Hdac3 is essential for the maintenance of chromatin structure and genome stability

Abstract

Hdac3 is essential for efficient DNA replication and DNA damage control. Deletion of Hdac3 impaired DNA repair and greatly reduced chromatin compaction and heterochromatin content. These defects corresponded to increases in histone H3K9,K14ac; H4K5ac; and H4K12ac in late S phase of the cell cycle, and histone deposition marks were retained in quiescent Hdac3-null cells. Liver-specific deletion of Hdac3 culminated in hepatocellular carcinoma. Whereas HDAC3 expression was downregulated in only a small number of human liver cancers, the mRNA levels of the HDAC3 cofactor NCOR1 were reduced in one-third of these cases. siRNA targeting of NCOR1 and SMRT (NCOR2) increased H4K5ac and caused DNA damage, indicating that the HDAC3/NCOR/SMRT axis is critical for maintaining chromatin structure and genomic stability.

Figure 1. Loss of Hdac3 impairs DNA repair

MEFs (Hdac3^FL/+ and Hdac3^FL/−) were treated with 0.1 µM tamoxifen for 48 hr and then treated with increasing concentrations of either doxorubicin (A) or cisplatin (B) and cell viability measured with the WST-1 assay. Values in the graph represent mean ± SD. of triplicate samples and the experiment was repeated at least twice. C. A schematic representation of the NHEJ substrate (left), the product formed (right) and the position of real-time PCR primers used to detect the repaired product (Zhuang et al., 2009). D. The reporter cassette used for HR detection is shown schematically. Upon induction of I-Sce1, gene conversion reconstitutes active GFP. The repaired GFP was then measured by FACS analysis. E. Western blot analysis of Hdac3 following siRNA transfection in 293T cells. GAPDH is shown as a loading control. Chromatin-based repair assays performed in 293T cells following knockdown of Hdac3 to measure the efficiency of NHEJ using Q-PCR (F) and HR using FACS (G). The values shown in F and G are the means ± SEM. H. Chromatin immunoprecipitation analysis of H3K9,K14ac and H3K9me3 at the NHEJ substrate before and after siRNA suppression of Hdac3. Quantitative PCR was used to compare the effects of non-targeting and Hdac3 siRNAs and the graphs show average of the relative levels of H3K9,K14ac ± SD. See also figure S1.

Figure 2. Loss of Hdac3 alters chromatin structure and decreases heterochromatin

A. Histological sections prepared from control and Hdac3-null livers at postnatal day 17 (p17) were used to examine nuclei using electron microscopy. Control hepatocytes contain dense staining of condensed chromatin whereas Hdac3-null cells have decreased amounts of heterochromatin, especially at the nuclear periphery. Arrows indicate the RNA rich nucleoli that remain in the Hdac3-null cells. Scale bar represents 4 µm. B. Loss of heterochromatic foci upon inactivation of Hdac3. Left panels show 2 examples of histological sections from control and Hdac3-null liver sections stained with Hoechst to detect heterochromatic foci. Scale bar represents 20µm. The graph at the right shows quantification of foci from at least 100 cells for each sample expressed as the mean ± SEM. C. HP1β localization to chromatin is reduced in Hdac3-null livers. Chromatin-containing fractions were isolated by cell fractionation and whole cell lysates (WCL) or chromatin fractions (Chrom.) assessed using immunoblot analysis for HP1β, or histone H3 or β-actin as loading controls. Graph shows quantification of the ratio of HP1β to H3 on chromatin from 3 independent experiments expressed as the mean ± SEM. D. Nuclei prepared from Alb-Cre;Hdac3^+/− control and Alb-Cre;Hdac3^−/− mice were digested with increasing concentrations of micrococcal nuclease (MNase, MN) and genomic DNA was analyzed using agarose gel electrophoresis. The position of size markers are shown at the left. E. Nucleosome integrity is reduced in Hdac3^−/− hepatocytes. Cells from Alb-Cre;Hdac3^+/− control and Alb-Cre;Hdac3^−/− mice were fractionated and nuclei extracted with buffer containing the indicated amounts of NaCl (M). Upper panels show the soluble histone H3 and lower panels the histone remaining in the chromatin pellet. See also figure S2.

Figure 3. Deletion of Hdac3 increases H4K5ac, H4K12ac, and H3K9,K14ac

Western blot analysis of nuclear extracts prepared from p17 Hdac3-null hepatocytes to examine histone acetylation and histone methylation levels. H3 and H4 served as loading controls.

Figure 4. Inactivation of Hdac3 increases H4K5 and H4K12 acetylation in synchronized cells

A. Wild-type NIH 3T3 cells (control) or Hdac3^FL/− NIH 3T3 cells infected with Ad-Cre for 48hr (Hdac3-null) were cultured in 0.5% serum-containing media for 48hr. Cells were then transferred to media containing 10% fetal calf serum, lysates prepared at the times indicated, and analyzed by Western blot to measure the levels of the indicated histone modifications. GAPDH served as a loading control. B. Immunofluorescence analysis of H4K5ac in S-phase cells following release of serum starved G₀/G₁ cells into regular media for 18hr. Late S-phase cells were identified by the punctate pattern of PCNA staining at the nuclear periphery. Zoomed images of individual nuclei shown and scale bar represents 10µm. C. Immunoprecipitation of quiescent control or Hdac3-null lysates with anti-H4K5ac and western blot analysis with anti-H4K5ac and anti-H4. P1, 1^st immunoprecipitation; P2, re-immunoprecipitation of the supernatant from P1; S, supernatant from P2. See also figure S3.

Figure 5. Loss of Hdac3 causes genomic instability

A. Metaphase spreads prepared from ER-Cre:Hdac3^FL/+ and ER-Cre:Hdac3^FL/− MEFs treated with either vehicle (top panel) or tamoxifen (bottom panel). A magnified view of a group of chromosomes outlined in the left panel is shown in the right panel. Arrows indicate broken pieces of chromosomes. B. The number of breaks and gaps observed in control or null MEFs (Ad-Cre or tamoxifen treated ER-Cre:Hdac3^FL/−) were quantified and the data in the graph represent the mean ± S.D. The number of breaks and gaps per cell were calculated from two different MEF preparations, in which a total of 50 cells were counted in each preparation. C. DNA repair is impaired in Hdac3-null livers. Hdac3-null hepatocytes are defective in DNA repair. Mice were irradiated with a 3Gy dose of IR and frozen sections of livers collected immediately or 1 hr and 6 hr later were prepared for immunofluorescence analysis of γH2AX and 53BP1. Arrows indicate Hdac3-null nuclei with 53BP1 foci. Scale bar represents 20µm. See also figure S4.

Figure 6. Loss of Hdac3 leads to hepatocellular carcinoma

A Representative livers of 5-month (middle panel) and 10-month (right panel) old Alb-Cre:Hdac3^−/− mice. B. Survival plot for Alb-Cre:Hdac3^−/− mice. Heterozygous mice showed no mortality within this time frame. C. Immunohistochemistry using anti-Hdac3 shows that normal hepatocytes express Hdac3 (left panel), whereas Alb-Cre:Hdac3^−/− mice lack Hdac3 both in the tumor and surrounding tissue. T, tumor; L, liver. Scale bar represents 60µm. D. Hematoxylin and eosin stained histological sections (H&E, top panels) and immunohistochemistry for Ki67 (bottom panels) from 10-month old Alb-Cre:Hdac3^−/− mice. Scale bar represents 60µm. See also figure S5.

Figure 7. β-catenin is mis-regulated in Hdac3-null HCC

A. Heat map of selected genes from a cDNA microarray analysis of control liver and Hdac3-null HCCs. The levels of mRNAs expressed from genes associated with the Ras, p53, and Wnt pathways are depicted as green (low) or red (high), where black indicates no change. B. Quantitative RT-PCR confirmation of the microarray results. The graph shows the expression levels of the indicated genes obtained on the microarrays and from QRT-PCR as the average fold increase or decrease over controls that were set to 1 ± SD. C. β-catenin is mis-localized in Hdac3-null livers. Control and p28 Hdac3-null hepatocytes were separated into cytoplasmic (C) and nuclear (N) fractions and β-catenin was detected by immunoblot. Tubulin was used to monitor cytoplasmic contamination of nuclei. D. β-catenin is nuclear in Hdac3-null HCC. Immunohistochemistry was used to determine the cellular localization of β-catenin (brown tint). Nuclei were counterstained with hematoxylin (blue tint). Arrows indicate cells with prominent nuclear β-catenin. Scale bar represents 20 µm. See also table S1.

Figure 8. NCOR1 is down regulated in human HCC and siRNA targeting of NCoR and SMRT causes DNA damage

A. Heat map representation of the analysis of NCOR1, NCOR2, Hdac3 and Hdac1 mRNA levels in Human HCC samples (GEO 5975 dataset). Green depicts low expression relative to the mean of control samples and red indicates higher expression. A 2-fold cut-off was used to define a significant change. B. NCOR1 is down regulated in a subset of human HCC. Immunohistochemical staining was used to detect NCOR1 in human HCC or normal matched surrounding tissue (brown stain). Cells were counterstained with hematoxylin (blue tint). Scale bar represents 20µm. C. HeLa cells were transfected with either non-targeting (NT) or NCoR/SMRT siRNA (N/S). Whole cell lysates were analyzed for histone modifications using western blot. D. Immunofluorescence analysis of H4K5 acetylation in HeLa cells transfected with either non-targeting (NT) or NCoR/SMRT siRNA (N/S). Scale bar represents 10µm. S-phase cells were identified by the punctate pattern of PCNA staining. E. Immunofluorescence analysis of 53BP1 in HeLa cells transfected with either non-targeting siRNA (NT) or NCoR/SMRT siRNA (N/S). Scale bar represents 10µm. See also figure S6.