The Schizosaccharomyces pombe HIRA-like protein Hip1 is required for the periodic expression of histone genes and contributes to the function of complex centromeres

Abstract

HIRA-like (Hir) proteins are evolutionarily conserved and are implicated in the assembly of repressive chromatin. In Saccharomyces cerevisiae, Hir proteins contribute to the function of centromeres. However, S. cerevisiae has point centromeres that are structurally different from the complex centromeres of metazoans. In contrast, Schizosaccharomyces pombe has complex centromeres whose domain structure is conserved with that of human centromeres. Therefore, we examined the functions of the fission yeast Hir proteins Slm9 and the previously uncharacterised protein Hip1. Deletion of hip1(+) resulted in phenotypes that were similar to those described previously for slm9 Delta cells: a cell cycle delay, synthetic lethality with cdc25-22, and poor recovery from nitrogen starvation. However, while it has previously been shown that Slm9 is not required for the periodic expression of histone H2A, we found that loss of Hip1 led to derepression of core histone genes expression outside of S phase. Importantly, we found that deletion of either hip1(+) or slm9(+) resulted in increased rates of chromosome loss, increased sensitivity to spindle damage, and reduced transcriptional silencing in the outer centromeric repeats. Thus, S. pombe Hir proteins contribute to pericentromeric heterochromatin, and our data thus suggest that Hir proteins may be required for the function of metazoan centromeres.

FIG. 1.

Schematic comparison of the S. pombe Hir proteins Slm9 and Hip1 with human HIRA. The percentage of similarity between regions of Hip1 and Slm9 with the equivalent regions of human HIRA is shown. The conserved C-terminal domain is shaded black, and the B-domain (29) is represented with an open circle. WD repeat motifs are represented as grey boxes. Note that the motif in Hip1 that is represented with a dashed line exhibits significant homology with the equivalent WD repeat in human HIRA but falls below the threshold to be designated a WD repeat as defined by Pfam (4).

FIG. 2.

Deletion of hip1⁺ results in growth defects. (A) A hip1Δ strain (SW137) was transformed with the indicated plasmid and cultured on EMM agar at 30 or 37°C for 3 to 4 days. (B) Wild-type (SW5), hip1Δ (SW137), and slm9Δ (JK2246) cells were grown overnight at 36°C and then processed for microscopy. Mean cell length (micrometers) ± standard deviation is indicated. (C) hip1 interacts genetically with cdc25. A cdc25-22 hip1Δ strain containing pRep41F-Hip1 was grown to early log phase at 25°C. Incubation was continued in either the presence or absence of thiamine (2 μM). Mean cell length (micrometers) ± standard deviation is indicated. (D) hip1Δ cells are defective in recovery from nitrogen starvation. DNA content analysis of wild-type (SW5) and hip1Δ (SW137) cells. Cells were grown to log phase in EMM medium (+N) and shifted to nitrogen-free EMM medium for 16 h (−N). Cells were then shifted to EMM medium and harvested at the indicated intervals.

FIG. 3.

Level and localization of Hip1 are not affected by cell cycle progression. (A) cdc25-22 hip1-Pk cells were grown to early log phase in EMM medium at 25°C before being shifted to 34°C for 4 h to arrest cells at the G₂/M transition. Cells were rapidly cooled to 25°C and harvested at 20-min intervals. Cell cycle synchronicity was determined by counting the percentage of cells with a septum (top panel). Whole-cell extracts were prepared from each sample and subjected to Western blotting with anti-Pk and anti-Res1 antibodies. (Res1 levels are known to remain constant throughout the cell cycle.) (B) Hip1-Pk localizes to the nucleus. hip1-Pk (SW187) cells were grown to mid-log phase in YE5S medium at 30°C. Cells were stained with DAPI and anti-Pk antibodies and processed for indirect immunofluorescence.

FIG. 4.

Hip1 and Slm9 interact. hip1-Pk (SW159) (lanes 1 and 2) and hip1-Pk slm9-HA6H (SW155) (lanes 3and 4) cells were grown to mid-log phase. Whole-cell extracts were prepared and partially purified with Ni²⁺-agarose, and the precipitates were analyzed by Western blotting with anti-Pk and anti-HA antibodies.

FIG. 5.

hip1Δ and slm9Δ cells have different sensitivities to heat shock and UV. (A) Wild-type (SW5), hip1Δ (SW137), and slm9Δ (JK2246) strains were grown to log phase, plated onto YE5S agar, and exposed to UV irradiation (254 nm) at the indicated dose with a Stratalinker. (B) Wild-type (SW5), hip1Δ (SW137), and slm9Δ (JK2246) strains were grown to log phase at 30°C and then shifted to 48°C. Aliquots were taken at the indicated intervals, diluted in cold YE5S medium, and then plated onto YE5S agar. Colony numbers were determined after 4 days of incubation at 30°C.

FIG. 6.

Hip1 is required for repression of histone mRNA levels in G₁-arrested cells. Cells were grown to mid-log phase at 25°C (exp G₂) and then shifted to 34°C for 4 h (arrest G₁). The cells were then shifted back to 25°C, and samples were taken at the indicated time points. DNA content was analyzed by flow cytometry (A), and gene expression was monitored by Northern blotting with the indicated probes (B). The strains used were SW47 (cdc10-129) and SW168 (cdc10-129 hip1Δ).

FIG. 7.

Loss of Hip1 leads to constitutive expression of histone genes. (A) Wild-type (SW5) and hip1Δ (SW137) cells were grown to mid-log phase in YE5S medium at 30°C. Cells were synchronized in early G₂ by centrifugal elutriation, inoculated into fresh YE5S medium, and sampled at 20-min intervals as indicated. The synchronicity of the cultures was determined by counting the percentage of cells with a septum (top panels). RNA was prepared from each sample and analyzed by Northern blotting with the indicated probes. cdc2⁺ mRNA levels served as a loading control. (B) Samples from the first cell cycle (0 to 120 min) were analyzed side by side as described above, and mRNA levels relative to the loading control were determined with a Phosphorimager. (C) Samples from the first cell cycle (0 to 120 min) were analyzed by Northern blotting with the indicated probes.

FIG. 8.

hip1Δ and slm9Δ cells are sensitive to spindle damage. (A) Strains SW5 (wild type) SW137 (hip1Δ), JK2246 (slm9Δ), and SW152(slm9Δ hip1Δ) were grown to log phase in YE5S medium. Cultures were diluted to approximately 0.4 × 10⁷ cells/ml, subjected to fivefold serial dilutions, and spotted on YE5S agar or YE5S agar supplemented with TBZ at the indicated concentration. Plates without TBZ were incubated at 30°C for 3 days, while plates supplemented with TBZ were incubated at 30°C for 6 days. (B) The elongated cell morphology of slm9Δ and hip1Δ cells is independent of Mad2 and Bub1. Comparison of the morphology of wild-type (w.t.), slm9Δ, hip1Δ, mad2Δ slm9Δ, mad2Δ hip1Δ, bub1Δ slm9Δ, and bub1Δ hip1Δ cells is shown. Cultures were grown at 30°C, and the mean cell length (micrometers) ± standard deviation is indicated.

FIG. 9.

Loss of Slm9 or Hip1 impairs transcriptional silencing. (A) Schematic of cen1 showing the relative insertion sites of the arg3⁺ and ade6⁺ marker genes. (B) Silencing in the outer repeats. Cells containing the otrR1::ade6⁺ allele in combination with the appropriate deletions were grown to log phase, subjected to 10-fold serial dilutions, and spotted onto YE5S agar lacking adenine. Plates were incubated for 3 to 4 days at 30°C. Row 1, SW26; row 2, SW4; row 3, FY1181; row 4, SW151; and row 5, SW118. (C) Silencing in the central core (cnt1). Cells containing the cnt1::arg3⁺ allele were spotted onto EMM plates lacking arginine (−Arg) or supplemented with arginine (+Arg) and incubated at either 25 or 30°C as indicated. Strains: row 1, 972; row 2, 1645; row 3, 2221; row 4, 6243; row 5, 6246; row 6, 4462. (D) Telomeric silencing. Cells containing the tel1L::his3⁺ insertion were spotted onto EMM plates lacking histidine (−His) or supplemented with histidine (+His) and incubated at 30°C. The strains used were: row 1, SW4; row2, FY86; row 3, FY1862; row 4, SW189; and row 5, SW188. (E) Silencing at the mating type loci. Cells containing the mat3-M::ade6⁺ allele in combination with the appropriate deletions were spotted onto YE5S agar lacking adenine. The strains were: row 1, SW26; row 2, SW4; row 3, PG1672; row 4, SW150; and row 6, SW149.

FIG. 10.

Hip1 and Slm9 are not required for the localization of Cnp1 or Swi6. (A) Chromatin immunoprecipitation analysis. The association of histone H3 and Cnp1 with centromeric sequences was compared in wild-type (SW61), hip1Δ (SW147), slm9Δ (JK2246), and hip1Δ slm9Δ (SW153) backgrounds. The positions of the specific PCR products are indicated, and fbp served as a euchromatic control. (B) Cells expressing green fluorescent protein-Swi6 were grown to mid-log phase at 30°C and analyzed by fluorescence microscopy. The wild-type (FY2214), hip1Δ (SW193), and slm9Δ (SW194) strains were used. (C) The frequency of wild-type, hip1Δ, and slm9Δ cells with 1, 2, 3, 4, or 5 Swi6 signals is shown. More than 100 cells were examined for each background.