Histone H3-K56 acetylation is catalyzed by histone chaperone-dependent complexes

Abstract

Acetylation of histone H3 on lysine 56 occurs during mitotic and meiotic S phase in fungal species. This acetylation blocks a direct electrostatic interaction between histone H3 and nucleosomal DNA, and the absence of this modification is associated with extreme sensitivity to genotoxic agents. We show here that H3-K56 acetylation is catalyzed when Rtt109, a protein that lacks significant homology to known acetyltransferases, forms an active complex with either of two histone binding proteins, Asf1 or Vps75. Rtt109 binds to both these cofactors, but not to histones alone, forming enzyme complexes with kinetic parameters similar to those of known histone acetyltransferase (HAT) enzymes. Therefore, H3-K56 acetylation is catalyzed by a previously unknown mechanism that requires a complex of two proteins: Rtt109 and a histone chaperone. Additionally, these complexes are functionally distinct, with the Rtt109/Asf1 complex, but not the Rtt109/Vps75 complex, being critical for resistance to genotoxic agents.

Figure 1. Detection and reconstitution of H3-K56 acetyltransferases

A. Asf1-dependent acetylation of histone H3-K56 in wild-type cell extracts. HAT assays were performed using 5 μg of wild-type yeast whole cell extract (WCE) proteins (lanes 1 and 2) or 2 μl of a Poros Q fraction from the extract (lanes 3–6). 6 pmol of chicken histones H3/H4 or Asf1/H3/H4 complexes were added, with or without 25μM AcCoA as indicated. The products were analyzed by immunoblotting with the indicated antibodies. B. Analysis of mutant cell lysates and extracts. (left) rtt109 and asf1 mutant cells lack H3-K56ac. Whole cell alkaline lysis extracts of the indicated mutant strains were analyzed by immunoblotting. (right) rtt109 mutant cell extracts lack HAT activity. HAT assays were performed using 6 pmol Asf1/H3/H4 and 5 μg of whole cell extracts of the indicated mutant strains. C. Purified recombinant His6-Rtt109 (lane 1), His6-Vps75 (lane2), and Rtt109/His6-Vps75 proteins were analyzed on a 15% SDS-PAGE gel and stained with Coomassie Blue R250. D. HAT activity of recombinant proteins. 0.3 pmol of recombinant Rtt109 and/or Vps75, 6 pmol of yeast H3/H4, and 3 pmol of Asf1 were assayed as indicated. E. Asf1 and Vps75 but not CAF-1 stimulate Rtt109 activity. 0.9 pmol of Asf1, Vps75 or CAF-1 and the indicated amount of Rtt109 with 6 pmol of yeast H3/H4 were used in the HAT assay.

Figure 2. Protein interactions that govern H3-K56 acetylation

A–E. Vps75 is a histone binding protein. 1 μg of the following proteins were incubated on ice for 1hr and separated on a 5 ml 15–35 % glycerol gradient centrifuged at 49 Krpm (225 K x g) for 24 hrs at 4°C: A: Vps75, B: yeast histones H3/H4, C: Vps75 + H3/H4, D: Rtt109, E: Rtt109 + H3/H4. 350 μl of each fraction were TCA-precipitated and analyzed on silver-stained 15% SDS-PAGE gels. F. Rtt109 binds Asf1. 40 pmol (1.2 μg) Asf1-FLAG protein was incubated with 20 pmol of the indicated His6-recombinant proteins and then precipitated with anti-FLAG antibody beads. Washed beads were analyzed on a 15% SDS-PAGE gel stained with Coomassie Blue G.

Figure 3. Biochemcial analyses of substrate specificity

A. HAT assays were performed with 0.3 pmol of Rtt109 and 3 pmol of Asf1 or Vps75 where indicated. Substrates were 2 pmol of either chicken (H3/H4)₂ tetramers in solution (lanes 1–3), tetramers deposited onto arrays of 5S DNA (lanes 4–6), or complete nucleosomes on 5S DNA arrays (lanes 7–9). Products were analyzed by SDS-PAGE and immunoblotting. B. SDS-PAGE analysis of HAT reactions performed using 2.0 μM [³H]-AcCoA, 3 pmol of Asf1 or Vps75, 1.0 pmol of Rtt109, and 6 pmol of yeast histones H3/H4. Products were analyzed by silver staining (left) or autoradiography for 6 days (right). C. Rate of Rtt109/Vps75 labeling of individual core histones. Filter binding assays were performed with 75 μM Acetyl-CoA, 20 μM histone, 0.4–0.8 μM Rtt109/Vps75 complex at 25 ºC. Rates are averages from triplicate experiments with standard deviation shown.

Figure 4. Mass spectrometric analyses of substrate specificity

A. Single acetylation of H3 by the Rtt109/Asf1 complex. Upper panel: Rtt109/Asf1 was incubated with recombinant yeast histones H3/H4 in the presence of AcCoA. Reaction products were analyzed by LCMS and a portion of the chromatograph is shown. Peaks were analyzed by MS/MS, resulting in the indicated mass observations. The expected mass of recombinant yeast H3 is 15,225 Da and H4 is 11,237 Da. Lower panel: As above, except that AcCoA was omitted. B. H3-K56 is the preferred substrate for the Rtt109/Asf1 complex. MS/MS Fragmentation ions from the selected acetylated peptide shown above. C. H3-K56 is the preferred substrate for the Rtt109/Vps75 complex. MS/MS Fragmentation ions from the selected acetylated peptide. No additional acetylated peptides were identified.

Figure 5. Enzymatic characterization of Rtt109

A. Acetyl-CoA saturation curve using the preformed Rtt109-Vps75 complex. Acetyl-CoA was varied from 0.25 to 13.5 μM at 30 μM histone H3 in filter binding assays containing 0.1 μM Rtt109-Vps75 complex. The k_cat was 0.19 ± 0.01 s⁻¹, with a K_m for acetyl-CoA of 1 ± 0.2 μM and a k_cat/K_m =1.9 ± 0.4 x 10⁵ M⁻¹ s⁻¹. Data from duplicate samples are shown. B. Time course of activity of Rtt109 on H3. Filter binding assay was performed with 75 μM Acetyl-CoA, 20 μM histone H3 at 25 ºC. Either 1 μM Rtt109 (squares) or 0.46 μM Rtt109-Vps75 complex (circles) was used in assays. Initial velocity for the Rtt109-Vps75 complex was 0.13 s⁻¹ with background levels of activity for Rtt109 alone. Experiments with a higher concentration of H3 showed similar results (data not shown). C. H3/H4 saturation curve using Rtt109 with Asf1. The rate of acetylation was determined at H3/H4 concentrations (as dimer) from 0.3 to 9.6 or 4.8 μM at 4.8 μM acetyl-CoA. Filter binding assays contained 0.06 μM Rtt109 and 0.3 μM of Asf1. k_cat was 0.021 ± 0.002 s⁻¹, with a K_m for histones H3/H4 of 1.19 ± 0.34 μM, and a k_cat/K_m value of 2.0 ± 0.5 x 10⁴ M⁻¹ s⁻¹. Data from duplicate samples are shown. D. As in C, with 0.3 μM Vps75 as the cofactor. k_cat was 0.34± 0.04 s⁻¹ with K_m for histones H3/H4 of 0.84 ± 0.28 μM, and a k_cat/K_m value of 4.4 ± 0.9 x 10⁵ M⁻¹ s⁻¹. Data from duplicate samples are shown. E. Mutant Rtt109 proteins bind Vps75. Wild-type or mutant His6-Rtt109 proteins containing residues 287–288 or 318–320 changed to alanines were coexpressed in bacteria with untagged Vps75. Purified complexes were analyzed on a 15% SDS-PAGE gel and stained with Coomassie Blue. These preparations contain a greater amount of a bacterial heat shock protein (*) than do the Rtt109/Vps75-His6 preparations shown in Figure 1C. F. HAT activity of mutant enzyme complexes. Filter binding assays were performed with 0, 80 or 320 ng of recombinant WT or mutant His6-Rtt109/Vps75 complexes, 24 pmol of yeast histones H3/H4 and 3.2 μM [³H]-AcCoA.

Figure 6. H4 is required for Rtt109/Asf1 activity

A. Free H3 saturation curve with the Rtt109-Vps75 complex. The rate of acetylation was determined at H3 concentrations from 0.8 to 86 μM at 75 μM acetyl-CoA. Filter binding assays contained 0.1 to 0.3 μM Rtt109-Vps75 complex. k_cat was determined to be 0.21± 0.04 s⁻¹, with a K_m for histone H3 of 5.9 ± 0.8 μM, and a k_cat/K_m value of 3.5 ± 0.9 x 10⁴ M⁻¹ s⁻¹. Experiments were done in triplicate with representative data shown. B. Vps75 but not Asf1 can stimulate acetylation of free H3. Filter binding assays contained 75 μM acetyl-CoA, 20 μM H3. The concentration of Rtt109 (alone or in the coexpressed Rtt109-Vps75 complex) was 0.4 μM, Asf1 was 0.5 μM (+) or 1 μM (++), and Vps75 was 0.7 (+) or1.4 μM (++). 5 μg (+) or 10 μg (++)BSA was used as a control for interaction specificity. Data are averages from two experiments with standard deviation shown. C. Asf1 stimulates H3 acetylation only in the presence of H4. Acetylation was monitored in filter binding assays with 75 μM acetyl-CoA, 20 μM each of H3 and/or H4, and 0.1 μM Rtt109. 0.7 μM Asf1 was added where noted. Data are averages from two experiments with standard deviation shown.

Figure 7. Asf1 and Vps75 have functionally different roles in vivo

A. Similar sensitivity to CPT caused by asf1Δ, rtt109Δ and H3-K56R mutations. Equal numbers of log-phase cells of the indicated genotypes were plated on rich, non-selective media (YPD) or on the same media containing 0.5 μg/ml CPT. Plates were grown at 30°C for 3 days prior to photography. B. As in A, comparing rtt109Δ and vps75Δ mutants. C. Chromatin immunoprecipitation analysis of PCNA enrichment at an early replication origin (ARS607). α factor-arrested cells were released into 0.2 M HU for 30 minutes and crosslinked with formaldehyde. Triplicate IP reactions with an anti-PCNA antibody were carried out using extracts from the indicated strains. PCR was used to detect the amount of ARS607 DNA immunoprecipitated relative to the +14kb band, which represents a DNA probe distal from the origin. D. As in C, comparing rtt109Δ and asf1Δ mutants. PI indicates immunoprecipitation with pre-immune control sera. E. As above, except that IPs were performed with anti-H3-K56ac antibodies.