Supplementary Figure 3: Profiles of the centromere deletion screens and resultant survivor characterizations.

(a) The viabilities of total cells (blue) and centromere-deleted cells (magenta), as determined by their resistance to G418 and 5-fluoroorotic acid (see METHODS). Error bars represents s.e.m. of n = 6 biological replicates. loxP-cen1, loxP-cen2, and loxP-cen3, the strains used for deleting cen1, cen2, and cen3, respectively. (b) The relative ratios between NC formation and telomere fusion in the survivors obtained in each screen. (c) Cell growth of the wild-type, NC survivors, and revertant cells growing on YES media at 33°C. The generation times of each strain were as follows: wild-type, 2.1 h; cd1-39, 1.9 h; cd1-60, 2.0 h; cd2-163, 2.1 h; cd2-166, 2.2 h; cd3-385, 2.7 h; cd3-386, 3.0 h; and cd3-386-r1, 2.0 h. (d) A typical fluorescent microscopic image of Hoechst 33342-stained anaphase chromosomes (magenta). Chromosome III visualization was obtained by tandem tethering of the GFP-LacI protein (green). Chromosome III was segregated equally to the daughter nuclei in wild-type cells but not Δcen3-NC cells (cd3-385). Scale bar, 10 μm. (e) The frequency of chromosome I segregation failure in the defective cells displaying lagging chromosomes of the indicated strains (n > 50). Cells were cultured at 33°C. One locus in chromosome I was visualized by tandem tethering of the GFP-LacI protein. Error bars represents s.e.m. of triplicate cultures. One third of the lagging chromosomes observed in heterochromatin-abrogated Δclr4 cells were derived from chromosome I, irrespective of whether it harbored canonical cen1 or Δcen1-NC, suggesting that not every NC results in lagging chromosomes.