Schizosaccharomyces pombe Hat1 (Kat1) is associated with Mis16 and is required for telomeric silencing

Abstract

The Hat1 histone acetyltransferase has been implicated in the acetylation of histone H4 during chromatin assembly. In this study, we have characterized the Hat1 complex from the fission yeast Schizosaccharomyces pombe and have examined its role in telomeric silencing. Hat1 is found associated with the RbAp46 homologue Mis16, an essential protein. The Hat1 complex acetylates lysines 5 and 12 of histone H4, the sites that are acetylated in newly synthesized H4 in a wide range of eukaryotes. Deletion of hat1 in S. pombe is itself sufficient to cause the loss of silencing at telomeres. This is in contrast to results obtained with an S. cerevisiae hat1Δ strain, which must also carry mutations of specific acetylatable lysines in the H3 tail domain for loss of telomeric silencing to occur. Notably, deletion of hat1 from S. pombe resulted in an increase of acetylation of histone H4 in subtelomeric chromatin, concomitant with derepression of this region. A similar loss of telomeric silencing was also observed after growing cells in the presence of the deacetylase inhibitor trichostatin A. However, deleting hat1 did not cause loss of silencing at centromeres or the silent mating type locus. These results point to a direct link between Hat1, H4 acetylation, and the establishment of repressed telomeric chromatin in fission yeast.

Fig 1

Mis16 copurifies with Hat1 during tandem affinity purification. Hat1 was affinity purified from KTP1 (Hat1-TAP) S. pombe cells. Isolated proteins were subjected to SDS-PAGE and analyzed using a LCQ Deca ion trap mass spectrometer. Apart from typical S. pombe contaminants also observed using a non-TAP strain (GAPDH [glyceraldehyde-3-phosphate dehydrogenase] and translation elongation factor 1 alpha 1), the only protein found in addition to Hat1 was Mis16 (14 peptides detected). Mis16 is an S. pombe orthologue of p46/Hat2.

Fig 2

Acetylation of H4 N-terminal peptides by S. pombe Hat1 in vitro. Unacetylated (UN) and K5/K12-diacetylated (5/12) H4 N-terminal peptides were incubated in vitro for 30 min at 37°C with S. pombe Hat1 and [³H]acetyl-CoA. Reactions were also performed without added peptide (NP) and with the unacetylated peptide minus Hat1 (UN-Hat1). Results are expressed as a percentage of radioactivity incorporated into the unacetylated peptide.

Fig 3

Acetylation of recombinant H4 by S. pombe Hat1. Recombinant H4 was incubated with Hat1 and unlabeled acetyl-CoA. Proteins from the reaction were resolved by electrophoresis and analyzed by Western blotting using antibodies that recognize total H4 (A) or H4 acetylated at K5 and/or K12 (B), K5 (C), K12 (D), or K8 (E). Acetylation at K16 was not detected above background (not shown).

Fig 4

hat1 deletion results in the loss of telomeric silencing in S. pombe. (A) Wild-type and experimental yeast strains were cultured on YEA plates in the presence (5FOA) or absence (YEA) of 5-FOA. Spot cultures represent 5-fold dilutions. Cells were grown for two (YEA) or three (5-FOA) days at 30°C. FY1872, ura4-telomeric marker (tel); KTP36, ura4-tel hat1Δ; KTP24, ura4 disrupted, hat1Δ; LBP6, hat1Δ. (B) Wild-type and experimental yeast strains were cultured on EMMG plates in the presence (EMMG) or absence (EMMG−ura) of added uracil. Spot cultures represent 5-fold dilutions. Cells were grown for two (EMMG) or three (EMMG−ura) days at 30°C. FY1872, ura4-tel; KTP36, ura4-tel hat1Δ; KTP24, ura4 disrupted, hat1Δ; LBP6, hat1Δ.

Fig 5

hat1 deletion does not result in the loss of centromeric silencing. Wild-type and experimental yeast strains were cultured on YEA plates in the presence (5FOA) or absence (YEA) of 5-FOA. Spot cultures represent 5-fold dilutions. Cells were grown for two (YEA) or three (5-FOA) days at 30°C. FY336, ura4-centromeric central core marker; KTP33, ura4-centromeric central core marker, hat1Δ; KTP24, ura4 disrupted, hat1Δ; LBP6, hat1Δ; FY496, ura4-centromeric inner most repeat marker; KTP29, ura4-centromeric inner most repeat marker, hat1Δ; FY648, ura4-centromeric outer repeat marker; KTP30, ura4-centromeric outer repeat marker, hat1Δ.

Fig 6

hat1 deletion results in an increase in H4 acetylation at the subtelomeric ura4 DNA sequence relative to that in the wild-type strain. Chromatin immunoprecipitation assays were performed on the FY1872 (ura4-telomeric marker) and KTP36 (ura4-telomeric marker, hat1Δ) strains using antibodies against H4 acetylated at K12 (H4-K12ac) (A) and H4 acetylated at K5 and/or K12 (H4-K5/K12ac) (B). Control immunoprecipitations were performed with rabbit IgG (A) or rabbit normal immune serum (RNIS) (B). Reverse transcription-PCRs (RT-PCRs) were performed using primers specific for ura4 and fus1. The average net C_T difference between the hat1Δ and wild-type strains was used to calculate the change (enrichment) of immunoprecipitated acetylated histones at the ura4 or fus1 DNA sequence in the hat1Δ strain relative to the wild type.

Fig 7

hat1 deletion results in an increase in H4-K8/K16 acetylation at the subtelomeric ura4 DNA sequence. Chromatin immunoprecipitation assays were performed on the FY1872 (ura4-telomeric marker) and KTP36 (ura4-telomeric marker, hat1Δ) strains using antibodies against H4 acetylated at K8 and/or K16. Results were analyzed as for Fig. 6.

Fig 8

Treatment with trichostatin A results in the loss of telomeric silencing. Wild-type (972) and telomeric marker (FY1872) S. pombe strains were cultured in medium containing (+TSA) or lacking (−TSA) trichostatin A and allowed to grow for ∼3 generations at 30°C. Cells were washed and then grown at 30°C in the absence of TSA on EMMG (2 to 3 days), 5-FOA (2 days), and EMMG-ura (4 days). Spot cultures represent 2.5-fold serial dilutions.