Stimulation of CREB binding protein nucleosomal histone acetyltransferase activity by a class of transcriptional activators

Abstract

The transcriptional coactivator CREB binding protein (CBP) possesses intrinsic histone acetyltransferase (HAT) activity that is important for gene regulation. CBP binds to and cooperates with numerous nuclear factors to stimulate transcription, but it is unclear if these factors modulate CBP HAT activity. Our previous work showed that CBP interacts with the Epstein-Barr virus-encoded basic region zipper (b-zip) protein, Zta, and augments its transcriptional activity. Here we report that Zta strongly enhances CBP-mediated acetylation of nucleosomal histones. Zta stimulated the HAT activity of CBP that had been partially purified or immunoprecipitated from mammalian cells as well as from affinity-purified, baculovirus expressed CBP. Stimulation of nucleosome acetylation required the CBP HAT domain, the Zta DNA binding and transcription activation domain, and nucleosomal DNA. In addition to Zta, we found that two other b-zip proteins, NF-E2 and C/EBPalpha, strongly stimulated nucleosomal HAT activity. In contrast, several CBP-binding proteins, including phospho-CREB, JUN/FOS, GATA-1, Pit-1, and EKLF, failed to stimulate HAT activity. These results demonstrate that a subset of transcriptional activators enhance the nucleosome-directed HAT activity of CBP and suggest that nuclear factors may regulate transcription by altering substrate recognition and/or the enzymatic activity of chromatin modifying coactivators.

FIG. 1

CBP HAT domain contributes to Zta coactivation. HeLa cells were cotransfected with 1 μg of BZLF1-luciferase reporter plasmid and expression plasmids for Zta, CBPwt, CBPΔH, CBPΔN, CBPΔNΔH, or the pCMV4 control vector as indicated in the figure. Luciferase units were expressed as fold activation. Values are the averages of at least three independent transfections.

FIG. 2

Zta stimulates CBP-associated HAT activity. (A) HeLa nuclear extract (Nxt) (250 μg) was incubated with [³H]acetyl CoA in the absence (lane 1) or presence (lane 2) of Zta (300 ng) and then assayed by SDS-PAGE and fluorography. (B) HeLa nuclear extract was incubated with [³H]acetyl CoA without (lane 1) or with Zta (lane 2). PC fractions of HeLa nuclear extracts were incubated with [³H]acetyl CoA and purified SONs with or without Zta as indicated. Acetylation of nucleosomes was assayed by SDS-PAGE and fluorography. The positions of core histones H3, H2B, H2A, and H4 are indicated at the right. (C) Western blot of HeLa nuclear extract and PC fractions probed with CBP-specific antisera. (D) HAT activity of CBP-specific immunoprecipitates can be stimulated by Zta. HAT assays containing SONs, [³H]acetyl CoA and either the PC-B fraction (lanes 1 and 2), CBP-specific immunoprecipitates (lanes 3 and 4), P/CAF-specific immunoprecipitates (lanes 5 and 6), or protein A only immunoprecipitates (lanes 7 and 8) were incubated in the absence (−) or presence (+) of Zta. Acetylated CBP and histones are indicated.

FIG. 3

CBP HAT activity is necessary and sufficient for activation by Zta. (A) Immunoprecipitates derived from NIH 3T3 cells transfected with Flag-tagged full-length CBP (lanes 1 and 2), Flag-tagged CBP ΔHAT (lanes 3 and 4) or vector (lanes 5 and 6) were assayed for HAT activity with purified SONs in the absence (−) or presence (+) of Zta. (B) Western blot of immunoprecipitates derived from NIH 3T3 cells transfected with Flag CBP, Flag CBPΔHAT, or vector that were used for the HAT assay in panel A. (C) Ni-NTA-purified His-tagged CBP expressed and purified from baculovirus was assayed for acetylation of SONs in the absence (−) or presence (+) of Zta. CBP was increased by threefold increments up to 200 ng. Acetylated products were visualized by fluorography of SDS-PAGE gels.

FIG. 4

Transcriptional activation domain of Zta is required for stimulation of HAT activity. (A) Schematic indicating Zta functional domains. The transcriptional activation domain (TAD) maps to amino acid residues 1 to 141. The DNA binding basic region (BR) and zipper-like dimerization domain (Z) map to residues 141 to 245. Alanine substitution mutations in the activation domains m.1 and m.2 and the DNA binding (dbd) domain are indicated. HeLa nuclear extracts were incubated alone, with full-length Zta, with Zta (Δ2-141), or with Zta (Δ141-245) and were assayed for HAT activation. (B) Zta mutants m.1 and m.2 abrogate CBP recruitment in magnesium agarose EMSA. The Z₅E4T promoter probe alone (lane 1, at left) and with wild-type Zta (lane 2), m.1 (lane 3), or m.2 (lane 4) are shown with GST (left panel), GST-CBP-C/H1 (middle panel), or GST-CBP-C/H3 (right panel). (C) Transiently transfected HeLa cells were assayed for CAT activity from the BHLF1 promoter after cotransfection of vector, wild-type Zta, m.1, or m.2. (D) Nucleosome acetylation was assayed with HeLa nuclear extract and [³H]acetyl CoA alone (lane 1) or with wild-type Zta (lane 2), m.1 (lane 3), or m.2 (lane 4). Wild-type Zta, m.1, and m.2 were compared by Coomassie brilliant blue staining of SDS-polyacrylamide gels (lower panel).

FIG. 5

DNA binding activity of Zta is required for stimulation of HAT activity. (A) Increasing concentrations (33, 100, and 300 ng) of wild-type Zta (lanes 1 to 4) and Zta-dbd (lanes 5 to 8) were compared for their ability to bind radiolabeled ZRE in EMSA. (B) Zta-wt (lanes 1 to 3) and Zta-dbd (lanes 4 to 6) were compared at 33, 100, and 300 ng for stimulation of nucleosomal HAT activity in HeLa nuclear extracts with [³H]acetyl CoA. Zta and Zta-dbd visualized by Coomassie brilliant blue staining of SDS-polyacrylamide gels (lower panel). (C) HeLa nuclear extracts with [³H]acetyl-CoA were incubated with Zta (lanes 2 to 8) and increasing concentrations of ZRE-wt oligonucleotide (lanes 3 to 5) or ZRE-m oligonucleotide (lanes 6 to 8). Acetylated histones were visualized by SDS-PAGE and fluorography. (D) ³²P-labeled SONs (20 ng) were incubated with Zta (10, 30, or 90 ng) or Zta-dbd (10, 30, or 90 ng) and assayed by EMSA. (E) Zta (30 ng) and ³²P-labeled SONs (20 ng) were incubated with ZRE or ZRE-m oligonucleotide (fivefold dilutions up to 1.0 μg) and assayed by EMSA.

FIG. 6

Activation of oligonucleosome-specific acetylation by Zta. (A) Increasing amounts of SONs (lanes 1 to 6), hydroxyapatite purified core histones (lanes 7 to 12), or acid-extracted free histones (lanes 13 to 18) were assayed for acetylation by phosphocellulose B fraction in the absence (−) or presence (+) of Zta. The amount of histones in each reaction is indicated above each set of reactions. (B) Characterization of DNA in oligonucleosomes. The DNA purified from SONs (lane 1) or SONs treated with 0.2 (lane 2), 2 (lane 3), or 20 (lane 4) U of MNase/ml were extracted, electrophoresed on a 1.5% agarose gel, and visualized by ethidium bromide staining. (C) MNase digestion inhibits HAT activation by Zta. SONs alone (lane 1) or SONs digested with 0.2 (lanes 2), 2.0 (lanes 3), or 20 (lanes 4) U of MNase/ml (as shown in panel B) were assayed as substrates for acetylation in the absence (−) or presence (+) of 300 ng of Zta.

FIG. 7

Template-specific targeting of CBP acetylation. Biotinylated ZRE and GRE templates (400 bp) were assembled into dinucleosomes and then incubated with CBP and [³H]acetyl CoA in the presence (+) or absence (−) of Zta. Biotinylated templates were isolated on magnetic beads after the HAT reaction and were assayed by autoradiography of SDS-polyacrylamide gels. The control lane contains a nonbiotinylated ZRE template, and a reaction not purified on magnetic beads is shown in the lanes marked SONs.

FIG. 8

A class of transcription activators stimulate nucleosomal HAT activity. (A) For each protein, 0.5 μg of purified recombinant Zta, NF-E2, C/EBPα, GATA-1, CREM, EKLF, Pit-1, Jun+Fos, phospho-CREB, and GCN4 was tested for its ability to stimulate acetylation of SONs using PC-B as the source of acetylase. (B) Coomassie brilliant blue staining of SDS-polyacrylamide gels containing 1.0 μg of the various recombinant activator proteins used in panel A. (C) Oligonucleosome binding of transcription factors was assayed by agarose gel EMSA with ³²P-labeled HeLa-derived SONs. Proteins were assayed as threefold dilutions from 20 to 180 ng for each reaction.