Mitotic Phosphorylation of Swi6/HP1 Regulates Its Chromatin Binding and Chromosome Segregation

Abstract

In eukaryotic cells, heterochromatin assembly is critical for chromosome segregation and transcriptional gene silencing. Heterochromatin protein 1 (HP1) is a conserved chromosomal protein that plays an important role in heterochromatin assembly. We have previously shown that mammalian HP1α and Schizosaccharomyces pombe Swi6 are phosphorylated by casein kinase II (CK2) and that this phosphorylation is essential for their function in heterochromatin assembly. In addition to CK2-mediated phosphorylation, several studies have shown that HP1 proteins undergo additional phosphorylation during mitosis. However, functional significance of the mitotic phosphorylation of HP1 remains unclear. Here, we identified mitotic phosphorylation sites within fission yeast Swi6 and showed that this phosphorylation is involved in chromosome segregation. Using an Escherichia coli co-expression system, we showed that Swi6 is phosphorylated by Ark1, a solo Aurora kinase in S. pombe, and mutational analyses revealed that serine residues in the conserved N-terminal region of Swi6 are the primary targets of Ark1. By expressing mutant Swi6, we confirmed that these serine residues are phosphorylated during mitosis in vivo. Although non-phosphorylatable or phosphomimic mutations in Swi6 had little effect on heterochromatic silencing, they caused defects in early chromosome segregation and modulated the temperature-sensitive growth of mutant cells for chromosome passenger complex components. These results suggest that the Ark1-mediated mitotic phosphorylation of Swi6 is involved in chromosome segregation during mitosis and implicates a conserved regulatory role for the mitotic phosphorylation of HP1 proteins.

Keywords: fission yeast; heterochromatin; heterochromatin protein 1; mitosis; phosphorylation.

FIGURE 1

Swi6 is hyperphosphorylated during mitosis. (A) Schematic of the procedure used to prepare cells in the asynchronous state, in the G2/M‐arrested state, or in the M phase using Schizosaccharomyces pombe cdc25‐22 mutant strains. (B) Phosphorylation patterns of Swi6 in asynchronously growing, G2/M‐arrested, and M‐phase cells. The cdc25‐22 cells were synchronized at the G2/M transition through incubation for 4 h at the restrictive temperature (36°C) and subsequently released to the permissive temperature (25°C). G2/M‐arrested cells were harvested immediately before release and M‐phase cells harvested 40 min thereafter. Whole‐cell lysates were resolved via standard (top, 10%) or Phos‐tag (bottom, 8%, 30 μM Phos‐tag) sodium dodecyl sulfate‐polyacrylamide gel electrophoresis and analyzed using immunoblotting with anti‐Swi6 antibody. Shrimp alkaline phosphatase (SAP)‐treated samples were used as unphosphorylated (or less phosphorylated) controls. Asterisks indicate the Swi6 species that exhibited an altered pattern during mitosis. (C) Time‐course experiments showing the dynamics of Swi6 phosphorylation. The cdc25‐22 cells were synchronized at 36°C and released via a temperature shift at 25°C. The cells were harvested every 20 min after release to prepare protein samples, and the Swi6 phosphorylation patterns examined by using standard and Phos‐tag gels as in (B). (D) The septation index is shown along with a schematic illustration of corresponding phases of the cell cycle. The septation peak roughly coincides with the S phase. All experiments were performed at least twice, and representative data are shown.

FIGURE 2

Swi6 is phosphorylated by Aurora kinase. (A) Schematic representation of the full‐length Swi6 and constructs for each of its domains; CD, chromodomain; CSD, chromoshadow domain; E, glutamic acid‐rich region; H, hinge region; N, N‐terminal domain. (B) Swi6 proteins produced in control or Ark1‐co‐expressed Escherichia coli cells were purified, resolved through standard or Phos‐tag polyacrylamide gel electrophoresis (30 μM), and visualized via CBB staining. (C) The Swi6 protein was divided into five parts as shown in (A), and each domain (except the E region) was produced in E. coli cells with or without Ark1 and analyzed as in (B). The arrowhead indicates peptides corresponding to the H with Ark1‐mediated phosphorylation. The asterisk indicates unrelated peptides that were also detected in the Ark1‐untreated control sample. (D) Recombinant proteins for the Swi6 N‐terminus and CD (Swi6‐NCD), as well as the Swi6‐NCD with a 1–30‐amino acid deletion (Swi6‐NCD_∆NN), were produced in E. coli cells with or without Ark1 and analyzed as in (B). (E) The E. coli co‐expression assay using a recombinant protein containing the Swi6 N‐terminus and CD (Swi6‐NCD). Wild‐type (WT) Swi6‐NCD and mutants with alanine substitutions at S12, S13, S8, S18, or both S12 and S13 were produced in E. coli cells with Ark1 and analyzed as in (B). (F) The E. coli co‐expression assay using recombinant protein containing the Swi6 hinge region (Swi6‐hinge). WT Swi6‐hinge and mutants with alanine substitutions at S142, S159, S162, S147, S165, S224 or S246 were produced in E. coli cells with Ark1 and analyzed as in (B). The arrowhead indicates peptides corresponding to the H with Ark1‐mediated phosphorylation. The asterisk indicates unrelated peptides that were also detected in the Ark1‐untreated control sample. (G) WT Swi6 and mutants with alanine substitutions at S12 and S13 (S12,13), T165, or S12,13 combined with T165 (S12,13, T165) were produced in E. coli cells with Ark1 and analyzed as in (B). (H) Schematic representation of full‐length Swi6 and the residues subject to Ark1‐mediated phosphorylation in the E. coli co‐expression system. The mitotically phosphorylated residues are shown in bold letters. All experiments were performed at least twice, and representative results are shown.

FIGURE 3

Conserved serine residues in the N‐terminal region of Swi6 are phosphorylated during mitosis in vivo. (A) Time‐course experiments showing the dynamics of the phosphorylation patterns of Swi6^S12,13A. The Schizosaccharomyces pombe cdc25‐22 cells expressing Swi6^S12,13A were synchronized at the G2/M transition through incubation for 4 h at the restrictive temperature (36°C) and subsequentlyreleased to the permissive temperature (25°C). The cells were harvested every 20 min after release to prepare protein samples, and the phosphorylation patterns examined by using standard and Phos‐tag gels. Whole cell lysates were resolved using standard (top, 10%) or Phos‐tag (bottom, 8%, 30 μM Phos‐tag) sodium dodecyl sulfate‐polyacrylamide gel electrophoresis and analyzed via immunoblotting with anti‐Swi6 antibody. (B) The septation index of cdc25‐22 swi6 ^S12,13A cells is shown along with a schematic illustration of corresponding phases of the cell cycle. (C) Schematic representation of full‐length Swi6 and the chimeric Swi6 containing the hinge region of human HP1β (mini‐swi6β). The positions of casein kinase II‐phosphorylatable serine residues are indicated by the black arrowheads. (D) Phosphorylation patterns of mini‐Swi6β in asynchronously growing, G2‐arrested, and M‐phase cells. The cdc25‐22 cells expressing mini‐Swi6β were analyzed. (E) Immunoblotting analysis of cells expressing mini‐Swi6β with amino acid substitutions. (F) Time‐course experiments showing the dynamics of miniSwi6β phosphorylation. The cdc25‐22 cells expressing miniSwi6β were synchronized at 36°C and released via a temperature shift at 25°C. The cells were harvested every 20 min after release to prepare protein samples, and the miniSwi6β phosphorylation patterns examined by using standard and Phos‐tag gels as in (B). (G) The septation index of cdc25‐22 mini‐swi6β cells is shown along with a schematic illustration of corresponding phases of the cell cycle. (H) Time‐course experiments showing the dynamics of miniSwi6β phosphorylation. The cdc25‐22 cells expressing miniSwi6β^S12,13A were synchronized at 36°C and released via a temperature shift at 25°C. The cells were harvested every 20 min after release to prepare protein samples, and the miniSwi6β^S12,13A phosphorylation patterns examined by using standard and Phos‐tag gels as in (B). (I) The septation index of cdc25‐22 mini‐swi6β ^S12,13A cells is shown along with a schematic illustration of corresponding phases of the cell cycle. All experiments were performed at least twice, and representative results are shown.

FIGURE 4

Unphosphorylatable Swi6 mutant enhances heterochromatin silencing. (A) Amino acid sequence of the Swi6 N‐terminal region (1–80 amino acids). Amino acids are shown, with acidic and basic amino acids indicated in orange and blue, respectively. Serine residues that could be phosphorylated by Ark1 and casein kinase II are shown in green. To replace the wild‐type swi6 ⁺ allele with the mutant swi6 allele, a BamHI site was introduced immediately after the ATG codon; the position of two additional amino acids, glycine and serine, is indicated by an asterisk (*). (B) Schematic diagram of full‐length Swi6 and Swi6 mutants with N‐terminal deletion (∆2–65 or ∆2–31) or amino acid substitutions (S12,13A or S12,13E). (C) Immunoblotting analysis of cells expressing wild‐type (WT) or mutant Swi6 proteins. Immunoblotting with an anti‐α‐tubulin antibody is shown as a loading control. (D) Diagram of the right side of centromere 1 (right) in Schizosaccharomyces pombe. The position of the silencing reporter gene, otr1R::Ura4 ⁺, is shown. (E) Heterochromatic silencing assays of WT and mutant swi6 strains. The silencing status at the centromeric otr1R::Ura4 ⁺ was evaluated. Tenfold serial dilutions of the indicated strains were spotted onto nonselective medium (N/S), minimal medium without uracil (–Ura), and minimal medium containing 5‐fluoroorotic acid (5‐FOA) (left). Expression levels of the ura4 ⁺ silencing reporter were evaluated via quantitative reverse‐transcription PCR (qRT‐PCR) analyses (right). Means and standard deviations of at least three independent experiments are shown. Statistical significance was determined using ANOVA. *p < 0.05. (F and G) Time‐course chromatin immunoprecipitation (ChIP) analysis of Swi6 (F) or Swi6^S12,13A (G) associated with pericentromeric dg regions, relative to act1 ⁺. Immunoprecipitated DNA was subjected to quantitative PCR analysis. Means and standard deviations of at least three independent experiments are shown.

FIGURE 5

Expression of Swi6 mutant defective in N‐terminal phosphorylation results in chromosome segregation defects at metaphase/early anaphase but not at late anaphase. (A and B) Analysis of chromosome segregation defects and lagging chromosomes in the swi6 mutants. EGFP‐tubulin (green); DNA was stained with DAPI (red). Scale bar, 3 μm. Chromosome segregation defects at metaphase/early anaphase (A) and the lagging chromosome frequency at late anaphase (B) in the mutants. The number of cells with deformed nuclei (A) or lagging chromosomes (B) is shown on the right of each representative image along with the number of cells examined. (C) Spotting assay to examine thiabendazole (TBZ) sensitivity for cells expressing wild‐type (WT) or each of the mutant Swi6 proteins. The clr4∆ cells were used as a control.

FIGURE 6

Swi6 N‐terminal phosphorylation is functionally linked to the chromosome passenger complex. (A–C) Serial dilutions of the indicated strains were spotted onto YEA plates and incubated at the indicated temperatures.

FIGURE 7

Ark1‐mediated phosphorylation of Swi6 reduces its DNA‐binding activity. (A) Wild‐type (WT) and mutant Swi6 (Swi6^∆2–17) proteins were purified, resolved using standard polyacrylamide gel electrophoresis (PAGE), and visualized by CBB staining. (B) Representative results of the electrophoretic mobility shift assays (EMSAs) performed with the WT and mutant Swi6. Different concentrations of the Swi6, from 0 to 20 μM (0.6‐fold dilutions), were incubated with 193‐bp 601 DNA. (C) Control and Ark1‐ or casein kinase II (CK2)‐phosphorylated Swi6 proteins were purified, separated with standard or Phos‐tag PAGE, and visualized via CBB staining. Shrimp alkaline phosphatase (SAP)‐treated samples were used as unphosphorylated (or less phosphorylated) controls. (D) Representative results of the EMSAs performed with control and Ark1‐ or CK2‐phosphorylated Swi6. (E) Quantification of the EMSAs performed with control and Ark1‐ or CK2‐phosphorylated Swi6. The bound DNA fractions were estimated from the intensity of the unbound DNA (1‐unbound fraction), and plotted against the Swi6 concentration. All EMSA experiments were repeated at least twice and representative results are shown.