Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 May;1861(5):481-496.
doi: 10.1016/j.bbagrm.2018.03.002. Epub 2018 Mar 8.

Transcriptional Regulation Mediated by H2A.Z via ANP32e-dependent Inhibition of Protein Phosphatase 2A

Affiliations
Free PMC article

Transcriptional Regulation Mediated by H2A.Z via ANP32e-dependent Inhibition of Protein Phosphatase 2A

Hyewon Shin et al. Biochim Biophys Acta Gene Regul Mech. .
Free PMC article

Abstract

The mechanisms that regulate H2A.Z and its requirement for transcription in differentiated mammalian cells remains ambiguous. In this study, we identified the interaction between the C-terminus of ANP32e and N-terminus of H2A.Z in a yeast two-hybrid screen. Knockdown of ANP32e resulted in proteasomal degradation and nuclear depletion of H2A.Z or of a chimeric green florescence protein fused to its N-terminus. This effect was reversed by inhibition of protein phosphatase 2A (PP2A) and, conversely, reproduced by overexpression of its catalytic subunit. Accordingly, knockdown of ANP32e inhibited phosphorylation of H2A.Z, whereas a mutation of serine-9 proved its requirement for both the protein's stability and nuclear localization, as did knockdown of the nuclear mitogen and stress-induced kinase 1. Moreover, ANP32e's knockdown also revealed its differential requirement for cell signaling and gene expression, whereas, genome-wide binding analysis confirmed its co-localization with H2A.Z at transcription start sites, as well as, gene bodies of inducible and tissue-specific genes. The data also suggest that H2A.Z restricts transcription, which is moderated by ANP32e at the promoter and gene bodies of expressed genes. Thus, ANP32e, through inhibition of PP2A, is required for nucleosomal inclusion of H2A.Z and the regulation of gene expression.

Keywords: Chromatin immunoprecipitation; Histone; MSK1; Yeast two-hybrid.

Figures

Figure 1
Figure 1. ANP32e interacts with H2A.Z
a. The bait plasmid vectors harboring fusion proteins of Gal4-DB with H2A.Z, H2A.X, the C-terminus of H2A.Z (CT), H2A.Z with a deletion of the N-terminal 16 aa (ΔN16), the N-terminal 16 aa (NT), or H2A.Z with a deletion of the N-terminal 6 aa (ΔN6), were transfected into the Y190 yeast strain along with the prey plasmid vector with the Gal4AD-ANP32e (nt 680–1070) fusion protein. The transfected yeast strain was streaked on medium deficient in leucine and tryptophan (-Leu-Trp), left plate, for selection of the cells that received both the bait and prey plasmids, or deficient in leucine, tryptophan, and histidine (-Leu-Trp-His), middle plate, for selection of cells that contain the bait and prey plasmids expressing fusion proteins that physically interact with one another. The colonies from the -Leu-Trp plate were also subjected to a filter-lift Lacz assay for confirming those interactions (right). b. Cardiac myocytes were transfected with plasmids harboring either GST or the GST-ANP32e fusing protein. After 24 h, cell lysates were subjected to immunoprecipitation with a rabbit polyclonal anti-GST, a mouse monoclonal anti-H2A.Z, or mouse IgG, as indicated. The immunoprecipitate was then analyzed by Western blotting using anti-H2A.Z, rabbit polyclonal anti-ANP32e, or anti-H2A, as indicated. The input is 1/20 the protein concentration that was used for immunoprecipitation. c. GFP fusion proteins with either the N-terminus domain of H2A.Z (zGFP), its C-terminal domain (GFPz), both domains (zGFPz), or the N-terminal domain of H2A.X and the C-terminal domain of H2A.Z (xGFPz) were constructed and delivered to cardiac myocytes via adenoviral vectors. After 24 h, cell lysates were subjected to immunoprecipitation with anti-GFP and the immunoprecipitate analyzed by Western blotting using anti-GFP and anti-ANP32e, as indicated. In b. and c., the input is 1/20 the protein concentration used for immunoprecipitation (200 μg).
Figure 2
Figure 2. ANP32e is necessary for H2A.Z stabilization and nuclear localization
a. Increasing doses (moi 10, 20, 30) of ANP32e or Lacz (control) were delivered to cardiac myocytes via adenoviral vectors. After 24 h, the cells were lysed, fractionated into cytosol (cyto), membrane (mem), nuclear (nuc), and cytoskeletal (cytoSk) fractions, and analyzed by Western blotting using anti-ANP32e, -H2A.Z, -H2A.X, -AKT, and –β1AR, as indicated. b. Short-hairpin RNA targeting ANP32e (iANP32e, moi 10), or a control construct, was delivered to cardiac myocytes via adenoviral vectors, in the presence or absence of 0.1 μM epoxomicin (Epox). After 24 h, the cells were lysed, fractionated into cytosol, membrane, nuclear, and cytoskeletal fractions, and analyzed by Western blotting using anti-ANP32e, -H2A.Z, -H2A, -pSer/pThr, and –actin, as indicated. c. The ANP32e Western blot signals in the nuclear fraction in (b.), and d. H2A.Z signals in the nuclear (black bars) and cytoskeletal fractions (orange bars) in (b.), were quantitated, normalized to H2A or actin, respectively, averaged, and plotted as relative levels (n=3). Error bars represent S.E.M., *p<0.05 v. nuclear control, # p<0.05 v. cytoSk control, $ p<0.05 v. Epox CytoSk.
Figure 3
Figure 3. ANP32e-induced stabilization of H2A.Z is mediated through its N-terminal tail
a.–b. Short-hairpin RNA targeting ANP32e (iANP32e, where indicated by plus a + signs), or a control construct (unmarked lanes), was delivered to cardiac myocytes via an adenoviral vector. After 24 h, GFP or GFP fusion proteins with either the N-terminus domain of H2A.Z (zGFP), its C-terminal domain (GFPz), or both domains (zGFPz), were delivered to cardiac myocytes via adenoviral vectors, as indicated by plus a + signs. After 24 h, the cells were lysed, fractionated into cytosol (cyto), membrane (mem), nuclear (nuc), and cytoskeletal (cytoSk) fractions, and analyzed by Western blotting using anti-Anp32e, -GFP, -H2A.Z, -H2A, -beta1 adrenergic receptor (β1AR), and -actin, as indicated. c. Western blot signals for GFP, GFPz, zGFP, and zGFPz, in the absence (black bars) or presence (orange bars) of iANP32e, shown in a. and b., in the cytosol, membrane, and nuclear fractions, were quantitated, normalized to actin, β1AR, or H2A, respectively, averaged, and plotted as relative values (n=3), error bars represent S.E.M, *p<0.01 v. control.
Figure 4
Figure 4. PP2A mediates the effects of ANP32e
a. Short-hairpin RNA targeting ANP32e (iANP32e, moi 10), or a control construct, was delivered to cardiac myocytes via adenoviral vectors, in the presence or absence of 10 nM okadiac acid (OA). After 24 h, the cells were lysed, fractionated into cytosol, membrane, nuclear, and cytoskeletal fractions, and analyzed by Western blotting using anti-ANP32e, -pSer/pThr, -H2A.Z, -H2A, and -AKT, as indicated. b. PP2A, with or without H2A.Z, or a control construct, were delivered to cardiac myocytes via adenoviral vectors (moi 10), in the presence or absence of 0.1 μM epoxomicin (Epox). After 24 h, the cells were lysed, fractionated into cytosol (cyto), membrane (mem), nuclear (nuc), and cytoskeletal (cytoSk) fractions, and analyzed by Western blotting using anti-H2A.Z, -PP2A, -pH3, and -actin, as indicated. c. Left graph, the H2A.Z and pSer/pThr signals in the nuclear fractions of the images in (a.) were quantitated, normalized to H2A, and graphed as relative values (n=3). c. Right graph, the H2A.Z signal in the nuclear and cytoskeletal fraction of the images in (b.) were quantitated, normalized to pH3 or actin, respectively, and graphed as relative values (n=3). Error bars represent S.E.M., *p<0.05 v. control.
Figure 5
Figure 5. Knockdown of ANP32e destabilized H2A.Z and inhibits its phosphorylation
a. Short-hairpin RNA targeting ANP32e (iANP32e, moi 10), or a control construct, was delivered to cardiac myocytes via adenoviral vectors. After 24 h, histones were acid extracted. Total histones (top 2 panels), or those immunoprecipitated with a pSer/pThr antibody (lower 2 panels), were separated by 2D gel electrophoresis followed by blotting with anti-H2A.Z or anti-H2A (n=2). b. Wild type H2A.Z, or Ser9 mutants, H2A.Z (A9 or D9, moi 10), in the presence or absence of 0.1 μM epoxomicin (EPOX). After 24 h, the cells were lysed, fractionated into cytosol (cyto), membrane (mem), nuclear (nuc), and cytoskeletal (cytoSk) fractions, and analyzed by Western blotting using anti-H2A.Z, -H2A.X, and -actin, as indicated. c. The H2A.Z signal in the nuclear and cytoskeletal fraction of the images in (b.) were quantitated, normalized to H2A.X or actin, respectively, and graphed as relative values (n=3). Error bars represent S.E.M., *p<0.05 v. control of equivalent fraction, #p<0.05 v. nuclear control.
Figure 6
Figure 6. Knockdown of MSK1 induces degradation of H2A.Z
a. Short-hairpin targeting MSK1 (iMSK1), or a control construct, was delivered to cardiac myocytes (moi 10, 15, 20, 30), in the presence or absence of 0.1 μM epoxomicin (EPOX). After 24 h, the cells were lysed, fractionated into cytosol (cyto), membrane (mem), nuclear (nuc), and cytoskeletal (cytoSk) fractions, and analyzed by Western blotting using anti-H2A.Z, -ubiquitin, -MSK1, -pH2B, -H2A.X, and actin, as indicated. b. The H2A.Z signal in the nuclear and cytoskeletal fraction of the images in (a.) were quantitated, normalized to H2A.X or actin, respectively, and graphed as relative values (n=3). Error bars represent S.E.M., *p<0.05 v. nuclear control.
Figure 7
Figure 7. ANP32e is required for FBS-induced phosphorylation of Erk1/2 and S6, and selective gene expression
a. Short-hairpin RNA targeting ANP32e (iANP32e, moi 10), or a control construct, was delivered to cardiac myocytes via adenoviral vectors. After 24 h, the cells were either stimulated or not with 10%FBS for 5 minutes, before they were lysed, and fractionated into cytosol (cyto), membrane (mem), nuclear (nuc), and cytoskeletal (cytoSk) fractions, and analyzed by Western blotting using anti-H2A.Z, -phospho-p42/p44 (p-p42/p44), -phospho-S6 (p-S6), and -H2A.X, as indicated. b. The p-p42/p44 and pS6 signals in the cytosolic fraction were quantitated, normalized to total p42/p44 and S6, respectively, and graphed as relative values (n=3). Error bars represent S.E.M., *p<0.05 v. control, #p<0.05 v. FBS-stimulated. c. Short-hairpin RNA targeting ANP32e (iANP32e, moi 10), PP2A catalytic subunit (PP2Ac), or a control construct, were delivered to cardiac myocytes via adenoviral vectors. After 24 h, the cells were either stimulated, or not, with 10%FBS for 24 h, before they were lysed, and fractionated into cytosol (cyto), membrane (mem), nuclear (nuc), and cytoskeletal (cytoSk) fractions, and analyzed by Western blotting using anti-H2A.Z, -ANP32e, -PP2Ac, -Cdk7, -Ac-H3, -H2A, -Akt, and -actin, as indicated. d. The H2A.Z, ANP32e, PP2Ac, Cdk7, and Ac-H3 in the nuclear fraction were quantitated, normalized to H2A, and graphed as relative values (n=3). Error bars represent S.E.M., *p<0.05 v. control, #p<0.05 v. FBS-stimulated.
Figure 8
Figure 8. Co-localization of H2A.Z and ANP32e at the TSS of transcriptional active and inactive genes
Mice were subjected to a sham or transverse aortic constriction surgical (TAC) operation. One week post-TAC, the hearts were isolated and subjected to pol II, H2A.Z, and ANP32e ChIP-Seq. a. Heat maps of the H2A.Z and ANP32e ChIP-Seq sequence fragments from sham, TAC, and input at the TSS. The Y-axis represents individual positions of bins, and the X-axis represents a region from −2000 to +2000 bp relative to the TSS. b. Graphs representing average peak values of H2A.Z and ANP32e ChIP-Seq fragments from sham, TAC, and input from −2000 to +2000 bp relative to the TSS. c. Histograms showing the distribution of fragments calculated from their overall frequencies in the ‘tac-h2az’ (on X-axis) v. ‘tac-ANP32e’ (on Y-axis) within the TSS (upper) and gene body (lower) regions. The X and Y-axes were segmented into 75 bins, and the number of fragments within each bin was counted, color coded, and plotted. The bar to the right of the plot illustrates the relationship between count and coloring. The plots represent pseudo-colored 2D matrices showing observed/expected distribution calculated from the overall frequencies of fragments on each of the axes. This plot shows the relation between H2A.Z and ANP32e levels relative to what is expected if they occurred by chance. The pseudo-color corresponds to the Obs/Exp ratio, and the color intensity is proportional to the log2 of the number of observed fragments within each bin. These plots suggest that there is a positive correlation between the levels of H2A.Z and ANP32e, where the red indicates that this occurs more frequently than expected by chance. This plot also shows that regions that have the lowest levels of ANP32e and highest levels of H2A.Z, or conversely, those with the lowest levels of H2A.Z and highest ANP32e, occur less frequently than expected by chance (blue). d.–f. The plots represent ChIP-Seq results for Cdk7, Hist3h2a, and Rps6, for which the sequence fragments of pol II, H2A.Z, and ANP32e are aligned across the gene coordinates in Integrated Gene Browser (IGB). g. A graph representing total H2A.Z and ANP32e sequence tags at the TSS of Cdk7, Hist3h2a, and Rps6. Genes were sorted in h. housekeeping, i. inducible-higher Anp32e, j. inducible-lower Anp32e, k. tissue-specific, and l. downregulated gene groups, sorted according to the changes in pol II densities during TAC vs. sham, both at the TSS and gene bodies. Sequence Tags are represented as box plots for the median, quartiles, minimum, and maximum number of Tags. The images represent the densities of pol II, H2A.Z, and ANP32e sequence tags for Cyr61, Myc, Eef1g, Ptrh2, Myl2, and Pln, Klf9, Foxo3, aligned across the gene coordinates, using integrated genome browser (IGB).
Figure 8
Figure 8. Co-localization of H2A.Z and ANP32e at the TSS of transcriptional active and inactive genes
Mice were subjected to a sham or transverse aortic constriction surgical (TAC) operation. One week post-TAC, the hearts were isolated and subjected to pol II, H2A.Z, and ANP32e ChIP-Seq. a. Heat maps of the H2A.Z and ANP32e ChIP-Seq sequence fragments from sham, TAC, and input at the TSS. The Y-axis represents individual positions of bins, and the X-axis represents a region from −2000 to +2000 bp relative to the TSS. b. Graphs representing average peak values of H2A.Z and ANP32e ChIP-Seq fragments from sham, TAC, and input from −2000 to +2000 bp relative to the TSS. c. Histograms showing the distribution of fragments calculated from their overall frequencies in the ‘tac-h2az’ (on X-axis) v. ‘tac-ANP32e’ (on Y-axis) within the TSS (upper) and gene body (lower) regions. The X and Y-axes were segmented into 75 bins, and the number of fragments within each bin was counted, color coded, and plotted. The bar to the right of the plot illustrates the relationship between count and coloring. The plots represent pseudo-colored 2D matrices showing observed/expected distribution calculated from the overall frequencies of fragments on each of the axes. This plot shows the relation between H2A.Z and ANP32e levels relative to what is expected if they occurred by chance. The pseudo-color corresponds to the Obs/Exp ratio, and the color intensity is proportional to the log2 of the number of observed fragments within each bin. These plots suggest that there is a positive correlation between the levels of H2A.Z and ANP32e, where the red indicates that this occurs more frequently than expected by chance. This plot also shows that regions that have the lowest levels of ANP32e and highest levels of H2A.Z, or conversely, those with the lowest levels of H2A.Z and highest ANP32e, occur less frequently than expected by chance (blue). d.–f. The plots represent ChIP-Seq results for Cdk7, Hist3h2a, and Rps6, for which the sequence fragments of pol II, H2A.Z, and ANP32e are aligned across the gene coordinates in Integrated Gene Browser (IGB). g. A graph representing total H2A.Z and ANP32e sequence tags at the TSS of Cdk7, Hist3h2a, and Rps6. Genes were sorted in h. housekeeping, i. inducible-higher Anp32e, j. inducible-lower Anp32e, k. tissue-specific, and l. downregulated gene groups, sorted according to the changes in pol II densities during TAC vs. sham, both at the TSS and gene bodies. Sequence Tags are represented as box plots for the median, quartiles, minimum, and maximum number of Tags. The images represent the densities of pol II, H2A.Z, and ANP32e sequence tags for Cyr61, Myc, Eef1g, Ptrh2, Myl2, and Pln, Klf9, Foxo3, aligned across the gene coordinates, using integrated genome browser (IGB).
Figure 8
Figure 8. Co-localization of H2A.Z and ANP32e at the TSS of transcriptional active and inactive genes
Mice were subjected to a sham or transverse aortic constriction surgical (TAC) operation. One week post-TAC, the hearts were isolated and subjected to pol II, H2A.Z, and ANP32e ChIP-Seq. a. Heat maps of the H2A.Z and ANP32e ChIP-Seq sequence fragments from sham, TAC, and input at the TSS. The Y-axis represents individual positions of bins, and the X-axis represents a region from −2000 to +2000 bp relative to the TSS. b. Graphs representing average peak values of H2A.Z and ANP32e ChIP-Seq fragments from sham, TAC, and input from −2000 to +2000 bp relative to the TSS. c. Histograms showing the distribution of fragments calculated from their overall frequencies in the ‘tac-h2az’ (on X-axis) v. ‘tac-ANP32e’ (on Y-axis) within the TSS (upper) and gene body (lower) regions. The X and Y-axes were segmented into 75 bins, and the number of fragments within each bin was counted, color coded, and plotted. The bar to the right of the plot illustrates the relationship between count and coloring. The plots represent pseudo-colored 2D matrices showing observed/expected distribution calculated from the overall frequencies of fragments on each of the axes. This plot shows the relation between H2A.Z and ANP32e levels relative to what is expected if they occurred by chance. The pseudo-color corresponds to the Obs/Exp ratio, and the color intensity is proportional to the log2 of the number of observed fragments within each bin. These plots suggest that there is a positive correlation between the levels of H2A.Z and ANP32e, where the red indicates that this occurs more frequently than expected by chance. This plot also shows that regions that have the lowest levels of ANP32e and highest levels of H2A.Z, or conversely, those with the lowest levels of H2A.Z and highest ANP32e, occur less frequently than expected by chance (blue). d.–f. The plots represent ChIP-Seq results for Cdk7, Hist3h2a, and Rps6, for which the sequence fragments of pol II, H2A.Z, and ANP32e are aligned across the gene coordinates in Integrated Gene Browser (IGB). g. A graph representing total H2A.Z and ANP32e sequence tags at the TSS of Cdk7, Hist3h2a, and Rps6. Genes were sorted in h. housekeeping, i. inducible-higher Anp32e, j. inducible-lower Anp32e, k. tissue-specific, and l. downregulated gene groups, sorted according to the changes in pol II densities during TAC vs. sham, both at the TSS and gene bodies. Sequence Tags are represented as box plots for the median, quartiles, minimum, and maximum number of Tags. The images represent the densities of pol II, H2A.Z, and ANP32e sequence tags for Cyr61, Myc, Eef1g, Ptrh2, Myl2, and Pln, Klf9, Foxo3, aligned across the gene coordinates, using integrated genome browser (IGB).

Similar articles

See all similar articles

Cited by 3 articles

Publication types

MeSH terms

Feedback