14-3-3 Binding to histone H3 is dependent on H3 phosphorylation and stabilized by additional acetylation. (A) Induction of phosphoacetylation increases histone H3 interaction with 14-3-3. Histones were isolated from resting 3T3 fibroblasts that were either left untreated (0) or stimulated for 1 h with anisomycin and TSA (A/T) and incubated with GST or GST–14-3-3ζ. Bound histones were analyzed by immunoblotting with antibodies against ph/ac histone H3 (panel i) and C-terminal histone H3 (H3 C-term) (panel ii). Loading of GST and GST–14-3-3 was monitored by Ponceau staining (panel iii). (B) In vitro modification of histone H3. Recombinant histone H3 was phosphorylated by MSK1 (lane2), acetylated by PCAF (lane 4) or phosphoacetylated with both enzymes (lane 3). Enzymes were omitted in control reactions (lane 1). The modification status was analyzed by sequential immunoblotting with the indicated antibodies. Corresponding modifications are denoted at the top. (C) Acetylation effects on the 14-3-3/histone H3 interaction are more dominant for the R23A28 mutant than for the A10R14 mutant. The indicated histone H3 mutants were in vitro modified as indicated and incubated with GST–14-3-3ζ proteins. Bound histone H3 proteins were analyzed by immunoblotting with C-terminal histone H3 antibodies. The panel shows one representative experiment for each mutant or WT histone H3.

Binding of 14-3-3 to an H3S10ph peptide is enhanced by additional lysine acetylation. (A) Lysine acetylation increases the affinity of 14-3-3 for the phosphorylated H3 peptide. Binding of 14-3-3ζ to the indicated H3 peptides was analyzed by fluorescence polarization measurements. The panel shows the average of at least three independent measurements (mean±s.d.). (B) Dissociation constants (K_d in μM) for the interaction of different 14-3-3 isoforms with the indicated histone H3 peptides determined by fluorescence polarization measurements. Values are average (mean±s.d.) of at least three independent measurements.

14-3-3ζ Is required for transcriptional activation of the HDAC1 gene by anisomycin and TSA. (A) Isoform-specific depletion of 14-3-3 proteins by RNA interference. HeLa cells were transfected with siRNAs against 14-3-3ɛ or ζ and either left untreated or stimulated with anisomycin and TSA (A/T). Whole-cell lysates were prepared and analyzed for 14-3-3 protein levels by immunoblotting. Equal loading was confirmed with a β-actin antibody. (B) 14-3-3 Depletion has no impact on the phosphoacetylation status of histone H3. HeLa cells, depleted of 14-3-3ɛ or ζ isoforms, and control cells were stimulated with TSA and anisomycin (sAn) for 1 h. ChIP assays were performed with the indicated antibodies. (C) Depletion of 14-3-3ζ interferes with transcriptional activation of the HDAC1 gene. HeLa cells were transfected with siRNAs as described for panel A. HDAC1 expression levels were determined by real-time RT–PCR, normalized to GAPDH levels and are depicted as fold increase compared with untreated samples, which were transfected with control siRNA. Values represent three independent experiments (mean±s.d.). Depletion of 14-3-3ζ interferes with the induction of HDAC1 by anisomycin and TSA treatment (lane 6, ^**P=0.002 and lane 8, ^*P=0.01, t-test), whereas 14-3-3ɛ depletion caused a more moderate reduction (lane 4, P=0.07, t-test).

Modulation of the histone H3/14-3-3 interaction by additional modifications. (A) Histone H3/14-3-3 interaction is modulated by additional lysine acetylation. IVT ³⁵S-methionine-labeled 14-3-3ζ was incubated with differentially modified gel-coupled histone H3 peptides. Bound proteins were analyzed by SDS–PAGE and fluorography. The panel shows one representative experiment. The signal intensity for each band was quantified and is depicted as summary of five independent measurements (mean±s.d.). Values were normalized relative to H3S10ph peptide-bound fraction (lane 2). Additional acetylation increased the association with 14-3-3 proteins (lane 4, ^*P=0.001, t-test). A similar effect of H3K14 acetylation was observed for the H3K9me2/S10ph peptide (lane 7, ^**P<0.001, t-test). (B) Nuclear 14-3-3 proteins preferentially bind to the S10phK14ac histone H3 peptide. Nuclear extracts were incubated with unmodified, S10ph or S10phK14ac H3 peptides. An aliquot of the nuclear extracts was used as input control. Bound proteins were analyzed on immunoblots with 14-3-3ζ antibodies. (C) Additional K14 acetylation increases the competitor potential of the S10ph histone H3 peptide. Binding reactions were performed as described for panel A (lanes 1–3). In addition, binding reactions on the ph/ac peptide were performed in the presence of a 20-fold molar excess of unmodified, H3S10ph or H3S10phK14ac free competitor peptides. Bound 14-3-3ζ proteins were analyzed by SDS–PAGE and fluorography. Each signal was normalized to the non-competed H3S10phK14ac peptide binding (lane 3) and is depicted as histogram showing the average of three independent experiments (mean±s.d.).

Localization of 14-3-3 proteins to the HDAC1 promoter correlates with histone H3 phosphoacetylation. (A) Phosphorylation and acetylation of histone H3 synergize in HDAC1 gene expression. Resting 3T3 fibroblasts were either left untreated (rest) or stimulated with TSA, anisomycin (An) or both drugs simultaneously (A/T) for 6 h. HDAC1 mRNA expression was analyzed by quantitative real-time RT–PCR and normalized to β-actin. Data are depicted relative to the untreated control as average of two independent experiments (mean±s.d.). (B) Localization of 14-3-3ζ to the HDAC1 promoter correlates with phosphoacetylation of histone H3. Chromatin was prepared from resting and TSA-, anisomycin- or TSA/anisomycin-stimulated 3T3 fibroblasts and ChIP assays were performed with the indicated antibodies. Rabbit pre-immune serum was used as unspecific antibody control (un). Precipitated and input DNAs were analyzed by semi-quantitative PCR with primers specific for the HDAC1 promoter and the β-actin promoter. Intensities were quantified relative to input DNA (arbitrary units) and are indicated above the panels. (C) The H3K9me2 and H3S10ph epitopes are inversely present at the HDAC1 promoter. Induction of histone H3 phosphoacetylation via anisomycin (An) or combinatorial treatment (A/T) leads to reduced epitope availability for the H3K9me2 antibody whereas the H3S10ph antibody is only slightly affected. (D) Transcriptional super-induction of HDAC1 requires the nucleosomal response. Resting 3T3 fibroblasts were treated with anisomycin and TSA (A/T), or in addition pretreated with H89 (10 μM) 15 min before A/T treatment (A/T+H89). Expression analysis was performed as described for panel A. (E) The nucleosomal response is required for localizing 14-3-3ζ to the HDAC1 promoter and to generate the phosphomethyl dual mark. ChIP analysis with antibodies specific for modified histone H3, 14-3-3ζ and HP1γ was performed as described for panel B.

Model for the role of histone H3 modifications in the regulation of HP1 and 14-3-3 binding during transcriptional induction of the HDAC1 gene. In the transcriptional silent state, the HDAC1 promoter region is occupied by H3K9me2-modified nucleosomes and HP1γ. Activation of MAP kinase signaling via anisomycin and inhibition of HDAC activity via TSA leads to the formation of triple modified histone H3, stable binding of 14-3-3 and transcriptional induction of the HDAC1 gene. Removal of the phosphorylation and acetylation marks via phosphatases and deacetylases, respectively, can regenerate K9 dimethylated histone H3 and allow the re-association of HP1 proteins.