Biochimica et Biophysica Acta 1839 (2014) 793–799

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbagrm

Akt1 mediates the posterior Hoxc gene expression through epigenetic modifications in mouse embryonic fibroblasts Kyoung-Ah Kong, Ji-Yeon Lee, Ji Hoon Oh, Youra Lee, Myoung Hee Kim ⁎ Department of Anatomy, Embryology Lab., Yonsei University College of Medicine, Seoul, Republic of Korea

a r t i c l e

i n f o

Article history: Received 5 August 2013 Received in revised form 9 May 2014 Accepted 11 June 2014 Available online 21 June 2014 Keywords: Hox gene expression Epigenetic modification Akt1 Hdac Ezh2 Kdm6b

a b s t r a c t The evolutionarily conserved Hox genes are organized in clusters and expressed colinearly to specify body patterning during embryonic development. Previously, Akt1 has been identified as a putative Hox gene regulator through in silico analysis. Substantial upregulation of consecutive 5′ Hoxc genes has been observed when Akt1 is absent in mouse embryonic fibroblast (MEF) cells. In this study, we provide evidence that Akt1 regulates the 5′ Hoxc gene expression by epigenetic modifications. Enrichment of histone H3K9 acetylation and a low level of the H3K27me3 mark were detected at the posterior 5′ Hoxc loci when Akt1 is absent. A histone deacetylase (HDAC) inhibitor de-repressed 5′ Hoxc gene expression when Akt1 is present, and a DNA demethylating reagent synergistically upregulated HDAC-induced 5′ Hoxc gene expression. A knockdown study revealed that Hdac6 is mediated in the Hoxc12 repression through direct binding to the transcription start site (TSS) in the presence of Akt1. Co-immunoprecipitation analysis revealed that endogenous Akt1 directly interacted with Hdac6. Furthermore, exogenous Akt1 was enriched at the promoter region of the posterior Hoxc genes such as Hoxc11 and Hoxc12, not the Akt1-independent Hoxc5 and Hoxd10 loci. The regulation of the H3K27me3 mark by Ezh2 and Kdm6b at the 5′ Hoxc gene promoter turned out to be Akt1 dependent. Taken together, these results suggest that Akt1 mediates the posterior 5′ Hoxc gene expression through epigenetic modification such as histone methylation and acetylation, and partly through a direct binding to the promoter region of the 5′ Hoxc genes and/or Hdac6 in mouse embryonic fibroblast cells. © 2014 Elsevier B.V. All rights reserved.

1. Introduction The Hox transcription factor plays critical roles in pattern formation and differentiation. In mammals, there are 39 Hox genes clustered on four different chromosomal loci. The genomic location of the Hox genes correlates to their spatio-temporal expression, termed colinearity [1,2]. During development, the combination of expressed Hox genes decide the cellular and positional identity of specific cell types along the anteroposterior body axis, by orchestrating downstream target gene expression [3]. Retinoic acid (RA) is one of the most characterized factors regulating Hox gene expression. Retinoic acid response elements (RAREs) have

Abbreviations: HDAC, histone deacetylase; ChIP, chromatin immunoprecipitation; MEF, mouse embryonic fibroblast; PRC, polycomb repressive complex; Hox, homeobox gene; Akt, v-akt murine thymoma viral oncogene homolog; PKB, protein kinase B; PI3K, phosphatidylinositol 3-kinase; RT-PCR, reverse transcription-polymerase chain reaction; DNMT, DNA methyltransferase; H3K27, histone H3 lysine 27; TSS, transcription start site; WT, wild type ⁎ Corresponding author at: Department of Anatomy, Embryology Lab., Brain Korea 21 PLUS Project for Medical Science, Yonsei University College of Medicine, C.P.O. Box 8044, Seoul 120-752, Republic of Korea. Tel.: +82 2 2228 1647; fax: +82 2 365 0700. E-mail address: [email protected] (M.H. Kim).

http://dx.doi.org/10.1016/j.bbagrm.2014.06.011 1874-9399/© 2014 Elsevier B.V. All rights reserved.

been identified in several Hox genes and aberrant RA activity has led to distinct homeotic transformations through alterations in Hox expression [4,5]. The upstream transcription factors involved in Hox gene expression include Pbx, Meis, Prep cofactors, Krox20, Sox/Oct family members, Cdx, Jpk, and the Hox proteins themselves [6–9]. Recently, diverse mechanisms particularly, epigenetic regulations such as histone modification [10–17], DNA methylation [16,18–21], and chromatin architecture [20,22–28] have been suggested to regulate Hox gene expression. The progressive changes of chromatin marks and 3-dimensional conformation changes seemed to facilitate the temporal and spatial Hox gene regulation during early embryonic development [28–31]. Long noncoding RNAs (lncRNAs) have also been implicated in the control of vertebrate Hox gene expression recruiting trithorax or polycomb group members to Hox clusters [32–35]. In addition, Hox genes are not only expressed during development, but also in diverse organs in adult and play multiple roles in coordinating cellular processes such as cell division, proliferation, and differentiation [36]. Endocrine regulation of Hox genes and the aberrant Hox gene expression have been described in numerous human cancers [37–40]. In a recent study, we identified Akt1 as an additional Hox regulator. Several unexpressed consecutive 5′ Hoxc genes in wild-type (WT) cells were derepressed in Akt1-null MEFs and the DNA methylation status showed reverse correlation with gene expression [21].

794

K.-A. Kong et al. / Biochimica et Biophysica Acta 1839 (2014) 793–799

Akt, also known as protein kinase B (PKB), is a serine–threonine protein kinase. It is a downstream effector of the phosphatidylinositol 3-kinase (PI3K) pathway and plays a key role in multiple cellular processes such as glucose metabolism, cellular proliferation, apoptosis, transcription, cell migration, and development. In mammals, there are three closely related isoforms of protein kinase B (PKB)/Akt family members: PKBα/Akt1, PKBβ/Akt2, and PKBγ/Akt3. Among the three Akt isoforms, Akt1 is the predominantly expressed isoform in most tissues [41]. Recently, Akt was shown to modulate gene expression through epigenetic modifications. Akt1 enhances DNA methylation through the expression of DNA methyltransferase (DNMTs) and DNMT stability [42–44]. In several cancers, Akt1 has been reported to mediate epigenetic silencing by promoter methylation [45–47]. In addition to DNA methylation, Akt also has been shown to be involved in histone modifications; the activity of histone methyltransferase EZH2 and histone acetyltransferase CBP were shown to be regulated by Akt through phosphorylation [48,49]. To address the molecular mechanism underlying Hox gene regulation by Akt1 [21], we investigated epigenetic modifications such as histone modifications along the Hox loci. The loss of Akt1 in MEFs resulted in induced histone H3K9 acetylation and reduced histone H3K27me3 interaction on the 5′ Hoxc genes Hoxc11 and Hoxc12. In Akt1-null MEFs, the expression of several Hdac genes was reduced and an RNAi experiment suggested that Hdac6 regulates acetylation status at Hoxc12. Expression analyses and ChIP assays have shown that Ezh2 and Kdm6b regulate histone H3K27me3 levels at the Hoxc11 and Hoxc12 genes, which were dependent on the presence of Akt1. Our data demonstrate that Akt1 is important for the regulation of Hox gene expression through epigenetic modifications, probably through the direct binding to the promoter region of the 5′ Hoxc genes and/or HDAC6 in mouse embryonic fibroblasts. 2. Materials and methods 2.1. Cell culture, plasmid and transfection Akt1+/+ (WT) and Akt1−/− MEFs (E13.5) were generated as previously described [50]. MEFs were maintained in Dulbecco's modified Eagle's medium (DMEM: WelGENE Inc., Daegu, Korea) supplemented with 10% fetal bovine serum (WelGENE Inc.), 100 units/ml penicillin, and 100 μg/ml streptomycin sulfate (WelGENE Inc.) at 37 °C in 5% CO2. Cells were transfected with siRNA against mouse Akt1 (s62215; Ambion Inc., Austin, TX, USA), HDAC6 (siRNA ID# s67425; Ambion), HDAC9 (siRNA ID# s95881; Ambion), or a universal negative control siRNA (ST Pharm. Co., Ltd. Seoul, Korea) using HiPerFect reagent (Qiagen, Hilden, Germany). For the long-term transfection, cells were transfected with 2 nM siRNA according to the manufacturer's protocol. The knockdown efficiency of siRNA was confirmed with RT-PCR and western blot analysis. A plasmid pcDNA-HA-Akt1 was constructed as follows: full-length mouse Akt1 gene was amplified from mouse P0 cDNA using the forward primer 5′-ATGCgaattcACCATGAACGACG TAGCC-3′ and reverse primer 5′-ATGCgtcgacCTTCT GGACTCGGCAATG3′, possessing polylinker EcoRI and XhoI restriction enzyme sites, respectively. For HA tagging, the PCR product was cloned into the EcoRI and XhoI sites of a pcDNA3 vector (Invitrogen, Carlsbad, CA, USA) containing an HA epitope (MASYPYDVPDYASPE). The plasmids were introduced into cells by transfection using a Lipofectamine 2000 Transfection Reagent (Invitrogen, Carlsbad, CA, USA), following the manufacturer's instructions. 2.2. Antibodies The following antibodies were used for Western blot and ChIP analysis: anti-Akt1 (#2938, Cell Signaling Technology, Danvers, MA, USA), anti-Akt2 (#2964, Cell Signaling Technology), anti-Akt3 (#3788, Cell Signaling Technology), anti-histone H3 (ab1791, Abcam, Cambridge,

UK), anti-histone H3 acetyl K9 (ab12179, Abcam), anti-histone H3 trimethyl K27 (ab6002, Abcam), anti-Ezh2 (#5246, Cell Signaling Technology), anti-Suz12 (#3737, Cell Signaling Technology), anti-KDM6A (ab36938, Abcam), anti-KDM6B (ab85392, Abcam), anti-HA tag antibody (ab9110, Abcam), anti-Hdac6 antibody (ab12179, Abcam), anti-β-actin antibody (ab6276, Abcam), and normal mouse and rabbit IgG (sc-2025 and sc-2027, Santa Cruz Biotechnology, Inc., CA, USA). 2.3. Immunoprecipitation (IP) Cells were harvested and resuspended in Nonidet-P40 (NP-40) lysis buffer (150 mM NaCl, 1.0% NP-40, 50 mM Tris–HCl [pH 8.0]) containing a protease inhibitor. After preclearing, the cell lysates were incubated with normal IgG (as a negative control) or primary antibodies overnight at 4 °C. The lysates were incubated with Protein A/G agarose beads and then washed with NP-40 lysis buffer. The precipitated proteins were released by boiling in loading buffer and resolved by SDS-PAGE. 2.4. RT- and real-time PCR analysis Total RNA was isolated using the TRIzol reagent (Invitrogen, Carlsbad, CA, USA), according to the manufacturer's protocol. Two micrograms of total RNA was reverse-transcribed using the ImProm-II™ Reverse Transcription System (Promega, Madison, WI, USA) and cDNA was amplified with the specific primers listed in Table S1. PCR amplification conditions and primers for detecting Hoxc genes were the same as described previously [21]. For quantitative PCR analysis, the StepOnePlus™ Real-Time PCR System (Applied Biosystems, Foster, CA, USA) and the Power SYBR Green PCR Master Mix (Applied Biosystems) kits were used. 2.5. Western blot After cell lysis, protein content was determined using a Pierce BCA Protein Assay Kit (Thermo Scientific, Rockford, IL, USA). Ten micrograms of protein was loaded on 10% SDS poly-acrylamide gels, electrotransferred to PVDF membranes (Bio-Rad, Hercules, CA, USA), and probed with appropriate antibodies. Immunoreactive bands were detected with anti-rabbit (Abcam) or anti-mouse (Zymed, South San Francisco, CA, USA) HRP secondary antibody and visualized using an enhanced chemiluminescent substrate (SuperSignal West Pico Chemiluminescent Substrate kit; Thermo Scientific). 2.6. Chromatin immunoprecipitation (ChIP) and ChIP-PCR assay ChIP analysis was performed following the protocol previously described by Min et al. [31] with minor modifications. Briefly, cells were cross-linked with 1% formaldehyde for 15 min at room temperature and the reaction was stopped by adding 125 mM glycine. After washing twice with cold PBS, cells were lysed for 10 min on ice in lysis buffer (1% SDS, 1% Triton X-100, 0.1% sodium deoxycholate, 10 mM EDTA, 50 mM Tris–HCl [pH 8.0]) containing a protease inhibitor. The chromatin was sonicated with 30 sets of 10-s pulses on ice to generate DNA fragments with an average length of 500–1000 bp. After preclearing, 1% of each sample was saved as the input fraction. Immunoprecipitation was performed using specific antibodies against the indicated proteins or IgG of a different species as the control. After cross-linking reversal, nucleic acids were prepared from the eluted complex and quantitative PCR analysis was performed using StepOnePlus™ Real-Time PCR System and Power SYBR Green PCR Master Mix. Primers for ChIP-qPCR are listed in Tables S2 and S3. 2.7. Statistical analysis All quantitative data were from experiments repeated at least three times, and significance levels were calculated using one-way ANOVA or

K.-A. Kong et al. / Biochimica et Biophysica Acta 1839 (2014) 793–799

Student's t test. A p-value less than 0.05 was considered statistically significant. 3. Results 3.1. Akt1 represses 5′ Hoxc gene expression in mouse embryonic fibroblast (MEF) cells In search of regulatory genes responsible for the colinear expression of Hox genes during embryogenesis, we previously identified Akt1 as a regulator of Hox genes through in silico analysis [21]. Several consecutive 5′ Hoxc genes that were repressed in WT cells were de-repressed in Akt1-null MEFs. In addition, the DNA methylation status showed reverse correlation with gene expression. We first confirmed that downregulation of the 5′ Hoxc gene expression was due to Akt1 itself and not a result of differences in WT vs. Akt1-null MEF status. WT MEFs were transfected with siRNA against Akt1 (siAkt1) for 21 days and the siRNA knockdown efficiency was confirmed by RT-PCR and Western blot analysis (Fig. 1A). As shown in Fig. 1, Akt1 siRNA specifically downregulated Akt1 production, without affecting the production of other Akt isoforms (Akt2 and Akt3). Similar to our previous findings in Akt1−/− MEFs [21], treatment with siAkt1 in WT MEFs significantly de-repressed 5′ Hoxc gene expression (Fig. 1B), confirming that Akt1 represses 5′ Hoxc gene expression in MEFs. 3.2. Histone modifications at the 5′ Hoxc gene loci To address the molecular mechanism underlying Hox gene regulation by Akt1, epigenetic modifications such as DNA methylation and histone modifications were examined. Previously, profound DNA methylation differences at 5′ Hoxc gene promoter regions such as Hoxc11 (Fig. S1) and Hoxc12 [21] were detected through combined bisulfite restriction analysis (COBRA) and bisulfite sequencing. Results from these experiments indicated that Akt1 modulated DNA methylation patterns at the 5′ Hoxc loci. Here, we examined whether histones were differentially modified in the presence or absence of Akt1. Hoxc11 and Hoxc12 genes, which are

795

significantly upregulated by siAkt1 (Fig. 1B) and the control Hoxc5 and Hoxd10 genes, whose expression levels were not affected by Akt1, were chosen for analysis. As shown in Fig. 2, histone H3 lysine 9 acetylation (H3K9ac), a marker for transcription activation, was enhanced in Akt1−/− MEFs at the Hoxc11 and Hoxc12 loci. However, the acetylation levels at Hoxc5 were strong in both Akt1+/+ and Akt1−/− MEFs, as Hoxc5 was expressed strongly in both cell types. In the Hoxd10 negative control, whose expression was not detected in both MEF types [21], the acetylation mark was not detected in either MEF type. Contrary to histone acetylation, histone H3 lysine 27 trimethylation (H3K27me3), a marker for transcriptional repression, was reduced in Akt1-null MEFs at Hoxc11 and Hoxc12 sites and levels were unchanged (or slightly reduced) at Hoxc5 and Hoxd10 sites compared with WT MEFs. 3.3. Histone deacetylase mediates the 5′ Hoxc gene repression when Akt1 is present To examine whether Akt1-mediated repression of the 5′ Hoxc gene is due to the histone deacetylase (HDAC), WT MEFs were treated with trichostatin A (TSA), an inhibitor of class I/II/IV HDACs. Expression of the 5′ Hoxc gene was upregulated in a dose-dependent manner (Fig. 3A). Furthermore, the treatment of 5-aza-2′ deoxycytidine (Aza), a specific inhibitor of DNA methylation, in combination with TSA synergistically de-repressed the 5′ Hoxc genes. A control 3′ Hoxc gene, Hoxc5, showed no change in its expression. There was no effect on the derepression of the 5′ Hoxc genes when MEFs were treated with the class III HDAC inhibitor, salermide (Fig. S2). To determine which HDAC is involved in the regulation of Hox gene expression, the relative expression level of class I, II and IV HDACs were examined in WT and Akt1-null MEFs. As shown in Fig. 3B, the expression levels of Hdac3, Hdac5, Hdac6, Hdac9 and Hdac11 were reduced in Akt1null MEFs, and the protein levels of Hdac6 and Hdac9, whose expressions were significantly reduced in the absence of Akt1, were confirmed through Western blotting (Fig. 3C). To see whether specific Hdacs repressed the 5′ Hoxc genes, a knockdown experiment with siRNA against Hdac6 and Hdac9 was performed. As shown in Fig. 3D, Hoxc12 expression was recovered when Hdac6 siRNA was treated in

Fig. 1. Effects of Akt1 on Hoxc gene expression. WT MEF cells were transfected with 2 nM control, or Akt1-siRNA for 3 to 21 days. (A) Effects of Akt1 siRNA on Akt isoforms' mRNA expression (left) and protein expression (right). (B) Quantitative analysis of 5′ Hoxc gene transcripts in Akt1 siRNA-transfected MEFs. Relative values ± S.E. were obtained from three independent assays. *p b 0.05; **p b 0.01, by one-way ANOVA.

796

K.-A. Kong et al. / Biochimica et Biophysica Acta 1839 (2014) 793–799

Fig. 2. Histone modifications at the 5′ Hoxc sites in MEFs. (A) Map of the Hoxc12, Hoxc11, Hoxc5, and Hoxd10 genes. Green boxes show CpG islands and the ChIP amplicons are designated as black bars. (B) Quantitative ChIP-PCR with chromatins extracted from Akt1+/+ and Akt1−/− MEFs using the indicated antibodies. Relative values ± S.E. were obtained from three independent assays. *p b 0.05; **p b 0.01, by Student's t-test.

the presence of Aza. Hdac6 specifically repressed Hoxc12, but not Hoxc11 or Hoxc13, in MEF cells, which is most likely explained due to the differences in the genomic context of these genes. Since both siRNA and Aza treatment downregulated Hdac9 production, the

association of Hdac6 along the Hoxc12 loci was investigated. When the direct binding of Hdac6 was examined using the ChIP-PCR method, the Hoxc12 promoter region was enriched with Hdac6 near the transcription start site (TSS) in WT MEFs (Fig. 3E).

Fig. 3. Effect of histone deacetylases on the 5′ Hoxc gene expression. (A) De-repression of Hoxc11 and Hoxc12 genes in the presence of HDAC inhibitor (TSA) in MEF cells. Akt1+/+ MEFs were treated with TSA at indicated concentrations for 24 h (left) and co-treated with 5-Aza-dC (5 μM for 72 h) and/or TSA (1 μM for 24 h) (right). Expression of 5′ Hoxc genes were determined by RT-PCR. (B) Transcript levels of Class I, IIa and IIb HDACs in Akt1+/+ and Akt1−/− MEFs. Relative values ± S.E. were obtained from three independent assays. *p b 0.05; **p b 0.01, by Student's t-test. (C) Protein levels of HDAC6 and HDAC9 in Akt1+/+ and Akt1−/− MEFs. (D) Effects of HDAC6 and HDAC9 on 5′ Hoxc gene silencing. WT MEF cells were transfected with indicated siRNA (10 nM) for 72 h and treated with 1 μM 5-Aza-dC. Gene expression was determined by RT-PCR. (E) Enrichment level of HDAC6 across the Hoxc12 gene in WT and Akt1-null MEFs. Data represent the average of three independent ChIP-PCR analyses. Multi Gauge V3.0 software (Fuji, Tokyo, Japan) was used for quantification. Relative values ± S.E. were obtained from three independent assays. *p b 0.05, by Student's t-test. (F) Co-immunoprecipitation assay of Akt1 and Hdac6. Cell lysates from Akt1+/+ MEFs were immunoprecipitated (IP) using Akt1 or IgG antibody and then the purified proteins were immunoblotted (IB) using antibody against Hdac6 or Akt1 (left). The endogenous Hdac6 in Akt1+/+ and Akt1-null MEFs was immunoprecipitated using Hdac6 or IgG (right). The purified proteins were immunoblotted (IB) using by Hdac6 and Akt1 antibodies. Black and white arrows indicate immunoprecipitated Akt1 and IgG, respectively. (G) Western blot analysis showed the overexpression of HA-tagged Akt1. The small size of HA proteins in pcDNA-HA transfected samples was not detected. Beta-actin was used as a loading control. (H). Quantitative ChIP-PCR using antibody against HA in Akt1+/+ MEF cells transfected with HA or HA-tagged Akt1. Relative values ± S.E. were obtained from four independent assays. *p b 0.05; **p b 0.01, by Student's t-test.

K.-A. Kong et al. / Biochimica et Biophysica Acta 1839 (2014) 793–799

To examine whether Hdac6 interacts with Akt1, a coimmunoprecipitation (Co-IP) analysis was performed with an antibody against Akt1 as well as Hdac6. As shown in Fig. 3F, direct association between the endogenous Hdac6 and Akt1 was detected. Although Akt1 is well known as a serine–threonine protein kinase, there is a still question whether Akt1 is recruited to the Hox promoters. To overcome the absence of ChIP-grade Akt1 antibody, a plasmid coding HA-tagged Akt1 was constructed and used for ChIP analysis using antibody against HA tag. Surprisingly, increased binding of overexpressed Akt1 was observed at both the Hoxc11 and Hoxc12 loci in wild type MEFs, while Akt1 was not enriched at Hoxc5 and Hoxd10 loci whose expression and repression were Akt1-independent, respectively (Fig. 3G and H). 3.4. Regulation of the H3K27me3 mark by Ezh2 and Kdm6b are Akt1-dependent

797

histone demethylase was observed at both the Hoxc11 and Hoxc12 loci in Akt1-null MEFs (Fig. 4B). Kdm6a binding levels were not changed along the Hoxc11 and Hoxc12 loci, which are similar to the Suz12 results in the presence or absence of Akt1 (Fig. S3). To confirm that these results are due to the absence of Akt1 (i.e. Akt1 specific), not due to the differences between WT and Akt1-null MEF cells, WT MEFs were treated with Akt1 siRNA, and both mRNA and protein levels were analyzed. As shown in Fig. 4C, both histone methyltransferases (Ezh2 and Suz12) and demethylases (Kdm6a and Kdm6b) were detected at similar levels compared to the Akt1-null MEFs (Fig. 4A). Enrichment of Ezh2 in the WT and Kdm6b in the Akt1 siRNA-treated MEFs were detected at the 5′ Hoxc gene loci (Fig. 4D), which resulted in a significant reduction of the H3K27me3 mark in siAkt1-transfected cells (Fig. 4E) as detected in Akt1-null MEF cells (Fig. 2B). 4. Discussion

Since H3K27me3 repressive marks are positioned by the histone methyltransferase contained in PRC2 and removed by the histone demethylases Kdm6a (Utx) and Kdm6b (Jmjd3), the levels of PRC2 members (Ezh2 and Suz12), Kdm6a, and Kdm6b were analyzed using RT-PCR and Western blot analysis. As shown in Fig. 4A, the relative expression levels of Ezh2 and Suz12 were not changed in the presence or absence of Akt1; however, the expression levels of Kdm6a and Kdm6b were increased in Akt1−/− MEFs. ChIP-PCR was used to test whether PRC2 proteins and histone demethylases were involved in H3K27me3 mark levels at the Hoxc11 and Hoxc12 gene loci. In WT MEFs, there was no significant change in the levels of PRC2; however, the binding of Ezh2 was increased at the Hoxc11 and Hoxc12 loci (Fig. 4B) while Suz12 remained unchanged (Fig. S3). Increased binding of the Kdm6b

We previously reported that Akt1 is a potent repressor of the 5′ Hoxc genes in mouse embryonic fibroblast cells [21]. siRNA techniques confirmed that a loss of Akt1 caused the upregulation of the 5′ Hoxc genes such as Hoxc11 and Hoxc12 (Fig. 1). These results strongly suggested that upregulation of the 5′ Hoxc genes in Akt1-null MEFs was not a result of the different status of MEF cells (i.e., WT vs. Akt1−/− MEFs), but was due to direct regulation by Akt1. In this study, we investigated how Hox gene expression was regulated by Akt1. In Akt1-null MEF cells, histone acetylation at H3K9, an active mark, was enriched at the 5′ Hoxc11 and Hoxc12 gene loci, which were coupled with their expression (Fig. 2). Since HDACs have been reported to be associated with many transcriptional repressive

Fig. 4. H3K27me3 modification at the 5′ Hoxc loci through Ezh2 and Kdm6b in MEFs. Expression level of polycomb proteins (Ezh2 and Suz12) and H3K27 demethylases (Kdm6a and Kdm6b) at the mRNA (left) and protein (right) levels in WT vs. Akt1-null MEF cells (A), and control siRNA vs. siAkt1-treated cells (C). Quantitative ChIP-PCR at the Hoxc11 and Hoxc12 loci using anti-Ezh2 and anti-Kdm6b antibody in WT and Akt1-null MEFs (B), and Akt1-siRNA transfected MEFs (D). (E) Quantitative analysis of ChIP-PCR at the 5′ Hoxc genes, Hoxc11 and Hoxc12 in Akt1-siRNA transfected MEFs. Relative values ± S.E. were obtained from three independent assays. *p b 0.05, by Student's t-test.

798

K.-A. Kong et al. / Biochimica et Biophysica Acta 1839 (2014) 793–799

complexes and participate in transcriptional repression, we questioned whether HDACs were involved in 5′ Hoxc gene repression in Akt1+/+ WT MEFs. Hdac6 was shown to participate in the transcriptional repression of Hoxc12 (Fig. 3). In the case of Hoxc11 and Hoxc13, different Hdac(s) or locus/cellular context-dependency might be involved in their repression instead of Hdac6 alone. In previous reports, HDAC4 was shown to repress the expression of Hoxb13 in androgen receptornegative prostate cancer cells [51], in which HDAC4 was recruited by transcription factor YY. Although the endogenous Akt1 seemed to interact with Hdac6 directly in MEFs and the exogenous Akt1 was enriched at the 5′ Hoxc genes (Hoxc11 and Hoxc12 loci), it is not clear how Akt1 is involved in Hdac recruitment and function. Because Hdac6 contains a potential Akt phosphorylation site, serine 129 (Fig. S4A), we wondered whether it could be modulated by Akt1. To determine whether Akt1 can phosphorylate Hdac6, we immunoprecipitated Hdac6 and probed it with an antibody that recognizes the phosphorylated form of Akt substrates. Surprisingly, the level of Hdac6 phosphorylation was enriched when Akt1 was absent (Fig. S4B). Since Akt2 and Akt3 isoforms are still in the cell, Hdac6 phosphorylation might be due to these isoforms. From these results along with the previous one (Fig. 3B), Akt1 seems to regulate the expression of Hdac6 (Fig. 3B) to repress 5′ Hoxc genes, rather than the direct phosphorylation of Hdac6. In Akt1-null MEFs, histone acetyltransferases (HATs) might be involved in increased acetylation levels at the 5′ Hoxc genes. Since histone acetylases such as Moz and Gcn5 have been reported to regulate Hox gene expression during development [52,53], we may need to discover which HAT is implicated in Hox gene expression in Akt1 signaling in the future. In histone H3K27 methylation, the activation of PI3K/AKT signaling has been reported to induce epigenetic silencing in which the enrichment of H3K27me3 in the promoter region repressed the CST6 tumor suppressor gene in breast cancer [46]. In another report, tumor suppressor C-terminal Src kinase (Csk)-binding protein (CBP) was downregulated by the enhanced enrichment of H3K27me3 by the MAPK/PI3K pathway [54]. In a mammary epithelial cell line, the forced expression of Akt1 induced genome-wide reprogramming of H3K27me3 in many downstream genes, and several Hox genes were included in these targets [55]. Here, as one piece of the evidence that Akt1 regulates Hox gene expression through histone H3K27 trimethylation, we showed a predominant enrichment of H3K27me3 at the 5′ Hoxc11 and Hoxc12 loci in the presence of Akt1 in MEF cells (Fig. 2). Although the level of Ezh2 histone methyltransferase expression was similar in both WT and Akt1-null MEFs, the recruitment of Ezh2 at the 5′ Hoxc loci was reduced when Akt1 was absent (Fig. 4B). In contrast to our result, the presence of an Akt signal has been shown to de-repress silent target genes by phosphorylating EZH2 in cancer cell lines, impeding EZH2 binding to histone H3 and resulting in decreased trimethylation at H3K27 [48]. Although this discrepancy could be due to the cell type used, how Akt1 regulates Ezh2 binding or activity on the target sites in MEFs remains to be revealed. Along with Ezh2, the histone demethylases, Kdm6a and Kdm6b, are also important regulators of gene expression through H3K27me3 and have been reported to regulate Hox genes during development and differentiation [17,56,57]. Interestingly, Lan et al. [57] showed that inhibition of Utx (Kdm6a) led to the mis-regulation of Hox genes and resulted in a concomitant posterior developmental defect in a Zebrafish model. Recently, Zhang et al. [58] showed that active Akt1 downregulated Kdm6a and Kdm6b expression, which is consistent with our finding here of both Kdm6a and Kdm6b gene upregulation in both Akt1-null and siAkt1-treated MEF cells (Fig. 4A,C). Furthermore, induced binding of Kdm6b was detected in the posterior 5′ Hoxc genes in both Akt1-null and siAkt1-treated MEFs (Fig. 4B,D). In this study, we suggest that Akt1 mediates the posterior 5′ Hoxc gene expression through epigenetic modification such as histone methylation and acetylation, and partly through a direct binding to the promoter region of the 5′ Hoxc genes and/orHdac6 in mouse embryonic fibroblast cells.

Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.bbagrm.2014.06.011. Acknowledgements This research was supported by the Basic Science Research Program through the National Research Foundation (NRF) funded by the Ministry of Education, Science and Technology (2010-0026759 and NRF2013R1A1A2008399), the Next-Generation BioGreen21 Program of the Rural Development Administration (PJ00905601), and partly from a faculty research grant (6-2013-0049) of Yonsei University College of Medicine, Korea. References [1] J.C. Pearson, D. Lemons, W. McGinnis, Modulating Hox gene functions during animal body patterning, Nat. Rev. Genet. 6 (2005) 893–904. [2] D. Lemons, W. McGinnis, Genomic evolution of Hox gene clusters, Science 313 (2006) 1918–1922. [3] D.M. Wellik, Hox genes and vertebrate axial pattern, Curr. Top. Dev. Biol. 88 (2009) 257–278. [4] J. Deschamps, J. van Nes, Developmental regulation of the Hox genes during axial morphogenesis in the mouse, Development 132 (2005) 2931–2942. [5] T. Alexander, C. Nolte, R. Krumlauf, Hox genes and segmentation of the hindbrain and axial skeleton, Annu. Rev. Cell Dev. Biol. 25 (2009) 431–456. [6] S. Tabariès, J. Lapointe, T. Besch, M. Carter, J. Woollard, C.K. Tuggle, L. Jeannotte, Cdx protein interaction with Hoxa5 regulatory sequences contributes to Hoxa5 regional expression along the axial skeleton, Mol. Cell. Biol. 25 (2005) 1389–1401. [7] D. Bansal, C. Scholl, S. Frohling, E. McDowell, B.H. Lee, K. Dohner, P. Ernst, A.J. Davidson, G.Q. Daley, L.I. Zon, D.G. Gilliland, B.J. Huntly, Cdx4 dysregulates Hox gene expression and generates acute myeloid leukemia alone and in cooperation with Meis1a in a murine model, Proc. Natl. Acad. Sci. U. S. A. 103 (2006) 16924–16929. [8] L. Kobrossy, M. Rastegar, M. Featherstone, Interplay between chromatin and transacting factors regulating the Hoxd4 promoter during neural differentiation, J. Biol. Chem. 281 (2006) 25926–25939. [9] E.Y. Lee, M.H. Kim, Jpk, a novel cell death inducer, regulates the expression of Hoxa7 in F9 teratocarcinoma cells, but not during apoptosis, Ann. N. Y. Acad. Sci. 1090 (2006) 182–187. [10] S. Lee, J.W. Lee, S.K. Lee, UTX, a histone H3-lysine 27 demethylase, acts as a critical switch to activate the cardiac developmental program, Dev. Cell 22 (2012) 25–37. [11] B. Biehs, J.K. Hu, N.B. Strauli, E. Sangiorgi, H. Jung, R.P. Heber, S. Ho, A.F. Goodwin, J.S. Dasen, M.R. Capecchi, O.D. Klein, BMI1 represses Ink4a/Arf and Hox genes to regulate stem cells in the rodent incisor, Nat. Cell Biol. 15 (2013) 846–852. [12] P. Schorderet, N. Lonfat, F. Darbellay, P. Tschopp, S. Gitto, N. Soshnikova, D. Duboule, A genetic approach to the recruitment of PRC2 at the HoxD locus, PLoS Genet. 9 (2013) e1003951. [13] M. Kanduri, B. Sander, S. Ntoufa, N. Papakonstantinou, L.A. Sutton, K. Stamatopoulos, C. Kanduri, R. Rosenquist, A key role for EZH2 in epigenetic silencing of HOX genes in mantle cell lymphoma, Epigenetics 8 (2013) 1280–1288. [14] Ö. Copur, J. Müller, The histone H3-K27 demethylase Utx regulates HOX gene expression in Drosophila in a temporally restricted manner, Development 140 (2013) 3478–3485. [15] M. Shahhoseini, Z. Taghizadeh, M. Hatami, H. Baharvand, Retinoic acid dependent histone 3 demethylation of the clustered HOX genes during neural differentiation of human embryonic stem cells, Biochem. Cell Biol. 91 (2013) 116–122. [16] H. He, X. Hua, J. Yan, Epigenetic regulations in hematopoietic Hox code, Oncogene 30 (2011) 379–388. [17] K. Agger, P.A. Cloos, J. Christensen, D. Pasini, S. Rose, J. Rappsilber, I. Issaeva, E. Canaani, A.E. Salcini, K. Helin, UTX and JMJD3 are histone H3K27 demethylases involved in HOX gene regulation and development, Nature 449 (2007) 731–734. [18] S. Branciamore, Z.X. Chen, A.D. Riggs, S.N. Rodin, CpG island clusters and proepigenetic selection for CpGs in protein-coding exons of HOX and other transcription factors, Proc. Natl. Acad. Sci. U. S. A. 107 (2010) 15485–15490. [19] X. Wu, Y. Gong, J. Yue, B. Qiang, J. Yuan, X. Peng, Cooperation between EZH2, NSPc1mediated histone H2A ubiquitination and Dnmt1 in HOX gene silencing, Nucleic Acids Res. 36 (2008) 3590–3599. [20] V. Kashyap, L.J. Gudas, F. Brenet, P. Funk, A. Viale, J.M. Scandura, Epigenomic reorganization of the clustered Hox genes in embryonic stem cells induced by retinoic acid, J. Biol. Chem. 286 (2011) 3250–3260. [21] K.A. Kong, H. Yoon, M.H. Kim, Akt1 as a putative regulator of Hox genes, Gene 513 (2013) 287–291. [22] D. Noordermeer, D. Duboule, Chromatin architectures and Hox gene collinearity, Curr. Top. Dev. Biol. 104 (2013) 113–148. [23] J. Fraser, M. Rousseau, S. Shenker, M.A. Ferraiuolo, Y. Hayashizaki, M. Blanchette, J. Dostie, Chromatin conformation signatures of cellular differentiation, Genome Biol. 10 (2009) R37. [24] M.A. Ferraiuolo, M. Rousseau, C. Miyamoto, S. Shenker, X.Q. Wang, M. Nadler, M. Blanchette, J. Dostie, The three-dimensional architecture of Hox cluster silencing, Nucleic Acids Res. 38 (2010) 7472–7484. [25] M. Rousseau, J.L. Crutchley, H. Miura, M. Suderman, M. Blanchette, J. Dostie, Hox in motion: tracking HoxA cluster conformation during differentiation, Nucleic Acids Res. 42 (2014) 1524–1540, http://dx.doi.org/10.1093/nar/gkt998.

K.-A. Kong et al. / Biochimica et Biophysica Acta 1839 (2014) 793–799 [26] Y.J. Kim, K.R. Cecchini, T.H. Kim, Conserved, developmentally regulated mechanism couples chromosomal looping and heterochromatin barrier activity at the homeobox gene A locus, Proc. Natl. Acad. Sci. U. S. A. 108 (2011) 7391–7396. [27] A.V. Pindyurin, B. van Steensel, Hox in space: gene cluster regulation linked to folding of chromatin, Nucleus 3 (2012) 118–122. [28] D. Noordermeer, M. Leleu, E. Splinter, J. Rougemont, W. De Laat, D. Duboule, The dynamic architecture of Hox gene clusters, Science 334 (2011) 222–225. [29] N. Soshnikova, D. Duboule, Epigenetic temporal control of mouse Hox genes in vivo, Science 324 (2009) 1320–1333. [30] M. Rastegar, L. Kobrossy, E.N. Kovacs, I. Rambaldi, M. Featherstone, Sequential histone modifications at Hoxd4 regulatory regions distinguish anterior from posterior embryonic compartments, Mol. Cell. Biol. 24 (2004) 8090–8103. [31] H. Min, J.Y. Lee, M.H. Kim, Structural dynamics and epigenetic modifications of Hoxc loci along the anteroposterior body axis in developing mouse embryos, Int. J. Biol. Sci. 8 (2012) 802–810. [32] J.S. Dasen, Long noncoding RNAs in development: solidifying the Lncs to Hox gene regulation, Cell Rep. 5 (2013) 1–2. [33] K.C. Wang, Y.W. Yang, B. Liu, A. Sanyal, R. Corces-Zimmerman, Y. Chen, B.R. Lajoie, A. Protacio, R.A. Flynn, R.A. Gupta, J. Wysocka, M. Lei, J. Dekker, J.A. Helms, H.A. Chang, A long noncoding RNA maintains active chromatin to coordinate homeotic gene expression, Nature 472 (2011) 120–124. [34] L. Li, B. Liu, O.L. Wapinski, M.C. Tsai, K. Qu, J. Zhang, J.C. Carlson, M. Lin, F. Fang, R.A. Gupta, J.A. Helms, H.Y. Chang, Targeted disruption of hotair leads to homeotic transformation and gene derepression, Cell Rep. 5 (2013) 3–12. [35] J.T. Lee, Epigenetic regulation by long noncoding RNAs, Science 338 (2012) 1435–1439. [36] R. Morgan, Hox genes: a continuation of embryonic patterning? Trends Genet. 22 (2006) 67–69. [37] N. Shah, S. Sukumar, The Hox genes and their roles in oncogenesis, Nat. Rev. Cancer 10 (2010) 361–371. [38] G.S. Daftary, H.S. Taylor, Endocrine regulation of HOX genes, Endocr. Rev. 27 (2006) 331–355. [39] H. Chen, S. Sukumar, HOX genes: emerging stars in cancer, Cancer Biol. Ther. 2 (2003) 524–525. [40] K.I. Ansari, I. Hussain, S. Kasiri, S.S. Mandal, HOXC10 is overexpressed in breast cancer and transcriptionally regulated by estrogen via involvement of histone methylases MLL3 and MLL4, J. Mol. Endocrinol. 48 (2012) 61–75. [41] M. Hanada, J. Feng, B.A. Hemmings, Structure, regulation and function of PKB/AKT—a major therapeutic target, Biochim. Biophys. Acta 1697 (2004) 3–16. [42] L. Sun, H. Zhao, Z. Xu, Q. Liu, Y. Liang, L. Wang, X. Cai, L. Zhang, L. Hu, G. Wang, X. Zha, Phosphatidylinositol 3-kinase/protein kinase B pathway stabilizes DNA methyltransferase I protein and maintains DNA methylation, Cell. Signal. 19 (2007) 2255–2263. [43] P.O. Estève, Y. Chang, M. Samaranayake, A.K. Upadhyay, J.R. Horton, G.R. Feehery, X. Cheng, S. Pradhan, A methylation and phosphorylation switch between an adjacent lysine and serine determines human DNMT1 stability, Nat. Struct. Mol. Biol. 18 (2011) 42–48. [44] C. Mei, L. Sun, Y. Liu, Y. Yang, X. Cai, M. Liu, W. Yao, C. Wang, X. Li, L. Wang, Z. Li, Y. Shi, S. Qiu, J. Fan, X. Zha, Transcriptional and post-transcriptional control of

[45]

[46]

[47]

[48]

[49] [50]

[51]

[52]

[53]

[54]

[55]

[56]

[57]

[58]

799

DNA methyltransferase 3B is regulated by phosphatidylinositol 3 kinase/Akt pathway in human hepatocellular carcinoma cell lines, J. Cell. Biochem. 111 (2010) 158–167. S.O. Lim, J.M. Gu, M.S. Kim, H.S. Kim, Y.N. Park, C.K. Park, J.W. Cho, Y.M. Park, G. Jung, Epigenetic changes induced by reactive oxygen species in hepatocellular carcinoma: methylation of the E-cadherin promoter, Gastroenterology 135 (2008) 2128–2140. H.J. Lin, T. Zuo, C.H. Lin, C.T. Kuo, S. Liyanarachchi, S. Sun, R. Shen, D.E. Deatherage, D. Potter, L. Asamoto, S. Lin, P.S. Yan, A.L. Cheng, M.C. Ostrowski, T.H. Huang, Breast cancer-associated fibroblasts confer AKT1-mediated epigenetic silencing of Cystatin M in epithelial cells, Cancer Res. 68 (2008) 10257–10266. R.K. Lin, Y.S. Hsieh, P. Lin, H.S. Hsu, C.Y. Chen, Y.A. Tang, C.F. Lee, Y.C. Wang, The tobacco-specific carcinogen NNK induces DNA methyltransferase 1 accumulation and tumor suppressor gene hypermethylation in mice and lung cancer patients, J. Clin. Invest. 120 (2010) 521–532. T.L. Cha, B.P. Zhou, W. Xia, Y. Wu, C.C. Yang, C.T. Chen, B. Ping, A.P. Otte, M.C. Hung, Akt-mediated phosphorylation of EZH2 suppresses methylation of lysine 27 in histone H3, Science 310 (2005) 306–310. Y. Liu, Z.B. Xing, Y. Fang, Akt kinase targets the association of CBP with histone H3 to regulate the acetylation of lysine K18, FEBS Lett. 587 (2013) 847–853. S.S. Bae, H. Cho, J. Mu, M.J. Birnbaum, Isoform-specific regulation of insulindependent glucose uptake by Akt/protein kinase B, J. Biol. Chem. 278 (2003) 49530–49536. G. Ren, G. Zhang, Z. Dong, Z. Liu, L. Li, Y. Feng, D. Su, Y. Zhang, B. Huang, J. Lu, Recruitment of HDAC4 by transcription factor YY1 represses HOXB13 to affect cell growth in AR-negative prostate cancers, Int. J. Biochem. Cell Biol. 41 (2009) 1094–1101. W. Lin, Z. Zhang, C.H. Chen, R.R. Behringer, S.Y. Dent, Proper Gcn5 histone acetyltransferase expression is required for normal anteroposterior patterning of the mouse skeleton, Dev. Growth Differ. 50 (2008) 321–330. A.K. Voss, C. Collin, M.P. Dixon, T. Thomas, Moz and retinoic acid coordinately regulate H3K9 acetylation, Hox gene expression, and segment identity, Dev. Cell 17 (2009) 674–686. K. Suzuki, C. Oneyama, H. Kimura, S. Tajima, M. Okada, Down-regulation of the tumor suppressor C-terminal Src kinase (Csk)-binding protein (Cbp)/PAG1 is mediated by epigenetic histone modifications via the mitogen-activated protein kinase (MAPK)/phosphatidylinositol 3-kinase (PI3K) pathway, J. Biol. Chem. 286 (2011) 15698–15706. T. Zuo, T.M. Liu, X. Lan, Y.I. Weng, R. Shen, F. Gu, Y.W. Huang, S. Liyanarachchi, D.E. Deatherage, P.Y. Hsu, C. Taslim, B. Ramaswamy, C.L. Shapiro, H.J. Lin, A.S. Cheng, V.X. Jin, T.H. Huang, Epigenetic silencing mediated through activated PI3K/AKT signaling in breast cancer, Cancer Res. 71 (2011) 1752–1762. F. De Santa, M.G. Totaro, E. Prosperini, S. Notarbartolo, G. Testa, G. Natoli, The histone H3 lysine-27 demethylase Jmjd3 links inflammation to inhibition of polycombmediated gene silencing, Cell 130 (2007) 1083–1094. F. Lan, P.E. Bayliss, J.L. Rinn, J.R. Whetstine, J.K. Wang, S. Chen, S. Iwase, R. Alpatov, I. Issaeva, E. Canaani, T.M. Roberts, H.Y. Chang, Y. Shi, A histone H3 lysine 27 demethylase regulates animal posterior development, Nature 449 (2007) 689–694. Y. Zhang, J. Li, X. Hu, N. Li, The activation of Akt/protein kinase B signaling affects the expression of histone H3 lysine 27 methylase and demethylase genes in mouse cell lines, J. Anim. Vet. Adv. 11 (2012) 2469–2479.