min shift-up at 28°C after cold shock treatment are shown below as in (B). The bar indicates 10 µm.

Fig. 5. Independence of Klp5 and Klp6 function from cohesin and the aurora kinase. (A) Sister chromatid cohesion in klp5 mutants. klp5 mutants containing centromere-marking Cen2–GFP were fixed and processed for immunofluorescence microscopy with anti-Sad1 antibody. Nuclear DNA was stained with DAPI. Merged images are shown on the bottom (Cen2–GFP, green; Sad1, red; DNA, blue). Two representative mitotic cells before chromosome segregation are shown. Note that Cen2–GFP overlaps or is proximal to one of the two SPBs. (B) Klp5 localization at the kinetochores in the cohesin mutant. The temperature-sensitive rad21-K1 mutant containing Klp5–GFP was incubated for 6 h at 36°C and localization of Klp5–GFP was examined. Anti-Sad1 antibody was used to visualize the SPBs. Merged images are shown on the right (Klp5–GFP, green; anti-Sad1, red; DAPI, blue). Arrowheads mark Klp5 dots at the kinetochores. (C) Localization of the aurora kinase (Ark1) in klp5 mutants. Ark1 was tagged with GFP in klp5 mutants and its localization during the cell cycle was examined (1 and 2, early mitosis; 3, prometaphase/metaphase; 4, anaphase; 5, late anaphase). Nuclear DNA was stained with DAPI. Merged images are shown on the right (Ark1–GFP, green; DAPI, red). The bar indicates 10 µm.

Fig. 7. Klp5/Klp6 and Dis1/Alp14 MAPs share an essential function in mitotic progression. (A–C) Defective mitotic phenotypes of alp14dis1, alp14klp5 or dis1klp5 mutants. Cells that were deleted simultaneously for alp14⁺ and dis1⁺, alp14⁺ and klp5⁺ or dis1⁺ and klp5⁺, but kept viable by episomal multicopy plasmids carrying the dis1⁺ or alp14⁺ gene were starved of nitrogen for 12 h, washed, resuspended in fresh rich medium and incubated further. At each time point, aliquots were taken for immunofluorescence microscopy with anti-tubulin antibody. Nuclear DNA was stained with DAPI. The frequency of cells that displayed defects in chromosome segregation was scored [(A), at least 100 cells were counted at each time point]. Representatives of cells displaying mis-segregation defects are shown in (B) (dis1klp5) and (C) (alp14dis1). DAPI (left), anti-tubulin (middle) and merged images (right, green for tubulin and red for DAPI) are shown. (D) Kinetochore localization of Alp14 in klp5 mutants. Alp14–GFP signals are shown in mitotic klp5 mutants. Images were taken from live time-lapse movies (0, 150 and 200 s). Arrows (red) and arrowheads (white) show the SPBs and the centromeres, respectively. (E) Localization of Klp5–GFP in alp14 mutants. Wild-type and alp14 deletion strains containing Klp5–GFP were shifted up to 36°C and samples were taken hourly. The percentage of cells that contain Klp5–GFP at the spindle and the kinetochore was determined. The bar indicates 10 µm in (B), (C) and (E), and 5 µm in (D).

Fig. 1. Mad2 localization at the kinetochores in klp5 mutants and enhanced chromosome loss in klp5mad2 double mutants. (A and B) Localization pattern of Mad2 and Cut12 during the cell cycle. Wild-type (A) or klp5 mutants (B), in which Mad2 and Cut12 (the SPB marker) were tagged with YFP and CFP, respectively, under their endogenous promoters, were grown and the localization patterns of these two proteins were examined. Nuclear DNA was stained with DAPI. In wild-type cells, only early mitotic cells show Mad2 at the kinetochores, whilst in klp5 mutants, Mad2 localizes to the kinetochores for a prolonged period. Representative images in different cell cycle stages are shown (1 and 2; early mitosis; 3–6, prometaphase/metaphase; 7, post-anaphase). Merged images are shown on the right (Cut12–CFP, red; Mad2–YFP, green). The bar indicates 10 µm. (C) Distribution of the spindle length in mitotic cells containing Mad2 dots. The distance between two Cut12 signals (equivalent to the length of the mitotic spindles) was measured in a total 40 or 120 mitotic cells in wild-type (left, blue columns) or klp5 mutants (right, red), respectively, which contain Mad2 at the kinetochores. (D) Minichromosome loss in klp5mad2 mutants. klp5 (left) or klp5mad2 (right) mutant cells carrying minichromosomes, which had been grown in minimal medium without adenine (selective conditions for minichromosomes), were plated on rich media plates and incubated at 30°C for 4 days. Colonies of adenine auxotrophs were red.