Global analysis of core histones reveals nucleosomal surfaces required for chromosome bi-orientation

Abstract

The attachment of sister kinetochores to microtubules from opposite spindle poles is essential for faithful chromosome segregation. Kinetochore assembly requires centromere-specific nucleosomes containing the histone H3 variant CenH3. However, the functional roles of the canonical histones (H2A, H2B, H3, and H4) in chromosome segregation remain elusive. Using a library of histone point mutants in Saccharomyces cerevisiae, 24 histone residues that conferred sensitivity to the microtubule-depolymerizing drugs thiabendazole (TBZ) and benomyl were identified. Twenty-three of these mutations were clustered at three spatially separated nucleosomal regions designated TBS-I, -II, and -III (TBZ/benomyl-sensitive regions I-III). Elevation of mono-polar attachment induced by prior nocodazole treatment was observed in H2A-I112A (TBS-I), H2A-E57A (TBS-II), and H4-L97A (TBS-III) cells. Severe impairment of the centromere localization of Sgo1, a key modulator of chromosome bi-orientation, occurred in H2A-I112A and H2A-E57A cells. In addition, the pericentromeric localization of Htz1, the histone H2A variant, was impaired in H4-L97A cells. These results suggest that the spatially separated nucleosomal regions, TBS-I and -II, are necessary for Sgo1-mediated chromosome bi-orientation and that TBS-III is required for Htz1 function.

Figure 1

A genetic screen for TBZ- and benomyl-sensitive mutants with the histone-GLibrary (Matsubara et al, 2007; Sakamoto et al, 2009). (A) Sensitivity to microtubule-depolymerizing agents was determined by dropping three-fold serial dilutions of histone point mutants on the FY406 or MSY748 background onto SC agar plates containing 25 μg/ml thiabendazole (TBZ) or 10 μg/ml benomyl diluted in 0.125% dimethyl sulphoxide (DMSO). The plates were incubated at 30 °C for 3 days. (B) The position of histone residues that conferred TBZ/benomyl sensitivity on the electrostatic surface of the nucleosome core (PDB ID: 1ID3; White et al, 2001) looking down from the DNA superhelix axis. The positions of each residue, which is visible or not directly visible, are indicated by solid and dashed lines, respectively. Classification of TBS-I, -II, and -III residues is listed in Table I. (C, D) Side (C) and top (D) view of the same structure shown in (B).

Figure 2

Histone H2A point mutants show chromosomal instability when released from mitotic arrest with nocodazole. (A) Nucleosome structure. Histones H2A, H2B, H3, H4, and DNA are yellow, red, blue, green, and white, respectively. (B, C) Enlarged views of the focused histone H2A C-terminal tail region (B) and the histone H3 residues located in proximity to H2A-I112 (C). Upper and lower panels represent the surface and cartoon models generated using the Pymol software application (see Materials and methods). (D) Multiple alignments of various histone H2A C-terminal sequences across species. Sc, Saccharomyces cerevisiae; Sp, Schizosaccharomyces pombe; Dm, Drosophila melanogaster; Xl, Xenopus laevis; Mm, Mus musculus; Hs, Homo sapiens. Red circles indicate the histone residues conferring TBZ/benomyl sensitivity. (E) Histone H2A point mutants in the C-terminal region show perturbation of cell-cycle progression when released from mitotic arrest with nocodazole (Noc). Samples were taken at the times indicated and analysed by flow cytometry as described in Materials and methods. Asyn, asynchronous cells. (F) The spindle assembly checkpoint in H2A-I112A and -L117A cells is functional. A deletion mutant of Mad2, a spindle assembly checkpoint protein, was used as a positive control. Samples were harvested at the indicated times and analysed by flow cytometry as described in Materials and methods. α-Factor was used to arrest cells in the G1 phase. (G) Experimental scheme for figure (H). Cells were arrested in G1 phase by the addition of α-factor for 3 h, and released from G1 arrest in the presence of nocodazole. (H) According to the scheme shown in (G), cells were lysed at the indicated times after release from G1 arrest and analysed by immunoblotting for Pds1-3HA and histone H3 (control).

Figure 3

The H2A-I112A mutation causes mono-polar attachment, particularly after nocodazole treatment. (A) A schematic representation of the tet operator (tetO)/TetR-GFP system with kinetochore-microtubule bi-polar (left) and mono-polar (right) attachment in budding yeast. GFP-fused TetR binds tetO sequences integrated into the pericentromeric regions of chromosome 5 (CEN5-GFP). Centromere stretching in metaphase is observed when cells are arrested by depletion of Cdc20, a mitotic APC/C activator (Tanaka et al, 2000). (B) Time course for the observation of chromosome missegregation. Cells were released from the mitotic arrest caused by depletion of Cdc20 with (+Noc) or without (NT, no treatment) additional nocodazole treatment for 2.5 h at 25 °C. Images of dividing cells were taken by time-lapse imaging. (C) Images shown are representative photographs merged with CEN5-GFP in cells in anaphase. The scale bar represents 1 μm. (D) The rates of chromosome missegregation in H2A-WT and -I112A cells. Chromosome missegregation was evaluated by the behaviour of CEN5-GFP dots in dividing cells. The rates are represented as the mean±the standard deviation. (E) Time course for the observation of kinetochore-microtubule attachment in metaphase after nocodazole treatment. Cells were arrested by Cdc20 depletion with or without nocodazole (+Noc or NT) for 2 h at 25 °C and released from the nocodazole block. After additional metaphase arrest by Cdc20 depletion for ∼2 h, images of metaphase-arrested cells were taken by time-lapse imaging. (F) Images are representative photographs merged with CEN5-GFP and CFP-Tub1 in H2A-WT and -I112A cells in metaphase. The cyan signal of CFP-Tub1 has been converted to red. The scale bar represents 1 μm. (G) The rate of mono-polar attachment in H2A-WT and -I112A cells. Kinetochore-microtubule attachment was evaluated by the movement of GFP-labelled CEN5 in metaphase-arrested cells. The rates were represented as mean with standard deviation.

Figure 4

The kinetochore localization of the CPC and Sgo1 is impaired in H2A-I112A cells. (A–D) H2A-WT and -I112A cells containing Ipl1-3HA (A), Bir1-13Myc (B), Sgo1-3HA (C), or Bub1-3HA (D) were grown in YPAD medium containing 15 μg/ml nocodazole for 3 h at 25 °C. Cells were fixed with 1% formaldehyde for 15 min and subjected to ChIP. Input DNA and DNA co-immunoprecipitated with the anti-Myc or anti-HA antibody (IP) were amplified with a primer set corresponding to CEN3 nucleotide sequences. The data are the average of two independent experiments. Error bars indicate the standard deviation. Dashed lines indicate the background level of the ChIP signal in an untagged strain. (E) The CPC (IPL1, SLI15, and BIR1) and SGO1 mRNA levels in nocodazole-treated H2A-WT and -I112A cells. The data are the average of two independent experiments. Error bars indicate the standard deviation. (F) The CPC (Ipl1, Sli15, and Bir1) and Sgo1 protein levels in nocodazole-treated H2A-WT and -I112A cells. (G–K) The localization of major kinetochore components was unchanged in H2A-I112A cells. The centromere-specific histone H3 variant, Cse4 (G); representative proteins of the inner kinetochore, Scm3 (H) and Mif2 (I); a representative protein of the outer kinetochore, Ctf3 (J); and a cohesin component, Scc1 (K), were analysed. Cells were subjected to ChIP as described in (A–D) with the exception that the cells were incubated in the presence of nocodazole for 1 h at 37 °C rather than at 25 °C. Since H2A-I112A cells are temperature sensitive (Supplementary Figure S6), the effect of the H2A-I112A point mutation is expected to be detectable at 37 °C by ChIP. (L) The kinetochore localization of Ipl1, a subunit of the CPC, was analysed in sgo1 cells. ChIP analysis was performed using WT and sgo1 cells containing Ipl1-3HA as described in (A–D). (M, N) The rates of missegregation (M) and mono-polar attachment (N) were increased in sgo1 cells. Under the same experimental conditions described in Figure 3B and E, the rates of missegregation and mono-polar attachment in WT and sgo1 cells were evaluated with or without prior treatment with nocodazole.

Figure 5

Sgo1 is a multi-copy suppressor of histones H2A and H3 point mutations in TBS-I and -II. (A) Illustration of an experimental approach to investigate the functional interaction between Sgo1 and TBS-I, -II, and -III residues (Figure 1B–D; Table I). (B) The sensitivity of histone H2A point mutants in the C-terminal region to 25 μg/ml TBZ and 10 μg/ml benomyl was determined using three-fold serial dilutions of cells. Each strain was transformed with a YEplac195 multi-copy vector (URA3 marker, 2 μ) containing the SGO1 gene. The plates were incubated for 3 days at 25 °C. (C–E) The sensitivity of representative TBS-I (C), TBS-II (D), and TBS-III (E) mutants to TBZ and benomyl was determined using three-fold serial dilutions of cells as described in (B).

Figure 6

The mitotic phenotypes of H2A-E57A (TBS-II mutant) and H4-L97A (TBS-III mutant) cells. (A) H2A-E57A and H4-L97A cells show perturbation of cell-cycle progression when released from mitotic arrest with nocodazole (+Noc). Samples were harvested at the times indicated and analysed by flow cytometry as described in Materials and methods. (B) The spindle assembly checkpoint in H2A-E57A and H4-L97A cells is functional. A deletion mutant of Mad2, a spindle assembly checkpoint protein, was used as a positive control (see Figure 2F). Samples were taken at the times indicated and analysed by flow cytometry as described in Materials and methods. (C) Immunoblotting of Pds1-3HA. Cells were subjected to immunoblotting as described in Figure 2H. (D) The rate of mono-polar attachment in H2A-E57A and H4-L97A cells. Kinetochore-microtubule attachment was evaluated by the behaviour of GFP-labelled CEN5 in metaphase-arrested cells. (E) The protein level of Sgo1 is reduced in nocodazole-treated H2A-E57A and H4-L97A cells. (F) The localization of Sgo1 is impaired in H2A-E57A cells and moderately impaired in H4-L97A cells. Cells were subjected to ChIP as described in Figure 4A–D. (G) Htz1 is a multi-copy suppressor of TBZ/benomyl sensitivity in H4-L97A cells but not in H2A-E57A or -I112A cells. The sensitivity of the indicated histone point mutants to 25 μg/ml TBZ and 10 μg/ml benomyl was determined using three-fold serial dilutions of cells. Each strain was transformed with a YEplac195 multi-copy vector (URA3 marker, 2 μ) containing the HTZ1 gene. The plates were incubated for 3 days at 25 °C. (H) The localization of Htz1 is severely impaired in H4-L97A cells, but not in H2A-I112A or -E57A cells. Cells were subjected to ChIP as described in Figure 4A–D. Input DNA and DNA co-immunoprecipitated with the anti-FLAG antibody (IP) were amplified with primer sets targeting the left or the right side of CEN3 nucleotide sequences.

Figure 7

Possible functions of TBS-I–III in the establishment of chromosome bi-orientation. The possible actions of TBS-I–III with Sgo1 or Htz1 in chromosome bi-orientation (bi-polar attachment) are described in Discussion. The nucleosome may be ‘the shugoshin (guardian) of shugoshin: Shugomaster’ to prevent its degradation by currently unknown proteases. The APC/C proteosomal complex is one candidate (Salic et al, 2004; Karamysheva et al, 2009).