The derived complex-to-complex genetic interaction network for HATs and HDACs. Nodes represent protein complexes and functional modules. Pairs of nodes were linked by an edge if they had significantly more genetic interactions than expected by chance. Node sizes proportional to the number of subunits in the corresponding complex or module; edge thicknesses signifies the statistical significance of the enrichment in genetic interactions (thick and thin edges indicate P < 10⁻¹⁶ and 10⁻¹⁶ ≤ P < 10⁻⁵, respectively).

Post-translational protein modification of the NuA4 core acetylation machinery choreographs DSB repair. (A) MG-132 alters the kinetics of histone H4 acetylation near an HO-induced DNA DSB. Chromatin immunoprecipitation (ChIP) assays with anti-H4-K8^Ac were used to assess kinetics of histone H4 acetylation. Enrichment of histone H4-K8^Ac was quantified by RT–PCR and normalized to the GEA2 internal control and also the local abundance of histone H3 (Supplemental Fig. S11). (B) NuA4 core acetylation machinery components show distinctive kinetics of recruitment to an HO-induced DNA DSB. The enrichment of tagged Esa1p, Yng2p, and Eaf1p after HO induction was assessed by ChIP assays with anti-HA. GAL-HO pdr5Δ strains were used in MG-132 experiments. (C) Proteasome and Rpd3C are recruited to an HO-induced DSB. Enrichment of tagged Rpn11p and Rpd3p after HO induction was assessed by ChIP with anti-HA.

Schematic model for the dynamic regulation of acetylation status near DSBs. Upon introduction of a DSB, such as that caused by HO endonuclease at the HO recognition site, NuA4 is actively recruited focally and rapidly hyperacetylates nearby histone H4. The function of the core acetylation machinery of NuA4 is disrupted when Yng2p is deacetylated through an Rpd3C-depedent mechanism followed by degradation by proteasome, followed by eviction of Esa1p and Epl1p. Three distinct processes appear to be at play to facilitate transition to the hypoacetylation phase crucial for proper DSB repair: (1) The breakdown of Piccolo NuA4 stops ongoing acetylation, (2) Rpd3C is recruited and could actively remove acetyl groups, and (3) ATP-dependent chromatin remodeling complexes, which conduct active nucleosome displacement.

Global features of genetic interaction patterns. (A) Full hierarchical clustering of genetic interaction patterns. Blue, yellow, and black boxes represent aggravating, alleviating, and no interaction, respectively. Only validated data were used in hierarchical clustering and all subsequent computational analysis. The dendrogram of query genes was expanded for visualization, which indicated the similarities of their interaction patterns. The cluster of essential query genes is highlighted inside the blue box. (B) esa1-531 and gcn5Δ partially rescue the growth defect of hda1Δ rpd3Δ double mutants. Growth of each strain was assessed by plating four 10-fold serial dilution on SC–Ura medium (for esa1-531 rescue experiment) or on YPD medium (for gcn5Δ rescue experiment) at 30°C, a semipermissive temperature for esa1-531. (C) esa1-531 and gcn5Δ partially reversed hyperacetylation of histone H4 and H3, respectively, in hda1Δ rpd3Δ double mutant at 30°C, a semipermissive temperature for esa1-531. The acetylation levels of H4 K5, H4 K8, H4 K12, and H4 K16, and H3 K9 and H3 K14 were analyzed by immunoblot.

Acetylation and deacetylation of Yng2p controls its protein stability and function. (A) Yng2p is acetylated in vivo through an Esa1p-dependent mechanism. Wild-type (WT) and esa1-531 strains stably expressing Myc-tagged Yng2p were grown at 25°C (permissive temperature) to OD₆₀₀ ∼0.3, then shifted to 37°C or kept for 4 h at 25°C; WCEs were then collected, immunoprecipitated with anti-Ac-K and probed with anti-Myc. (B) K170 is the major acetylated lysine residue of Yng2p in vivo. K170 was identified by tandem mass spectrometry. Substitution of K170 with arginine (K170R) diminishes acetylation of Yng2p. (C) Effects of K170 substitutions on Yng2p stability. Cells were grown in conditions describes in Figure 3A. WCEs were collected and probed with anti-HA, and band intensities were quantified. K170R mutation causes decreased protein abundance of Yng2p, while K170Q causes no detectable change. (D) Effects of K170 substitutions on Yng2p function. Drug sensitivities were assessed by plating 10-fold serial dilution on YPD medium containing MMS (0.03%) or benomyl (15 μg/mL). (E) Yng2p is deacetylated through an Rpd3p-dependent mechanism in vivo. Cells were grown at 30°C to OD₆₀₀ ∼0.6, then WCEs were collected, immunoprecipitated with anti-Ac-K and probed with anti-Myc.

HDA complex deacetylates Htz1-K14^Ac in vitro and in vivo. (A) Overlap of the enzyme–substrate relationship with genetic interactions. A schematic model depicts genetic and enzyme–substrate interactions between HATs (NuA4 and SAGA), HDACs (HDA complex), and their substrates (Htz1p and histone H3). (B) Deletion of HDA1 rescues the slow-growth and drug-sensitivity phenotype of htz1Δ. Four tetrads dissected from an hda1Δ/HDA1 htz1Δ/HTZ1 diploid strain on YPD medium are shown. Drug sensitivities were assessed by plating 10-fold serial dilution on YPD medium containing HU (200 mM), MMS (0.03%), or benomyl (15 μg/mL). (C,D) Hda1p deacetylates Htz1-K14Ac in vivo and in vitro. The K14 acetylation level in wild-type and HDAC mutant strains was analyzed by immunoblot. (D) In vitro deacetylation activity of Hda1-TAP purified by tandem affinity purification was assessed by incubating it with Flag-tagged Htz1p purified from hda1Δ strain and then monitoring the K14 acetylation level by immunoblot. No protein was added to the control samples. (E) Deletion of HDA1 rescues the sensitivity of htz1-K14R to benomyl. The experimental conditions are described in B. (F) Genetic interactions of an N-tail deletion of histone H3 [hht1(Δ1–36)] with hda1Δ (alleviating) and htz1Δ (aggravating) mutants, suggesting that histone H3 reside in the same functional module as Htz1p.

Functional dissection of the NuA4 complex. (A) Hierarchical clustering of NuA4 component genes based on genome-wide genetic interaction patterns shows functional association. Font colors indicate distinct functional modules (see legend). Subsets of genetic interactions between NuA4 and other complexes were organized for visualization. (B) Core acetylation machinery mutants are differentially sensitive to TSA. Growth curves of esa1-531, epl1-1, and yng2Δ mutants in SC–Ura medium at 37°C (restrictive temperature) with or without TSA (100 μM) are shown. Error bars indicate ±1 SEM from three biological replicates. (C) ESA1 stabilizes and controls the protein abundance of Yng2p. Wild-type (WT) and esa1-531 strains stably expressing tagged Yng2p or Epl1p were grown at 25°C (permissive temperature) to OD₆₀₀ ∼0.3, then shifted to 37°C or kept for 4 h at 25°C; whole-cell extracts (WCEs) were then collected and probed with anti-HA, and band intensities were quantified. (D) Yng2-3HA turnover kinetics in wild-type and esa1-531 strains investigated by cycloheximide turnover experiments. Cells were grown at 25°C to OD₆₀₀ ∼0.3, then either shifted to 37°C or kept at 25°C, and treated with 0.1 μg/mL cycloheximide simultaneously. Equal amount of cells were collected at indicated time points and analyzed by immunoblot as shown in Supplemental Figure S9A. The fraction of Yng2-3HA remaining after cycloheximide addition was plotted. Error bars indicate ±1 SEM. (E) Yng2p is stabilized by MG-132 in esa1-531 cells. Wild-type and esa1-531 cells stably expressing tagged Yng2p were incubated with MG-132 (75 μM) for 4 h at 25°C (permissive temperature) or 37°C (restrictive temperature), after which WCEs were analyzed by immunoblot.