The histone deacetylase inhibitor sodium valproate causes limited transcriptional change in mouse embryonic stem cells but selectively overrides Polycomb-mediated Hoxb silencing
© Boudadi et al.; licensee BioMed Central Ltd. 2013
Received: 25 January 2013
Accepted: 16 April 2013
Published: 1 May 2013
Histone deacetylase inhibitors (HDACi) cause histone hyperacetylation and H3K4 hypermethylation in various cell types. They find clinical application as anti-epileptics and chemotherapeutic agents, but the pathways through which they operate remain unclear. Surprisingly, changes in gene expression caused by HDACi are often limited in extent and can be positive or negative. Here we have explored the ability of the clinically important HDACi valproic acid (VPA) to alter histone modification and gene expression, both globally and at specific genes, in mouse embryonic stem (ES) cells.
Microarray expression analysis of ES cells exposed to VPA (1 mM, 8 h), showed that only 2.4% of genes showed a significant, >1.5-fold transcriptional change. Of these, 33% were down-regulated. There was no correlation between gene expression and VPA-induced changes in histone acetylation or H3K4 methylation at gene promoters, which were usually minimal. In contrast, all Hoxb genes showed increased levels of H3K9ac after exposure to VPA, but much less change in other modifications showing bulk increases. VPA-induced changes were lost within 24 h of inhibitor removal. VPA significantly increased the low transcription of Hoxb4 and Hoxb7, but not other Hoxb genes. Expression of Hoxb genes increased in ES cells lacking functional Polycomb silencing complexes PRC1 and PRC2. Surprisingly, VPA caused no further increase in Hoxb transcription in these cells, except for Hoxb1, whose expression increased several fold. Retinoic acid (RA) increased transcription of all Hoxb genes in differentiating ES cells within 24 h, but thereafter transcription remained the same, increased progressively or fell progressively in a locus-specific manner.
Hoxb genes in ES cells are unusual in being sensitive to VPA, with effects on both cluster-wide and locus-specific processes. VPA increases H3K9ac at all Hoxb loci but significantly overrides PRC-mediated silencing only at Hoxb4 and Hoxb7. Hoxb1 is the only Hoxb gene that is further up-regulated by VPA in PRC-deficient cells. Our results demonstrate that VPA can exert both cluster-wide and locus-specific effects on Hoxb regulation.
KeywordsHoxb genes Valproic acid Histone deacetylase Polycomb repression Mouse embryonic stem cells Histone modification Microarray expression analysis Retinoic acid Transcriptional activation
Histone deacetylase inhibitors (HDACi) have long been known to cause global histone hyperacetylation, often accompanied by increased H3K4 methylation, in a variety of model systems ( and references therein). Two structurally unrelated HDACi, suberoylanilide hydroxamic acid (SAHA) and depsipeptide (a bicyclic peptide) are remarkably effective against cutaneous T-cell lymphoma (CTCL) [2, 3] and have been Food and Drug Administration (FDA) approved for treatment of this cancer (Additional file 1: Table S1). HDACi have great potential as chemotherapeutic agents, prompting searches for new HDACi and a growing number of trials against various cancers [4, 5]. A major barrier to improving the clinical effectiveness of HDACi is that their mechanisms of action are varied and complex, and generally not well understood (discussed in ). There are at least six different structural classes of HDACi, four of which are in clinical trials (Additional file 1: Table S1). All exert multiple effects on cell function, including induction of differentiation, cell cycle disruption and apoptotic death [5, 6]. The situation is further complicated by the fact that there are 18 different histone deacetylases (HDACs) in human cells, split into four classes [5, 7]. Eleven of these enzymes, classes I, IIa, IIb and IV, have a very similar catalytic site, but differ in subtle ways in their sensitivities to HDACi (Additional file 1: Table S1) . Class III enzymes, the sirtuins, are NAD-dependent and are insensitive to all classes of HDACi in clinical use . In addition, HDACs, despite their name, act on a variety of proteins in addition to histones , including transcription factors, enzymes and HDACs themselves . They usually operate in vivo as part of multi-protein complexes, the composition of which can influence their catalytic activity, their location within the cell and their targeting to specific genes [7, 9].
Valproic acid (VPA) is a branched, short-chain fatty acid that inhibits class I and IIa HDACs, most likely through binding to the catalytic site . VPA has been used clinically for many years as an anti-epileptic agent and mood stabiliser, usually as the sodium salt [11, 12]. Because it is well tolerated and has been shown to induce differentiation and apoptosis of carcinoma cells, it has recently been tested in clinical trials as a potential chemotherapeutic agent for a variety of cancers [4, 13]. One long-appreciated side effect of VPA is its teratogenicity, causing problems for those women who must depend on it during pregnancy because alternative drugs are ineffective or unavailable [14, 15]. Teratogenic effects include musculoskeletal, neurological and behavioural aberrations, leading to identification of a distinctive valproate syndrome [15–17]. VPA acts on class I and IIa HDACs, which collectively deacetylate a variety of nuclear and cytoplasmic proteins, so its effects are inevitably pleiotropic. For example, VPA induces oxidative stress, with increased levels of reactive oxygen species, and anti-oxidants can alleviate some of its teratogenic effects [18, 19].
Surprisingly, global changes in histone modification induced by HDACi are usually not accompanied by changed levels at individual genes, as measured by chromatin immunoprecipitation (ChIP). In fact, butyrate has been reported to cause an unexpected decrease in histone acetylation at some transcription start sites in HepG2 and HT29 cells, with associated down-regulation of transcription . A wide ranging ChIP-seq study of quiescent human T-cells showed that a combination of butyrate and Trichostatin A (TSA) at high concentrations induced acetylation primarily at the promoters of active genes and of the small proportion of silent genes that showed high levels of H3K4 methylation . In human lymphoblastoid cells we found that only a small proportion of genes showed altered transcription after treatment with VPA. Of these, almost as many were down-regulated as were up-regulated , consistent with early findings of a surprising lack of transcriptional stimulation by HDACi . Nor did we find any consistent increase in histone acetylation or H3K4 methylation at selected promoter regions, leading to the suggestion that many genes are sheltered from the global effects of HDACi .
The complex relationship between the histone modifications induced by HDACi and the transcription of individual genes reflects the general uncertainty surrounding the role of histone modifications in transcriptional control. The recent achievements in genome-wide mapping of chromatin marks have confirmed that some histone modifications are usually associated with transcriptionally active, or potentially active, genes (for example, H3K9ac, H3K4me3), while others are typical of silent genes (for example, H3K9me2/3, H3K27me3) [24, 25]. However, such data cannot establish the cause-effect relationship between histone modification and gene activity. It is important to establish for particular loci in particular developmental contexts, whether a histone modification can by itself drive a change in chromatin function, whether the modification is part of a chain of events by which chromatin transitions from one functional state to another, or whether the modification is simply a by-product of other processes and of no functional consequence in its own right [26, 27].
In the experiments reported here, we have tested the ability of VPA to alter patterns of gene expression in mouse embryonic stem (ES) cells, looking in detail at its effects on members of the Hoxb cluster. The Hox (Homeobox) genes encode highly conserved transcription factors that act as master regulators of development and whose mis-expression has been implicated in a variety of human diseases, including cancer [28, 29], and possibly in the teratogenicity of VPA . In vertebrates, 39 Hox genes are divided between four clusters (A, B, C and D) located on different chromosomes. They are silent or very weakly expressed in the pre-implantation mammalian embryo and in ES cells, but become activated during development along the antero-posterior axis according to their positions in the cluster . Previous work has indicated that the chromatin that packages the Hoxb genes in embryonic cells can be influenced by HDAC inhibitors [31, 32]. Here, we use them as a model system to explore the relationship between changes in histone modifications and transcription as genes switch from silenced to active states.
VPA causes global histone modification changes in ES cells but has little effect on gene expression or differentiation state
Ontology analysis of this dataset, using DAVID software, revealed several annotations to be significantly over-represented amongst genes whose activities were altered by VPA. The most significant are listed in Additional file 1: Table S2. The highest scoring annotation was Cytoskeleton and cytoskeleton-interacting proteins (P = 5.13 × 10-5). Cell cycle (P = 6.76 × 10-4) and chromatin organisation (P = 6.46 × 10-3) were also featured. The effect on cytoskeleton-related genes is intriguing in light of the morphological changes induced by VPA. However the short timescale involved and the reversibility of the effect makes it likely that morphological changes were the result of a combined effect of altered acetylation of cytoskeletal proteins and gene expression changes .
VPA-induced changes in transcription at specific genes are not consistently linked to changes in histone acetylation
Hoxb genes show increased H3 acetylation following VPA treatment, but this is associated with increased transcription only at certain loci
In pre-implantation mouse embryos in culture, VPA-induced increases in both histone acetylation and H3K4me3 were heritable through mitosis in the absence of the inhibitor, at least until the blastocyst stage . In contrast, in ES cells the VPA-induced increases in H3K9ac were consistently lost within 24 h of removal of the inhibitor. A typical experiment is shown in Additional file 2: Figure S2. It will be interesting to explore possible reasons for the different behavior of pre-blastocyst embryos and cultured cells (ES cells) derived from the Inner Cell Mass of the blastocyst.
In view of these findings, it is interesting to note that of three genes showing strongly increased transcription after VPA (Figure 3C), two showed relatively high levels of H3K27me3 (H1f0, Egr1), while one (Ndrg4) did not (Figure 4). Thus, Polycomb-mediated silencing is not a prerequisite for VPA-induced increases in either transcription or H3K9ac and cannot, in itself, explain the characteristic and unusual response of Hoxb genes to this inhibitor.
Hoxb genes are all induced by retinoic acid in differentiating ES cells but longer-term expression varies from gene to gene
Loss of Polycomb components generally increases Hoxb gene expression and modulates the transcriptional response to VPA in a locus-specific manner
To test the role of Polycomb in silencing Hoxb in undifferentiated ES cells, we measured expression of Hoxb genes in an ES cell line in which both PRC1 and PRC2 were non-functional . Total lack of expression of Ring1b (PRC1) and Eed (PRC2) in these dKO cells was confirmed by qPCR (Additional file 2: Figure S3). Loss of Polycomb components in those cells led to a strong diminution of the H3K27me3 mark, as expected, and a corresponding increase in global H3K27ac (Additional file 2: Figure S3).
The comparatively weak activation of Hoxb1 in response to Polycomb depletion (Figure 9A), its strong response to VPA in Polycomb null cells (Figure 9B) and the unique diminution of RA-induced Hoxb1 expression in differentiating cells over time (Figure 7), all suggest that Hoxb1 is subject to a regulatory regime that differs from other Hoxb genes. Hoxb1 is distinguished by its proximity to two RA response elements (RAREs), DR2 and DR5, which lie 1.5Kb and 6.5Kb respectively, 3′ of the Hoxb1 coding region. We find both RAREs in CCE/R ES cells are strongly enriched in the silencing mark H3K27me3, with below average levels of H3K27ac, H3K9ac, H4K8ac and H3K4me3 (Figure 5C). VPA treatment caused no changes in the modifications tested at DR5, but DR2 showed an approximately 2-fold increase in H3K9ac and H3K27ac, along with a small diminution of H3K27me3 (Figure 5C). These changes, though modest, may contribute to the distinctive transcriptional responses of Hoxb1.
ES cells treated with VPA showed dramatic global changes in histone modification, including increased acetylation of all histones and H3K4 methylation, but surprisingly limited changes in both transcript levels and histone modification levels at individual promoters. In line with the limited transcriptional response, we found only very occasional evidence for genes showing altered histone acetylation or H3K4 methylation following growth in VPA, even among those genes showing the strongest changes in transcription. This discrepancy between global hypermodification and minimal local modification changes is consistent with our recent studies on human lymphoblastoid cell lines , and suggests that in ES cells, as in other cell types, many genes are sheltered from the global changes in histone modification induced by environmental inhibitors. This may also help explain the low frequency of VPA-induced changes in gene expression. Following completion of the microarray experiments presented here, Hezroni and colleagues  published data showing limited transcriptional change following treatment of ES cells with VPA. Comparison of our data with the nearest equivalent set from Hezroni and colleagues, shows a strong overlap, despite differences in treatment time and concentration (0.5 mM VPA for 4 h vs 1 mM for 8 h). Of our 241 up-regulated (>1.5-fold) genes, 52 were classified as up-regulated by Hezroni et al. . The overlap for down-regulated genes was less striking at 118 and 18 respectively. Genes in common are listed in Additional file 1: Table S3. Ontology analysis (DAVID) of these common genes showed that enrichment scores >1.3 and P- values <0.01 were restricted to the terms, steroid metabolism/homeostasis, protein transport/localization and negative regulation of differentiation/neurogenesis.
A discrepancy between global changes in histone modification and local transcriptional effects has been noted in ES cells depleted in DPY30, a key component of the MLL2 histone methyltransferase complex . Dpy30 -/- cells showed minimal transcription changes, despite major global and local reductions in H3K4 methylation . However, DPY30 depleted cells showed reduction or absence of expression of lineage-specific genes following RA treatment and compromised neuronal differentiation .
In quiescent human CD4+ T-cells, hyperacetylation induced by the inhibition of HDACs with a mixture of butyrate and TSA, occurred on only a minority of genes, primarily those that were transcriptionally active . Among silent genes, a subset of less than 15% was found to be susceptible to hyperacetylation in the presence of HDACi. The great majority of this subset were marked by elevated levels of H3K4 methylation and could therefore be categorised as bivalent, poised genes [42, 43]. Transcriptionally active genes are generally enriched in H3K4 methylation, particularly at their promoters , and it may be that levels of H3K4me3 are a more important determinant of sensitivity to HDACi than transcription per se. In contrast, Hezroni et al.  found only modest local changes in H3K9ac in ES cells treated with 0.5 mM VPA (less than 20% overall after 4 h treatment); there was variation between genes but no evidence that acetylation was selectively increased at regions classed as bivalent .
Our results illustrate how the Hoxb genes in ES cells are regulated by a combination of cluster-wide and locus-specific mechanisms. Thus, all Hoxb genes are kept in a silent state, at least in part, by Polycomb silencing complexes, being marked by a blanket of the Polycomb-associated histone modification H3K27me3 . Similarly, all Hoxb genes are activated when differentiating cells are exposed to RA (see  and references therein). Remarkably, there is evidence that a binding site for RA receptor gamma in the Hoxa1 upstream enhancer is a key player in cluster-wide activation of Hoxb genes . Locus-specific effects became apparent when the time-course of the RA response was followed, with clear differences in the times at which individual Hoxb genes are maximally expressed, from day 3 to day 8 of differentiation. In contrast to in vivo data [30, 47], in ES cells grown and RA-treated in culture, we found no clear correlation between the timing of maximal expression (transcript level) of individual Hoxb genes and their position in the cluster. The clear-cut and consistent differences in timing show that regulatory mechanisms acting at the level of individual loci, or groups of loci, can moderate the cluster-wide effect triggered by RA.
All Hoxb loci showed increased H3K9 acetylation in the presence of the HDACi, VPA. VPA consistently increased levels of H3K9ac at upstream, promoter and coding regions across the entire Hoxb cluster. H3K27ac and H3K4me3 both showed more modest, though occasionally significant, VPA-induced increases. Perhaps surprisingly, H4K8ac, a modification that shows a global increase after VPA, remained unchanged at all Hoxb loci tested. It is interesting that levels of H3K27me3 were consistently lower across Hoxb genes in VPA-treated cells, though the reductions were always small and statistically insignificant for any single locus. The HDACi-induced increase in H3K9ac at Hoxb loci shows that H3 acetate groups are turning over at these loci, which must therefore contain both HDACs and histone acetyltransferases (HATs). It will be interesting to identify these enzymes. In human T-cells, regions that were sensitive to HDACi-induced hyperacetylation have not always been found to be rich in HDACs, leading to the suggestion that the association of HDACs with such regions was transient and unstable . As the specificity of HDACs for particular histone types tends to be low [9, 48], it is likely that the absence of VPA-induced acetylation of certain H3 and H4 lysines across Hoxb loci reflects the absence of catalytically-active HATs of the necessary specificity, rather than variations in HDAC levels or types. A strong overlap has been noted in mouse ES cells between peaks of H3K9ac and the HAT p300 . However, the significance of this is uncertain and much remains to be learned about the in vivo specificity of the various HAT complexes. Knock-out experiments have shown that depletion of CBP/p300 specifically reduces acetylation at H3K18 and H3K27, while GCN5 depletion selectively reduces H3K9ac .
It is interesting that only Hoxb4 and Hoxb7 showed a significant increase in transcription following VPA treatment of wild-type ES cells. As VPA treatment increased H3K9ac across Hoxb4 and Hoxb7, and across Hoxb1, 2 and 5, where transcription does not change, we can conclude that increased H3K9ac levels are not caused by increased transcription, nor are they sufficient, in themselves, to trigger transcription in all Hoxb genes. Of the modifications tested, only H3K27ac showed VPA-induced increases selectively at Hoxb4 and 7, thus correlating with changes in transcription. It has been suggested previously that H3K27ac might exert its effects by displacing, and mitigating the silencing effects of, H3K27me3 , and such a process is consistent with our data. The results so far also indicate that increased transcription of Hoxb genes in the presence of continued Polycomb repressor complex (PRC) silencing can only be induced at loci that are pre-marked by higher than average levels of H3K4me3, H3K9ac and H3K27ac. Perhaps these modifications are part of a permissive chromatin environment that facilitates VPA-induced transcription. Our results show that a high level of H3K4me3 is not, in itself, necessary for the VPA-sensitive chromatin state at all Hoxb loci, though it may play a locus-specific role, perhaps in facilitating the up-regulation of transcription in response to HDACi .
As expected, the absence of functional PRC complexes in Eed/Ring1b dKO cells led to increased Hoxb transcription. However, the increase varied from 8-fold and highly significant for Hoxb7, to small and insignificant for Hoxb1 and Hoxb5. Interestingly, removal of Polycomb did not make the Hoxb genes more susceptible to the effects of HDACi-induced acetylation. With the exception of Hoxb1, treatment of dKO cells with VPA caused little further increase in Hoxb expression. Hoxb1, for which there was a relatively small (not significant) increase in transcription following PRC inactivation, showed a robust increase when dKO cells were treated with VPA. Collectively, these findings suggest that increased H3K9ac acetylation (and perhaps acetylation at other lysines not yet tested) induced by VPA can partially override PRC-mediated silencing of some Hoxb genes (particularly Hoxb4 and Hoxb7), but causes no further increase in transcription in the absence of PRC. Thus, with the exception of Hoxb1, an increase in activating marks even in the absence of Polycomb-associated silencing, does not, in itself, induce transcription. The results emphasize that Hoxb1 has regulatory features that distinguish it from other genes in the Hoxb cluster. It may be significant that chromatin opening, perhaps progressing 3′ to 5′ through the cluster, has been shown to be important for Hox gene activation [31, 51]. If Hoxb1 chromatin is selectively decondensed in PRC-deficient cells, it could be more sensitive to VPA-induced acetylation, perhaps because this facilitates the transient attachment of HATs and HDACs .
Gene silencing is a multi-layered process, involving the combinatorial actions of histone and DNA modifications, chromatin condensation, DNA binding proteins, small RNAs and the intra-nuclear location of genes and gene clusters. Gene activation involves stripping away some or all of these silencing components and may not be an all or nothing process. As we have seen, the Hoxb genes can be significantly up-regulated by treatment with VPA, or by removal of the Polycomb complex, though the level of activation is far below that seen following induction of differentiation and treatment with RA. In contrast, most genes in ES cells are insensitive to VPA treatment, and even within the Hoxb cluster, the response of Hoxb1 sets it apart from the other Hoxb genes. Definitive conclusions about mechanisms of action of HDAC inhibitors are made difficult by the wide range of HDAC substrates, including both histones and non-histone proteins, but of the modifications tested, H3K9ac and H3K27ac are associated in both general and locus-specific ways with Hoxb regulation. These related modifications could provide both a context and the final trigger by which VPA activates Hoxb expression, and it is important now to define the regulatory pathways by which they are put in place and through which they operate.
The HDACi, valproic acid, induces global histone hyperacetylation and increased H3K4 methylation in mouse ES cells, but significantly alters the expression of only a small proportion of genes, of which approximately one third are down-regulated.
There is no correlation between changes in expression and levels of H3K9ac, H4K8ac or H3K4me3 at gene promoters, which overall show little sign of VPA-induced change in histone modifications. It seems that most genes in mouse ES cells are sheltered from the global increases in histone modification induced by VPA.
Hoxb genes were exceptional in showing consistently increased H3K9ac at promoter and intragenic regions in cells exposed to VPA. H3K27ac and H3K4me3 showed smaller increases at only some Hoxb loci while H4K8ac was unaffected. VPA consistently caused small reductions in H3K27me3, though these never achieved statistical significance for any one locus. Only Hoxb4 and Hoxb7 showed significantly increased transcription after VPA treatment. These two loci have the highest levels of H3K9ac, H3K27ac and H3K4me3 in untreated cells and show strong increases in acetylation in response to VPA.
Inactivation of PRC in Eed/Ring1b dKO cells, activated Hoxb genes several-fold, with the exception of Hoxb1 (only weakly activated) and Hoxb5, which remained essentially silent. Uniquely, Hoxb1 activity in PRC-negative cells was increased several-fold by VPA. Thus, Hoxb genes respond to VPA in a Polycomb-dependent and locus-specific manner.
RA increased transcription of all Hoxb genes within 24 h, with an accompanying increase in H3K9ac, but expression thereafter, up to day 8, varied between Hoxb loci. In differentiating ES cells there is no clear relationship between the change in Hoxb expression over time and 3′ to 5′ position within the cluster.
Control of Hoxb genes is multilayered. Both PRC-mediated silencing, with high levels of H3K27me3, and RA activation, affect all Hox genes in the cluster. However, these overall control mechanisms are modulated in different ways at individual Hoxb loci. H3K9ac, H3K27ac and H3K4me3 may all help provide a VPA-responsive chromatin environment at specific Hoxb loci, but H4K8ac seems not to be involved.
Primary cells and cell culture
The male mouse feeder-independent 129/Sv-derived ES cell line, CCE/R , was grown on 0.1% gelatin-coated flasks and 129S4/SvJae-derived ES cell line, J1, and the Eed/Ring1B dKO, PRC1/PRC2-deficient) daughter cell line were grown on mitotically inactivated mouse embryonic fibroblast (MEF) feeder cells . All ES cell lines were cultured in DMEM supplemented with 20% foetal calf serum, 1× penicillin/streptomycin, 1× L-glutamine, 1× non-essential amino acids, 0.25% 2-mercaptoethanol and 1U/μl ESGRO (Millipore, Watford, UK). All supplies were sourced from Invitrogen (Paisley, UK) unless otherwise noted.
For RA-induced differentiation, ES cells were plated on non-adherent plastic dishes in ES cell media minus ESGRO with addition of 1μM all-trans-RA on day 2 of differentiation . Cell viability was assessed by Trypan Blue staining and by the MTS assay (Promega, Southampton, UK), a calorimetric assay to detect metabolically active cells (used according to the manufacturer’s instructions). ES pluripotency was tested by AP staining of undifferentiated fixed adherent cells [53, 54]. Cell cycle analysis was performed using PI (10 μg/ml) staining in the presence of RNaseA (200 mg/ml) using a FACScan flow cytometer (BD FACSCalibur™, Becton Dickinson, Oxford, UK) and analysed using WinMDi 2.9 and Cylchred software.
Rabbit polyclonal antisera to H4K8ac, H3K9ac, H3K4me3 and H2BK12/15 ac were raised in-house by immunisation with synthetic peptides conjugated to ovalbumin as previously described . Sources of other antisera are shown in Additional file 1: Table S4. Specificities were assayed by inhibition ELISA for all in-house and commercial antisera  and checked by western blotting.
Histones were extracted from ES cells by acid extraction and analysed by electrophoresis in 15% SDS-polyacrylamide gels and western blotting as previously described  Protein loading was tested by Ponceau S staining before proteins were probed with appropriate primary antibodies. Primary antibody binding was detected by fluorescent tagged anti-rabbit IgG secondary antibodies (Rockland) and detected by scanning (Odyssey system; LI-COR, Cambridge, UK).
Native chromatin immunoprecipitation (NChIP)
Immunoprecipitation of native chromatin was performed based on the method described previously  and DNA was analysed using qPCR. Briefly, cells were lysed to release nuclei prior to micrococcal nuclease digestion. The amount of micrococcal nuclease added and digestion time were adjusted to obtain a mix of mono- and short oligo-nucleosomes, optimal for immunoprecipitation . Digested chromatin was immunoprecipitated and bound (B) and unbound (UB) fractions were separated. DNA from B and UB fractions was isolated and analysed by qPCR using region specific primers. Sequences of all ChIP-qPCR primers are listed in Additional file 1: Tables S5-S7.
Gene expression analysis
Total RNA was extracted from cells using the RNeasy mini kit (Qiagen, Crawley, UK) according to the manufacturer’s instruction. An on-column DNase digestion was performed during the extraction in order to yield a pure RNA fraction (RNase-Free DNase Set, Qiagen). cDNA was prepared using the RT-Superscript-III Kit (Invitrogen) according to manufacturer’s instructions. cDNA was then analysed using qPCR or microarrays.
For each RT-qPCR assay, 50 ng of cDNA were mixed with 5μl of SYBR green Mix (Qiagen), forward and reverse primers (0.4 μM each) and RNase-free water in a final volume of 10 μl. The PCR reaction was run on the ABI 7900HT system (95°C for 10 minutes; 40 cycles of 94°C for 15 s, primer set-specific annealing temperature for 30 s 72°C for 30 s, with a final denaturation step 95°C for 15 s, 60°C for 15 s, 95°C for 15 s). Standard curves made from serial dilutions of cDNA allowed relative quantification. Sequences of all primers and Tm used for qPCR are listed in Additional file 1: Table S8.
For microarray expression analysis we used the NIA 15 K mouse cDNA array  as previously described . cDNA samples to be compared were labelled with two different fluorochromes, cy3 and cy5 (Amersham, Little Chalfont, UK) using a Bioprime Labelling kit (Invitrogen). Following their purification (PCR purification Kit; Qiagen), equal amounts of probes were combined and precipitated before hybridization onto slides at 42°C for 20 h. Slides were scanned using GenePix 4000A scanner. Spot alignment was checked automatically and then manually using GenePix software. Data were normalised using the DNMAD web-software. T-tests were performed on total M-values using biological triplicates for each condition and standard P-values were then corrected by the Benjamini-Hochberg FDR method using the statistical environment “R”. Finally, genes with associated FDR less than 10% were considered statistically significant and selected for further analysis. Original microarray data files are available through the GEO database [GSE45610].
cutaneous T-cell lymphoma
double knockout (Eed/Ring1b negative)
Dulbecco’s modified Eagles medium
enzyme-linked immunosorbent assay
fluorescence activated cell sorter
Food and Drug Administration
False Discovery Rate
histone deacetylase inhibitor
leukaemia inhibitory factor
mouse embryonic fibroblast
Polycomb repressor complex
retinoic acid response element
real time quantitative polymerase chain reaction
suberoylanilide hydroxamic acid
valproic acid (administered as the sodium salt)
EB was supported by a studentship from the Epigenome Network of Excellence (FP7) and HS by a studentship from the UK Medical Research Council. The Chromatin and Gene Expression Group is funded by grants from Cancer Research UK. AW was supported by a Wellcome Trust Senior Research Fellowship (grant reference 087530/Z/08/A).
- Nightingale KP, Gendreizig S, White DA, Bradbury C, Hollfelder F, Turner BM: Cross-talk between histone modifications in response to histone deacetylase inhibitors: MLL4 links histone H3 acetylation and histone H3K4 methylation. J Biol Chem. 2007, 282: 4408-4416.View ArticlePubMedGoogle Scholar
- Marks PA, Xu WS: Histone deacetylase inhibitors: potential in cancer therapy. J Cell Biochem. 2009, 107: 600-608. 10.1002/jcb.22185.PubMed CentralView ArticlePubMedGoogle Scholar
- StatBite: FDA oncology drug product approvals in 2009. J Natl Cancer Inst. 2009, 102 (4): 219.Google Scholar
- Batty N, Malouf GG, Issa JP: Histone deacetylase inhibitors as anti-neoplastic agents. Cancer Lett. 2009, 280: 192-200. 10.1016/j.canlet.2009.03.013.View ArticlePubMedGoogle Scholar
- Mercurio C, Minucci S, Pelicci PG: Histone deacetylases and epigenetic therapies of hematological malignancies. Pharmacol Res. 2010, 62: 18-34. 10.1016/j.phrs.2010.02.010.View ArticlePubMedGoogle Scholar
- Xu WS, Parmigiani RB, Marks PA: Histone deacetylase inhibitors: molecular mechanisms of action. Oncogene. 2007, 26: 5541-5552. 10.1038/sj.onc.1210620.View ArticlePubMedGoogle Scholar
- Haberland M, Montgomery RL, Olson EN: The many roles of histone deacetylases in development and physiology: implications for disease and therapy. Nat Rev Genet. 2009, 10: 32-42. 10.1038/nrg2485.PubMed CentralView ArticlePubMedGoogle Scholar
- Haigis MC, Guarente LP: Mammalian sirtuins–emerging roles in physiology, aging, and calorie restriction. Genes Dev. 2006, 20: 2913-2921. 10.1101/gad.1467506.View ArticlePubMedGoogle Scholar
- Marmorstein R, Trievel RC: Histone modifying enzymes: structures, mechanisms, and specificities. Biochim Biophys Acta. 2009, 1789: 58-68. 10.1016/j.bbagrm.2008.07.009.PubMed CentralView ArticlePubMedGoogle Scholar
- Qiu Y, Zhao Y, Becker M, John S, Parekh BS, Huang S, Hendarwanto A, Martinez ED, Chen Y, Lu H, Adkins NL, Stavreva DA, Wiench M, Georgel PT, Schiltz RL, Hager GL: HDAC1 acetylation is linked to progressive modulation of steroid receptor-induced gene transcription. Mol Cell. 2006, 22: 669-679. 10.1016/j.molcel.2006.04.019.View ArticlePubMedGoogle Scholar
- Phiel CJ, Zhang F, Huang EY, Guenther MG, Lazar MA, Klein PS: Histone deacetylase is a direct target of valproic acid, a potent anticonvulsant, mood stabilizer, and teratogen. J Biol Chem. 2001, 276: 36734-36741. 10.1074/jbc.M101287200.View ArticlePubMedGoogle Scholar
- Gerstner T, Bell N, Konig S: Oral valproic acid for epilepsy–long-term experience in therapy and side effects. Expert Opin Pharmacother. 2008, 9: 285-292. 10.1517/14656522.214.171.1245.View ArticlePubMedGoogle Scholar
- Duenas-Gonzalez A, Candelaria M, Perez-Plascencia C, Perez-Cardenas E, de la Cruz-Hernandez E, Herrera LA: Valproic acid as epigenetic cancer drug: preclinical, clinical and transcriptional effects on solid tumors. Cancer Treat Rev. 2008, 34: 206-222. 10.1016/j.ctrv.2007.11.003.View ArticlePubMedGoogle Scholar
- Alsdorf R, Wyszynski DF: Teratogenicity of sodium valproate. Expert Opin Drug Saf. 2005, 4: 345-353. 10.1517/147403126.96.36.1995.View ArticlePubMedGoogle Scholar
- Kluger BM, Meador KJ: Teratogenicity of antiepileptic medications. Semin Neurol. 2008, 28: 328-335. 10.1055/s-2008-1079337.PubMed CentralView ArticlePubMedGoogle Scholar
- Ardinger HH, Atkin JF, Blackston RD, Elsas LJ, Clarren SK, Livingstone S, Flannery DB, Pellock JM, Harrod MJ, Lammer EJ: Verification of the fetal valproate syndrome phenotype. Am J Med Genet. 1988, 29: 171-185. 10.1002/ajmg.1320290123.View ArticlePubMedGoogle Scholar
- Massa V, Cabrera RM, Menegola E, Giavini E, Finnell RH: Valproic acid-induced skeletal malformations: associated gene expression cascades. Pharmacogenet Genomics. 2005, 15: 787-800. 10.1097/01.fpc.0000170914.11898.3a.View ArticlePubMedGoogle Scholar
- Chuang CM, Chang CH, Wang HE, Chen KC, Peng CC, Hsieh CL, Peng RY: Valproic Acid Downregulates RBP4 and Elicits Hypervitaminosis A-Teratogenesis-A Kinetic Analysis on Retinol/Retinoic Acid Homeostatic System. PLoS One. 2012, 7: e43692-10.1371/journal.pone.0043692.PubMed CentralView ArticlePubMedGoogle Scholar
- Tung EW, Winn LM: Valproic acid increases formation of reactive oxygen species and induces apoptosis in postimplantation embryos: a role for oxidative stress in valproic acid-induced neural tube defects. Mol Pharmacol. 2011, 80: 979-987. 10.1124/mol.111.072314.View ArticlePubMedGoogle Scholar
- Rada-Iglesias A, Enroth S, Ameur A, Koch CM, Clelland GK, Respuela-Alonso P, Wilcox S, Dovey OM, Ellis PD, Langford CF, Dunham I, Komorowski J, Wadelius C: Butyrate mediates decrease of histone acetylation centered on transcription start sites and down-regulation of associated genes. Genome Res. 2007, 17: 708-719. 10.1101/gr.5540007.PubMed CentralView ArticlePubMedGoogle Scholar
- Wang Z, Zang C, Cui K, Schones DE, Barski A, Peng W, Zhao K: Genome-wide mapping of HATs and HDACs reveals distinct functions in active and inactive genes. Cell. 2009, 138: 1019-1031. 10.1016/j.cell.2009.06.049.PubMed CentralView ArticlePubMedGoogle Scholar
- Halsall J, Gupta V, O’Neill LP, Turner BM, Nightingale KP: Genes are often sheltered from the global histone hyperacetylation induced by HDAC inhibitors. PLoS One. 2012, 7: e33453-10.1371/journal.pone.0033453.PubMed CentralView ArticlePubMedGoogle Scholar
- Van Lint C, Emiliani S, Verdin E: The expression of a small fraction of cellular genes is changed in response to histone hyperacetylation. Gene Expr. 1996, 5: 245-253.PubMedGoogle Scholar
- Dunham I, Kundaje A, Aldred SF, Collins PJ, Davis CA, Doyle F, Epstein CB, Frietze S, Harrow J, Kaul R, Khatun J, Lajoie BR, Landt SG, Lee BK, Pauli F, Rosenbloom KR, Sabo P, Safi A, Sanyal A, Shoresh N, Simon JM, Song L, Trinklein ND, Altshuler RC, Birney E, Brown JB, Cheng C, Djebali S, Dong X, Ernst J: An integrated encyclopedia of DNA elements in the human genome. Nature. 2012, 2012 (489): 57-74.View ArticleGoogle Scholar
- Bannister AJ, Kouzarides T: Regulation of chromatin by histone modifications. Cell Res. 2011, 21: 381-395. 10.1038/cr.2011.22.PubMed CentralView ArticlePubMedGoogle Scholar
- Henikoff S, Shilatifard A: Histone modification: cause or cog?. Trends Genet. 2011, 27: 389-396. 10.1016/j.tig.2011.06.006.View ArticlePubMedGoogle Scholar
- Turner BM: The adjustable nucleosome: an epigenetic signaling module. Trends Genet. 2012, 28: 436-444. 10.1016/j.tig.2012.04.003.View ArticlePubMedGoogle Scholar
- Faiella A, Wernig M, Consalez GG, Hostick U, Hofmann C, Hustert E, Boncinelli E, Balling R, Nadeau JH: A mouse model for valproate teratogenicity: parental effects, homeotic transformations, and altered HOX expression. Hum Mol Genet. 2000, 9: 227-236. 10.1093/hmg/9.2.227.View ArticlePubMedGoogle Scholar
- Grier DG, Thompson A, Kwasniewska A, McGonigle GJ, Halliday HL, Lappin TR: The pathophysiology of HOX genes and their role in cancer. J Pathol. 2005, 205: 154-171. 10.1002/path.1710.View ArticlePubMedGoogle Scholar
- Dolle P, Izpisua-Belmonte JC, Falkenstein H, Renucci A, Duboule D: Coordinate expression of the murine Hox-5 complex homoeobox-containing genes during limb pattern formation. Nature. 1989, 342: 767-772. 10.1038/342767a0.View ArticlePubMedGoogle Scholar
- Chambeyron S, Bickmore WA: Chromatin decondensation and nuclear reorganization of the HoxB locus upon induction of transcription. Genes Dev. 2004, 18: 1119-1130. 10.1101/gad.292104.PubMed CentralView ArticlePubMedGoogle Scholar
- VerMilyea MD, O’Neill LP, Turner BM: Transcription-independent heritability of induced histone modifications in the mouse preimplantation embryo. PLoS One. 2009, 4: e6086-10.1371/journal.pone.0006086.PubMed CentralView ArticlePubMedGoogle Scholar
- Kargul GJ, Dudekula DB, Qian Y, Lim MK, Jaradat SA, Tanaka TS, Carter MG, Ko MS: Verification and initial annotation of the NIA mouse 15K cDNA clone set. Nat Genet. 2001, 28: 17-18.PubMedGoogle Scholar
- Lin H, Gupta V, Vermilyea MD, Falciani F, Lee JT, O’Neill LP, Turner BM: Dosage compensation in the mouse balances up-regulation and silencing of X-linked genes. PLoS Biol. 2007, 5: e326-10.1371/journal.pbio.0050326.PubMed CentralView ArticlePubMedGoogle Scholar
- Lin H, Halsall JA, Antczak P, O’Neill LP, Falciani F, Turner BM: Relative overexpression of X-linked genes in mouse embryonic stem cells is consistent with Ohno’s hypothesis. Nat Genet. 2011, 43: 1169-1170. 10.1038/ng.992. author reply 1171–1162View ArticlePubMedGoogle Scholar
- Kim SC, Sprung R, Chen Y, Xu Y, Ball H, Pei J, Cheng T, Kho Y, Xiao H, Xiao L, Griskin NV, White M, Yang XJ, Zhao Y: Substrate and functional diversity of lysine acetylation revealed by a proteomics survey. Mol Cell. 2006, 23: 607-618. 10.1016/j.molcel.2006.06.026.View ArticlePubMedGoogle Scholar
- Wilson MA, Ricci AR, Deroo BJ, Archer TK: The histone deacetylase inhibitor trichostatin A blocks progesterone receptor-mediated transactivation of the mouse mammary tumor virus promoter in vivo. J Biol Chem. 2002, 277: 15171-15181.PubMedGoogle Scholar
- Jones PA, Archer TK, Baylin SB, Beck S, Berger S, Bernstein BE, Carpten JD, Clark SJ, Costello JF, Doerge RW, Esteller M, Feinberg AP, Gingeras TR, Greally JM, Henikoff S, Herman JG, Jackson-Grusby L, Jenuwein T, Jirtle RL, Kim Y-J, Laird PW, Lim B, Martienssen R, Polyak K, Stunnenberg H, Tlsty TD, Tycko B, Ushijima T, Zhu J: Moving AHEAD with an international human epigenome project. Nature. 2008, 454 (7205): 711-715. 10.1038/454711a.View ArticleGoogle Scholar
- Leeb M, Pasini D, Novatchkova M, Jaritz M, Helin K, Wutz A: Polycomb complexes act redundantly to repress genomic repeats and genes. Genes Dev. 2010, 24: 265-276. 10.1101/gad.544410.PubMed CentralView ArticlePubMedGoogle Scholar
- Hezroni H, Sailaja BS, Meshorer E: Pluripotency-related, valproic acid (VPA)-induced genome-wide histone H3 lysine 9 (H3K9) acetylation patterns in embryonic stem cells. J Biol Chem. 2011, 286: 35977-35988. 10.1074/jbc.M111.266254.PubMed CentralView ArticlePubMedGoogle Scholar
- Jiang H, Shukla A, Wang X, Chen WY, Bernstein BE, Roeder RG: Role for Dpy-30 in ES cell-fate specification by regulation of H3K4 methylation within bivalent domains. Cell. 2011, 144: 513-525. 10.1016/j.cell.2011.01.020.PubMed CentralView ArticlePubMedGoogle Scholar
- Azuara V, Perry P, Sauer S, Spivakov M, Jorgensen HF, John RM, Gouti M, Casanova M, Warnes G, Merkenschlager M, Fisher AG: Chromatin signatures of pluripotent cell lines. Nat Cell Biol. 2006, 8: 532-538. 10.1038/ncb1403.View ArticlePubMedGoogle Scholar
- Bernstein BE, Mikkelsen TS, Xie X, Kamal M, Huebert DJ, Cuff J, Fry B, Meissner A, Wernig M, Plath K, Jaenisch R, Wagschal A, Feil R, Schreiber SL, Lander ES: A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell. 2006, 125: 315-326. 10.1016/j.cell.2006.02.041.View ArticlePubMedGoogle Scholar
- Santos-Rosa H, Schneider R, Bannister AJ, Sherriff J, Bernstein BE, Emre NC, Schreiber SL, Mellor J, Kouzarides T: Active genes are tri-methylated at K4 of histone H3. Nature. 2002, 419: 407-411. 10.1038/nature01080.View ArticlePubMedGoogle Scholar
- Mikkelsen TS, Ku M, Jaffe DB, Issac B, Lieberman E, Giannoukos G, Alvarez P, Brockman W, Kim TK, Koche RP, Lee W, Meudenhall E, O’Donovan A, Presser A, Russ C, Xie X, Meissner A, Wernig M, Jaenisch R, Nusbaum C, Lander ES, Bernstein BE: Genome-wide maps of chromatin state in pluripotent and lineage-committed cells. Nature. 2007, 448: 553-560. 10.1038/nature06008.PubMed CentralView ArticlePubMedGoogle Scholar
- Kashyap V, Gudas LJ, Brenet F, Funk P, Viale A, Scandura JM: Epigenomic reorganization of the clustered Hox genes in embryonic stem cells induced by retinoic acid. J Biol Chem. 2011, 286: 3250-3260. 10.1074/jbc.M110.157545.PubMed CentralView ArticlePubMedGoogle Scholar
- Lempradl A, Ringrose L: How does noncoding transcription regulate Hox genes?. Bioessays. 2008, 30: 110-121. 10.1002/bies.20704.View ArticlePubMedGoogle Scholar
- Johnson CA, White DA, Lavender JS, O’Neill LP, Turner BM: Human class I histone deacetylase complexes show enhanced catalytic activity in the presence of ATP and co-immunoprecipitate with the ATP-dependent chaperone protein Hsp70. J Biol Chem. 2002, 277: 9590-9597. 10.1074/jbc.M107942200.View ArticlePubMedGoogle Scholar
- Jin Q, Yu LR, Wang L, Zhang Z, Kasper LH, Lee JE, Wang C, Brindle PK, Dent SY, Ge K: Distinct roles of GCN5/PCAF-mediated H3K9ac and CBP/p300-mediated H3K18/27ac in nuclear receptor transactivation. EMBO J. 2011, 30: 249-262. 10.1038/emboj.2010.318.PubMed CentralView ArticlePubMedGoogle Scholar
- Pasini D, Malatesta M, Jung HR, Walfridsson J, Willer A, Olsson L, Skotte J, Wutz A, Porse B, Jensen ON, Helin K: Characterization of an antagonistic switch between histone H3 lysine 27 methylation and acetylation in the transcriptional regulation of Polycomb group target genes. Nucleic Acids Res. 2010, 38: 4958-4969. 10.1093/nar/gkq244.PubMed CentralView ArticlePubMedGoogle Scholar
- Eskeland R, Freyer E, Leeb M, Wutz A, Bickmore WA: Histone acetylation and the maintenance of chromatin compaction by Polycomb repressive complexes. Cold Spring Harb Symp Quant Biol. 2010, 75: 71-78. 10.1101/sqb.2010.75.053.View ArticlePubMedGoogle Scholar
- Potocnik AJ, Nielsen PJ, Eichmann K: In vitro generation of lymphoid precursors from embryonic stem cells. EMBO J. 1994, 13: 5274-5283.PubMed CentralPubMedGoogle Scholar
- Bernstine EG, Koyama H, Ephrussi B: Enhanced expression of alkaline phosphatase in hybrids between neuroblastoma and embryonal carcinoma. Somatic Cell Genet. 1977, 3: 217-225. 10.1007/BF01551816.View ArticlePubMedGoogle Scholar
- Hashemi SM, Soudi S, Shabani I, Naderi M, Soleimani M: The promotion of stemness and pluripotency following feeder-free culture of embryonic stem cells on collagen-grafted 3-dimensional nanofibrous scaffold. Biomaterials. 2011, 32: 7363-7374. 10.1016/j.biomaterials.2011.06.048.View ArticlePubMedGoogle Scholar
- White DA, Belyaev ND, Turner BM: Preparation of site-specific antibodies to acetylated histones. Methods. 1999, 19: 417-424. 10.1006/meth.1999.0878.View ArticlePubMedGoogle Scholar
- O’Neill LP, Turner BM: Immunoprecipitation of native chromatin: NChIP. Methods. 2003, 31: 76-82. 10.1016/S1046-2023(03)00090-2.View ArticlePubMedGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.