STEM CELLS AND DEVELOPMENT Volume 27, Number 17, 2018 Mary Ann Liebert, Inc. DOI: 10.1089/scd.2018.0059

Role of Epigenetic Regulation by the REST/CoREST/HDAC Corepressor Complex of Moderate NANOG Expression in Chicken Primordial Germ Cells

Hyun Gyo Jung,* Young Sun Hwang,* Young Hyun Park, Ho Yeon Cho, Deivendran Rengaraj, and Jae Yong Han

Primordial germ cells (PGCs), the precursors of gametes, have regulatory mechanisms involving factors and epigenetic modifications. The NANOG is a key regulator of and embryonic stem cell characteristics. However, the epigenetic regulation of NANOG with a histone deacetylase (HDAC) complex in PGCs has not been studied. In this study, we investigated the epigenetic regulation, and in particular the histone acetylation, of NANOG in chicken PGCs. Intriguingly, although NANOG was highly activated in chicken PGCs, the upstream region of its promoter was moderately suppressed by histone deacetylation. HDAC inhibition induced histone H3 lysine 9 acetylation (H3K9ac) and derepressed NANOG expression. Furthermore, knockdown studies revealed that HDAC complex members, such as RE1-silencing transcription factor (REST) and REST corepressor 3 (RCOR3), are important epigenetic modulators of NANOG expression in chicken PGCs. We demonstrate that moderate regulation of NANOG by the REST/CoREST/HDAC complex might be crucial for maintaining the integrity of PGCs.

Keywords: chicken, primordial germ cells, NANOG, RE1-silencing transcription factor, REST corepressor, histone deacetylation

Introduction VASA and deleted in azoospermia like (DAZL), in a pre- determined manner [7,8]. In contrast with mammals, chicken rimordial germ cells (PGCs) are the precursors of PGCs migrate to the genital ridge through blood vessels after Psperm and oocytes, which are the only cells that transmit specification [9]. genetic information to the next generation. PGCs undergo In terms of epigenetic regulation, the dynamics and land- epigenetic dynamics during embryogenesis to establish cel- scape of histone modification have not been established yet in lular specification and suppress somatic lineage differentia- avian PGCs. In recent study, global enrichment of H3K9 tri- tion [1]. The epigenetic status of PGCs is regulated by DNA methylation (H3K9me3) was observed in chicken PGCs [10]. methylation and histone modifications (histone acetylation or Also, H3K9ac formed foci in euchromatin in chicken PGCs as deacetylation), which have roles in regulating the expression the activating histone acetylation marker [10], indicating this Downloaded by Seoul National Univ. Medical College from www.liebertpub.com at 11/01/18. For personal use only. of genes related to pluripotency and germness [2]. mark could play an important role in epigenetic regulation in Specifically, histone H3 lysine 9 acetylation (H3K9ac) chicken PGCs. However, regulation by his- transiently and markedly increases in PGCs to establish tone modification in chicken PGCs has been largely unknown. transcriptional activation of chromatin at the point of entry NANOG is an important transcription factor that regulates into the genital ridge and decreases during colonization [3]. pluripotency in early embryos, embryonic stem cells (ESCs), Also, maintenance of H3K9ac is involved in proper chro- and induced pluripotent stem cells (iPSCs) [11]. In mammals, mosomal segregation during the formation of gametes [4,5]. NANOG expression was also found from migrating germ cells Thus, histone acetylation is a crucial epigenetic marker of to gonadal germ cells [12,13], and shown to prevent apoptosis germ cell specification and development. of migrating germ cells [14]. Moreover, in later stage, NANOG Avian PGCs are important for biotechnology and germ cell is involved in differentiation and maturation of mouse PGCs in biology due to their unique characteristics [6]. Chicken PGCs genital ridge [15–17]. In chicken, NANOG is required for the are specified by maternally inherited germ plasm, including maintenance of pluripotency in chicken ESCs [18] and was

Department of Agricultural Biotechnology, Research Institute of Agriculture and Life Sciences, College of Agriculture and Life Sciences, Seoul National University, Seoul, Korea. *These authors contributed equally to this work.

1215 1216 JUNG ET AL.

detected during early embryonic development from preovula- DMEM (Gibco) supplemented with 10% FBS. The cultured tory oocyte to intrauterine stages [19]. Also, Jak1/Stat3 sig- PGCs and DF-1 were treated with HDAC inhibitors, 0.25 mM naling supports NANOG expression and pluripotency in early valproic acid (VPA; Sigma-Aldrich), and 50 nM romidepsin chicken embryos [20]. (Selleckchem, Boston, MA), for three consecutive days. In chicken PGCs, NANOG is expressed in the germinal crescent at Hamburger and Hamilton stage 5 (HH5) to HH8 Immunostaining [21]. In addition, NANOG expression was similar to the PGC-specific genes CVH and dead end homolog 1 (DND1) VPA- and romidepsin-treated cells were harvested and wa- during PGC migration and colonization in gonads [21], in- shed twice in phosphate-buffered saline (PBS) and fixed with dicating a role for NANOG in maintaining the self-renewal 4% paraformaldehyde for 20 min. After permeabilization with ability of chicken PGCs in various stages. The recent report 0.1% Tween-20 and 1% Triton X-100, nonspecific binding was showed that NANOG solely activated enhancers upstream blocked with 1% normal goat serum. Cells were then incubated of germ cell specifiers to generate PGC-like cells from with rabbit anti-acetyl-histone H3K9 (06-942; Milli- epiblast-like cells, but not in ESCs in mouse [22], indicating pore, Bedford, MA), mouse anti-acetyl-histone H3 lysine 27 that its function differs according to cell type. (H3K27ac) antibody (GTX50903; GeneTex, CA), rabbit anti- The role of histone acetylation and regulation of the trimethyl-histone H3K9 antibody (07-442; Millipore), and rabbit histone deacetylase (HDAC) complex in pluripotency and anti-trimethyl-histone H3K27 (H3K27me3) (07-449, Millipore) somatic cell reprogramming in ESCs and iPSCs has been at 4C overnight. Cells incubated in the absence of primary reported [23–25]. Despite the emerging role of NANOG in antibody were used as a negative control. After extensive germ cell development, the role of histone acetylation and washing, cells were incubated with Alexa Fluor 594-conjugated the HDAC complex in the regulation of NANOG expression goat anti-rabbit IgG H&L (ab150080; Abcam, Cambridge, in avian PGCs is unclear. United Kingdom), Alexa Fluor 488-conjugated goat anti- In this study, we demonstrated the epigenetic regulation of rabbit IgG H&L (ab150077; Abcam), and FITC-conjugated NANOG expression by H3K9ac and the HDAC corepressor goat anti-mouse IgG (sc-2010; Santa Cruz, CA) for 1 h. complex in chicken PGCs. We used an HDAC inhibitor to After washing, cells were mounted with 4¢,6-diamidino-2- induce aberrant histone acetylation and evaluated the ex- phenylindole (DAPI) and visualized using Ti-U fluorescence pression of germ cell-specific genes, including NANOG. In microscope (Nikon, Tokyo, Japan). addition, upstream regulators of NANOG, such as DNA- dependent factors and HDAC complexes, were examined to RNA isolation and reverse transcription-quantitative evaluate epigenetic regulation in chicken PGCs. This study polymerase chain reaction increases our understanding of the molecular mechanisms Total RNA was extracted using TRIzol reagent (Invitrogen, involved in the fundamental specification of germ cells. Carlsbad, CA) according to the manufacturer’s protocol. One microgram of total RNA from each sample was reverse- Materials and Methods transcribed using the Superscript III First-strand Synthesis Experimental animals and care System (Invitrogen) according to the manufacturer’s proto- col. The polymerase chain reaction (PCR) mixture contained The care and experimental use of White Leghorn (WL) 2 mL of PCR buffer, 0.5 mL of 10 mM dNTP mixture, 10 pmole chickens were approved by the Institute of Laboratory Animal each of forward and reverse primers (Table 1), 1 mL of com- Resources, Seoul National University (SNU-150827-1). Chick- plementary DNA (cDNA), and 1 U of Taq DNA polymerase ens were maintained according to a standard management pro- in a 20 mL final volume. PCR amplification involved an initial gram at the University Animal Farm, Seoul National University, incubation at 95C for 10 min, followed by 40 cycles at 95C Korea. The procedures for animal management, reproduction, for 30 s, 60C for 30 s, and 72C for 30 s. PCR was terminated and embryo manipulation adhered to the standard operating with a final incubation at 72C for 5 min, and at the dissoci- protocols of our laboratory. ation temperature. Reverse transcription-quantitative poly- merase chain reaction (RT-qPCR) was performed using a

Downloaded by Seoul National Univ. Medical College from www.liebertpub.com at 11/01/18. For personal use only. Culture of chicken PGCs and chemical treatment StepOnePlus real-time PCR system (Applied Biosystems, Foster City, CA) with EvaGreen (Biotium, Hayward, CA). Chicken PGCs were cultured in accordance with our stan- Each test sample was assayed in triplicate. Relative quanti- dard procedure [26,27]. Briefly, PGCs from male WL embry- fication of target gene expression was performed using onic gonads at 6 days old (HH stage 28) were maintained in the following formula: 2-DDCt,whereDDCt = (Ct of the target knockout Dulbecco’s modified Eagle’s medium (DMEM; gene - (Ct of GAPDH + Ct of ACTB)/2) treatment - (Ct of the Gibco, Grand Island, NY) supplemented with 20% fetal bovine target gene - (Ct of GAPDH + Ct of ACTB)/2) control. serum (FBS; Hyclone, South Logan, UT), 2% chicken serum (Sigma-Aldrich, St. Louis, MO), 1· nucleosides (Millipore, Histone extraction and western blot analysis Billerica, MA), 2 mM L-glutamine (Gibco), 1· nonessential amino acids (Gibco), b-mercaptoethanol (Gibco), 1 mM so- Cells were harvested and washed in ice-cold PBS for two dium pyruvate (Gibco), and 1· antibiotic/antimycotic (Gibco). times. Harvested cells were resuspended in Triton extraction Human basic fibroblast growth factor (Koma Biotech, Seoul, buffer [TEB; PBS containing 0.5% Triton · 100 (v/v), 2 mM Korea) at 10 ng/mL was used for PGC self-renewal. The cul- phenylmethylsulfonyl fluoride, 0.02% (w/v) NaN3]. PBS tured PGCs were subcultured onto mitomycin-inactivated and TEB contain 5 mM sodium butyrate to preserve levels mouse embryonic fibroblasts at 5- to 6-day intervals by gentle of histone acetylation. After harvest of cell pellets, the nu- pipetting without any enzyme treatment. DF-1 was cultured in clear pellets were extracted with 0.2 N HCl for overnight at THE MODERATE EXPRESSION OF NANOG IN CHICKEN PGCS 1217

Table 1. Primers Used for Reverse Transcription-Quantitative Polymerase Chain Reaction Gene symbol Description Accession no. Primer sequence (5¢ / 3¢) NANOG Nanog NM_001146142 F: AACTCTGCGGGGCTGTCTTG R: AAAAGTGGGGCGGTGAGATG POUV/POU5F3 POU domain class 5 transcription NM_001110178 F: TGAAGGGAACGCTGGAGAGC factor 3 R: ATGTCACTGGGATGGGCAGAC VASA Chicken VASA homolog NM_204708 F: CCTTGCAGCCTTTCTTTGTC R: ACGACCAGTTCGTCCAATTC DAZL Deleted in azoospermia like NM_204218 F: CAACTATCAGGCTCCACCAC R: CTCAGACGGTTTTCAGGGTT T brachyury transcription factor NM_204940 F: ACGCCATGTACTCCTTCCTG R: TGTTGGTGAGCTTGACCTTG GSC Goosecoid homeobox NM_205331 F: AACGGAGCTGCAGTTACTCAA R: TTCCATTTCTGTGCATTTTCC SOX7 SRY (sex determining region Y)- XM_420036 F: CTGGGAGAGATGGACCGTAA box 7 R: CCTTCCTCAGCAGGTCTACG GATA4 GATA-binding 4 NM_420041 F: ACCAGTCAGCCATCAGCCAG R: AGAGCAGGGGAGGAGGGAAA SOX3 SRY (sex determining region Y)- NM_204195 F: GACAAAACGAAAAACCGACCC box 3 R: ATTGCGGACACGAACCGAG AFP Alpha-fetoprotein XM_015276399 F: CTCCTATGCTCTCAGGAGGC R: TTCTTTGCAGCACTGTTCCC MMP7 Matrix metallopeptidase 7 NM_001006278 F: GCTGCCATTCTCCTACCTCA R: CACATCTGGGCTGTTGCATT NR6A1 subfamily 6 group XM_015279584 F: GTTTGCCAGGACTTCACAGAG A member 1 R: CGGGACATTCACCATCTTTC CXCR4 C-X-C motif chemokine receptor 4 NM_204617 F: TCTGTGGCTGACCTCCTCTT R: TCAGGCACTGTGAGAAGCAC REST RE1-silencing transcription factor XM_015276447 F: AAAGAGCAAACAAAAGGGGA R: TTGCTCGTTGGCTTCTTTT SIN3A SIN3 transcription regulator family NM_001293184 F: GGAAGAGGAGGAAGAAGAGG member A R: CTCTCCCTTGTCTCCTCTTC SIN3B SIN3 transcription regulator family XM_015300175 F: GGGGAGAAAAGAAAAGACCAA member B R: CCTCGCTTCCTTTCGTTC MTA1 Metastasis-associated 1 NM_001012953 F: AACAAGCCAAACCCCAACC R: GTTTGGTCCTGGTCTCTCTC MTA3 Metastasis-associated 1 family, XM_015283840 F: GCTCTTCCTTTCTCGCCAGT member 3 R: ACGGTCTGTAAGGGGGCTAT RCOR1 REST corepressor 1 XM_015287816 F: AACAGAGCGAAGAGGAAACC R: CTCTTTGCCGTGTTCTGC RCOR3 REST corepressor 3 NM_001079727 F: ATCGACAGGCTCGAAAGCTTGCTA R: CATTCTCAGCGTTCCAAATGCCG GAPDH Glyceraldehyde-3-phosphate NM_204305 F: GGTGGTGCTAAGCGTGTTAT dehydrogenase R: ACCTCTGTCATCTCTCCACA ACTB Actin, beta NM_205518 F: AGGAGATCACAGCCCTGGCA R: CAATGGAGGGTCCGGATTCA

Downloaded by Seoul National Univ. Medical College from www.liebertpub.com at 11/01/18. For personal use only. 4C. Supernatants containing histone protein are neutralized levels were quantitated with ImageJ (National Institutes of by 2 M NaOH at 10% volume of 0.2 N HCl. Protein (*10 mg) Health, Bethesda, MD) [28]. from total cell lysate was loaded in each lane for separation on 15% SDS-polyacrylamide gels. Resolved were Prediction of the RE1-silencing transcription transferred onto a Hybond 0.45 PVDF membrane (GE factor-binding site in the proximate promoter Healthcare Life Sciences, Little Chalfont, United Kingdom) of NANOG, POUV, and VASA and blocked with 3% skim milk for 1 h at room tempera- ture (Sigma-Aldrich). The blocked membrane was incu- The region upstream of NANOG, POUV,andVASA tran- bated overnight at 4C in the presence of primary antibody scriptional start site (TSS) was predicted by MatInspector, a (dilution: 1:3,000) and then for 1 h with horseradish Genomatix program (www.genomatix.de) using TRANSFAC peroxidase-conjugated secondary antibody (31460; Ther- matrices (vertebrate matrix; core similarity 1.0 and matrix mo Fisher Scientific, Inc., Waltham, MA) at a dilution of similarity 0.8). 1:4,000. Primary were as follows: anti-histone Chromatin immunoprecipitation H3 (ab1791; Abcam) and anti-H3K9Ac antibody (06-942; Millipore). Immunoreactive proteins were visualized with Treated cells were fixed in 1% methanol-free formalde- the ECL western blot detection system (Thermo Fisher Sci- hyde (Thermo Fisher Scientific, Inc.) in DMEM for 20 min at entific, Inc.). All blots were performed in triplicate, and protein room temperature, followed by 10-min blocking in 125 mM 1218 JUNG ET AL.

Table 2. Primers Used for Chromatin Immunoprecipitation-Polymerase Chain Reaction and Chromatin Immunoprecipitation-Quantitative Polymerase Chain Reaction

Primer Description Primer sequence (5¢ / 3¢) ChIP-REST REST binding site on NANOG promoter F: GAGGCGTAGGTGGGTAGCAAGAC R: GTCAGTTGATTGGAGGGGAGTGG

glycine. Cells were rinsed two times in ice-cold PBS and buffer, 0.5 mL of 10 mM dNTP mixture, 10 pmol of each of harvested. Suspension cultured cells were fixed in the plate the forward and reverse primers (Table 2), 1 mL of ChIP and harvested by pipetting after washing with PBS. Fixed DNA, and 1 U of Taq DNA polymerase in a 20 mL final cell pellets were dissociated in radioimmunoprecipitation volume. ChIP-PCR amplification involved an initial incu- (RIPA) lysis buffer (10 mM Tris-HCl, pH 7.5, 1 mM EDTA, bation at 95C for 10 min, followed by 30 cycles at 95C 0.5 M EGTA, 1% Triton X-100, 0.1% SDS, 0.1% Na- for 30 s, 60C for 30 s, and 72C for 30 s. The reaction deoxycholate, and 140 mM NaCl; Thermo Fisher Scientific, was terminated by a final incubation at 72C for 5 min. Inc.) with protease and phosphatase inhibitors (Sigma- For chromatin immunoprecipitation-quantitative polymerase Aldrich). Chromatin was sheared to an average length of chain reaction (ChIP-qPCR), the reaction mixture, thermal 500–700 base pairs (bp) using a sonicator (Q500; Qsonica, conditions, and all other procedures were as stated above. Newtown, CT). DNA concentration and the efficiency of ChIP-qPCR was performed using a StepOnePlus real-time shearing were determined by agarose gel electrophoresis. PCR system (Applied Biosystems) with EvaGreen (Bio- An anti-H3K9Ac antibody (06-942; Millipore) was used for tium). Each test sample was assayed in triplicate. Relative immunoprecipitation. Ten micrograms of sonicated chroma- quantification of gene expression was performed using the tin samples was used. Chromatin immunoprecipitation (ChIP) following formula: ((2-DCt) · 100%), where DCt = (Ct [IP] - was performed using a One-Step ChIP Kit (ab117138; Ab- Ct [Input · DF]). The DF (default input fraction) was 1%, cam) according to the manufacturer’s instructions (ab117138; which is a dilution factor of 6.644 cycles (ie, log2 of 100). Abcam) and nonimmune IgG was used as a negative control. The bound chromatin/antibody complex was incubated in Small interfering RNA transfection in chicken PGCs reverse crosslinking buffer for 2 h at 65C, followed by treatment with proteinase K. Chicken PGCs were seeded at 2.0 · 105 in 1 mL of me- dium per well of a 12-well plate. Cells were transfected with Chromatin immunoprecipitation-polymerase chain 100 pmol of each HDAC small interfering RNA (siRNA) reaction and chromatin immunoprecipitation- (Table 3) using RNAiMAX (Invitrogen). Negative control quantitative polymerase chain reaction siRNA with no complementary sequence in the chicken genome was used as a control. After transfection for 48 h, For total chromatin control samples (inputs) purified be- total RNA was extracted using TRIzol reagent (Invitrogen) fore immunoprecipitation, add 10 mL of each chromatin according to the manufacturer’s protocol. The knockdown sample to 88 mL of buffer and 2.5 mL of proteinase K. For efficiency of HDAC complex siRNAs and their effects on chromatin immunoprecipitation-polymerase chain reaction the expression of pluripotency-, germ cell-, and three germ (ChIP-PCR), the reaction mixture comprised 2 mL of PCR layer-specific marker genes were assayed by RT-qPCR.

Table 3. List of Small Interfering RNA Sequences for Knockdown Experiments Gene symbol Accession no. Primer sequence (5¢ / 3¢) REST XM_015276447 S: CAGAGACTGTCAGGAAGCTCAGAAA Downloaded by Seoul National Univ. Medical College from www.liebertpub.com at 11/01/18. For personal use only. AS: UUUCUGAGCUUCCUGACAGUCUCUG SIN3A NM_001293184 S: ACAUCUUCAUGAGGCUGCAUCAGAU AS: AUCUGAUGCAGCCUCAUGAAGAUGU SIN3B XM_015300175 S: CAGCGGGCUGUUGUUGCAUUACUCA AS: UGAGUAAUGCAACAACAGCCCGCUG MTA1 NM_001012953 S: CCGCCAAUGGAAAUGUGGAAGCAAA AS: UUUGCUUCCACAUUUCCAUUGGCGG MTA3 XM_015283840 S: CAUGUACCGAGUGGGAGAUUAUGUU AS: AACAUAAUCUCCCACUCGGUACAUG RCOR1 XM_015287816 S: CCGACGUCGUUUCAACAUAGAUGAA AS: UUCAUCUAUGUUGAAACGACGUCGG RCOR3 NM_001079727 S: CGGAUCUCCCUAACUUCACUCCCUU AS: AAGGGAGUGAAGUUAGGGAGAUCCG Nonspecific siRNAa S: CCUACGCCACCAAUUUCGU AS: GGAUGCGGUGGUUAAAGCA

aNonspecific siRNA has no complementary sequence in the chicken genome and was used as a control for gene silencing. siRNA, small interfering RNA. THE MODERATE EXPRESSION OF NANOG IN CHICKEN PGCS 1219

Statistical analysis H3K9ac compared with untreated control PGCs (Fig. 1A). Next, PGCs were treated with 50 nM Romidepsin in vitro, a Data are expressed as mean – standard error from three specific inhibitor of HDAC1 and 2 [30], which also resulted independent biological experiments. Student’s t-test was per- in a global increase in H3K9ac (Fig. 1A). Quantification of formed to evaluate differences between experimental groups H3K9ac after chemical treatment confirmed the significant using GraphPad Prism v.5 (GraphPad Software, La Jolla, CA). elevation of H3K9ac in VPA- and romidepsin-treated PGCs A value of P < 0.05 was considered indicative of statistical compared with control PGCs (Fig. 1B, C). However, ro- significance (***P < 0.001, **P < 0.01, and *P < 0.05). midepsin is a more effective HDAC inhibitor in PGCs compared with VPA (Fig. 1C). Therefore, only romidepsin Results was used for subsequent treatments. In addition to H3K9ac, Upregulation of H3K9ac and gene expression we examined H3K27ac, H3K9me3, and H3K27me3 in in PGCs by HDAC inhibitors romidepsin-treated PGCs (Supplementary Fig. S1; Supple- mentary Data are available online at www.liebertpub.com/ To assess the influence of HDAC on epigenetic regulation scd). The global increase in H3K27ac and decrease in and gene expression in chicken PGCs, we treated PGCs with H3K9me3 and H3K27me3 were shown in chicken PGCs by 0.25 mM VPA, an inhibitor of class I HDACs, closely as- romidepsin treatment, indicating the treatment modulated sociated with transcriptional activity [29]. Treatment of overall epigenetic marks in an active status. Next, we examined PGCs with VPA in vitro resulted in a global increase in romidepsin effects on the expression of a set of genes in PGCs. Downloaded by Seoul National Univ. Medical College from www.liebertpub.com at 11/01/18. For personal use only.

FIG. 1. Effect of HDAC inhibitors on H3K9ac and expression of NANOG in chicken PGCs. (A) Cultured PGCs were treated with 0.25 mM VPA or 50 nM romidepsin for 3 days, and subjected to immunocytochemical analysis of H3K9ac levels. Scale bar, 100 mm. (B) Western blot analysis of H3K9ac after chemical treatment. Histone H3 was used as control. (C) Relative quantification of H3K9ac to H3 in B. (D) Expression of NANOG in PGCs treated with romidepsin by RT- qPCR. RT-qPCR was conducted in triplicate, and the relative gene expression in treatments was normalized to expression of GAPDH and ACTB. Error bars represent SE of biological triplicate analyses. Significant differences between groups are indicated as ***P < 0.001 and *P < 0.05. H3K9ac, histone H3 lysine 9 acetylation; HDAC, histone deacetylase; PGCs, primordial germ cells; RT-qPCR, reverse transcription-quantitative polymerase chain reaction; SE, standard error; VPA, valproic acid. Color images available online at www.liebertpub.com/scd FIG. 2. Effect of romidepsin on acetylated H3K9 levels at the REST-binding site in the NANOG promoter. (A) The REST- binding site in the proximal NANOG promoter. Arrows indicate primer-binding site. (B) Agarose gel electrophoresis of ChIP- PCR after treatment with HDAC inhibitor. %Input = 100-fold diluted DNA before selection; IgG = control for nonspecific antibody binding to DNA. (C) ChIP-qPCR amplification of the proximal REST-binding site (188 bp from the TSS) after treatment with HDAC inhibitor. Untreated PGCs were used as the control. Error bars represent SE of biological triplicate analyses. Significant differences between groups are indicated as *P < 0.05. bp, base pairs; ChIP-PCR, chromatin immuno- precipitation-polymerase chain reaction; ChIP-qPCR, chromatin immunoprecipitation-quantitative polymerase chain reaction; H3K9ac, histone H3 lysine 9 acetylation; HDAC, histone deacetylase; IgG, immunoglobulin G; PGCs, primordial germ cells; RE1-silencing transcription factor; SE, standard error; TSS, transcriptional start site. Color images available online at www.liebertpub.com/scd Downloaded by Seoul National Univ. Medical College from www.liebertpub.com at 11/01/18. For personal use only.

FIG. 3. REST regulates H3K9ac to repress NANOG expression in chicken PGCs. (A) Expression of REST in DF-1 cells and chicken PGCs by RT-qPCR. (B) Effect of siRNA-mediated knockdown of REST determined by RT-qPCR. (C) RT- qPCR analysis of POUV, NANOG, VASA, and DAZL expression in PGCs after knockdown of REST. RT-qPCR was conducted in triplicate, and the relative gene expression in treatments was normalized to expression of GAPDH and ACTB. (D, E) Suppression of REST increased H3K9ac levels at the REST-binding site in the proximal NANOG promoter. %Input = 100-fold diluted DNA before selection; IgG = control for nonspecific antibody binding to DNA. Nonspecific siRNA-treated PGCs were used as the control. Error bars represent SE of biological triplicate analyses. Significant dif- ferences between groups are indicated as ***P < 0.001, **P < 0.01, and *P < 0.05. siRNA, small interfering RNA; REST, RE1-silencing transcription factor; H3K9ac, histone H3 lysine 9 acetylation; PGCs, primordial germ cells; IgG, im- munoglobulin G; RT-qPCR, reverse transcription-quantitative polymerase chain reaction; SE, standard error.

1220 THE MODERATE EXPRESSION OF NANOG IN CHICKEN PGCS 1221

NANOG expression was significantly upregulated following was efficiently knocked down using REST-targeting siRNA. romidepsin treatment (Fig. 1D). Furthermore, expression of the Following successful knockdown of REST (Fig.3B),we pluripotency and germness-related genes, POUV and VASA, quantified the expression of pluripotency- and germness- was significantly upregulated following romidepsin treatment related genes by RT-qPCR. NANOG and VASA expression (Supplementary Fig. S2A), indicating that active genes in was significantly upregulated; the increase in NANOG ex- chicken PGCs, NANOG, POUV, and VASA, were further in- pression was considerable. In contrast, POUV and DAZL duced by HDAC inhibitor. Therefore, expression of these expression was unchanged (Fig. 3C). REST-binding sites genes may be suppressed by an HDAC complex in chicken were also found in the proximal POUV and VASA promoters PGCs. In contrast, the expression of three germ layer markers, (Supplementary Fig. S4). Otherwise, REST knockdown in- BRACHYURY, GSC, SOX7, GATA4, and SOX3, which are not creased VASA expression but not POUV expression (Fig. 3C), constitutively expressed in chicken PGCs, was unchanged even indicating REST to be involved in transcriptional regulation following romidepsin treatment (SupplementaryFig. S2B). We of VASA but not POUV. ChIP-PCR and ChIP-qPCR showed found that the germ cell tumor markers, such as AFP, MMP7, that suppression of REST increased H3K9ac at the REST- and NR6A1, were upregulated in romidepsin-treated PGCs binding site in the proximate NANOG promoter (Fig. 3D, E). (Supplementary Fig. S2C). While romidepsin-treated PGCs These results indicate that the REST regulates showed downregulated expression of migration-related gene, the level of H3K9ac to control NANOG transcription in CXCR4 (Supplementary Fig. S2D). In chicken DF-1 cells, only chicken PGCs. NANOG was expressed after romidepsin treatment than that of POUV, DAZL,andVASA (Supplementary Fig. S3). The CoREST complex member RCOR3 mediates epigenetic regulation of NANOG expression HDAC inhibitor increases acetylated H3K9 levels Next, we elucidated the HDAC complex that mediates the at RE1-silencing transcription factor-binding sites change in H3K9ac in PGCs. We first determined the expression in the NANOG promoter of candidate members of the HDAC complex, including To identify the transcriptional repressor involved in metastasis-associated 1 (MTA1), MTA3, REST corepressor 1 modulation of NANOG expression by romidepsin, we pre- (RCOR1), RCOR3, SIN3 transcription regulator family dicted the regions affected by HDAC inhibition in the pu- member A (SIN3A), and SIN3B, in chicken PGCs and DF-1 tative promoter of NANOG.UpstreamofNANOG, we found cells. Reverse transcription-polymerase chain reaction (RT- a binding site of the transcriptional repressor, RE1-silencing PCR) showed that all HDAC complex members, except for transcription factor (REST), 188 bp from the TSS using SIN3A, were expressed in PGCs and DF-1 cells (Fig. 4A). MatInspector (Fig. 2A). Next, to examine the effect of the Second, the HDAC complex members MTA1, MTA3, RCOR1, HDAC inhibitor on the REST-binding site, we performed RCOR3,andSIN3B were knocked down in PGCs using spe- ChIP analysis on this site of the NANOG promoter (188 bp cific siRNAs. After successful knockdown of HDAC complex from TSS). ChIP-PCR and ChIP-qPCR analyses showed members in PGCs (Fig. 4B), we examined the expression of that romidepsin treatment resulted in a significant increase POUV, NANOG, VASA,andDAZL using RT-qPCR. Among of H3K9ac in the region nearby the REST-binding site the five members of the HDAC complex, only knockdown of in the NANOG promoter (Fig. 2B, C). Therefore, H3K9ac RCOR3 showed significant upregulation of NANOG expres- level and the expression of NANOG could be regulated by sion (Fig. 4C). In addition, repression of RCOR3 by siRNA an HDAC complex possibly via the REST-binding site in enriched the acetylated H3K9 residue on the REST site in the chicken PGCs. NANOG promoter (Fig. 4D, E). These results suggest that RCOR3, together with REST, deacetylates the region upstream REST represses NANOG expression via H3K9ac of the NANOG promoter, which represses NANOG expression in chicken PGCs in chicken PGCs.

We next performed an expression and knockdown anal- Discussion ysis of REST in chicken PGCs. RT-qPCR showed that REST Downloaded by Seoul National Univ. Medical College from www.liebertpub.com at 11/01/18. For personal use only. expression was significantly higher in PGCs than DF-1 cells NANOG is a notable homeodomain transcription factor (Fig. 3A). Subsequently, we conducted a loss-of-function involved in maintaining and establishing pluripotency in study of REST on gene expression in PGCs. REST expression mammals [11,31,32]. During somatic cell reprogramming ‰ FIG. 4. The CoREST complex member RCOR3 modulates epigenetic regulation of NANOG expression. (A) Expression of HDAC complexes in chicken PGCs and DF-1 cells by RT-PCR. (B) Effect of siRNA-mediated knockdown of HDAC complex members by RT-qPCR. (C) RT-qPCR analysis of POUV, NANOG, VASA, and DAZL expression in PGCs after knockdown of HDAC complex members. RT-qPCR was conducted in triplicate, and the relative gene expression in treatments was normalized to expression of GAPDH and ACTB. (D, E) Suppression of RCOR3 resulted in an increased level of H3K9ac at the REST-binding site in the proximal NANOG promoter. %Input = 100-fold diluted DNA before selection; IgG = control for nonspecific antibody binding to DNA. Nonspecific siRNA-treated PGCs were used as the control. Error bars represent SE of biological triplicate analyses. Significant differences between groups are indicated as ***P < 0.001, **P < 0.01, and *P < 0.05. RCOR3, REST corepressor 3; HDAC, histone deacetylase; REST, RE1-silencing transcription factor; H3K9ac, histone H3 lysine 9 acetylation; PGCs, primordial germ cells; RT-PCR, reverse transcription-polymerase chain reaction; RT-qPCR, reverse transcription-quantitative polymerase chain reaction; IgG, immunoglobulin G; siRNA, small interfering RNA; SE, standard error. Downloaded by Seoul National Univ. Medical College from www.liebertpub.com at 11/01/18. For personal use only.

1222 THE MODERATE EXPRESSION OF NANOG IN CHICKEN PGCS 1223

FIG. 5. Epigenetic regulation of NANOG expression in chicken PGCs. Epigenetic regulation of HDAC complex controls NANOG expression in chicken PGCs. HDAC inhibitor represses deacetylation activity of the HDAC, resulting in upre- gulation of NANOG. The transcription factor REST binds to the upstream region of the NANOG promoter and represses NANOG expression with RCOR3 and HDAC. REST/CoREST/HDAC complex maintains moderate NANOG expression in chicken PGCs. HDAC, histone deacetylase; REST, RE1-silencing transcription factor; RCOR3, REST corepressor 3; PGCs, primordial germ cells. Color images available online at www.liebertpub.com/scd

and self-renewal in ESCs, NANOG cooperates with the mors [39]. Therefore, gene expression in PGCs may need to SIN3A/HDAC corepressor complex to activate pluripotency be regulated to the proper level. genes and repress reprogramming barrier genes [25]. In REST is a zinc finger transcription factor that represses terms of germ cell specification from the epiblast in vitro, the expression of multiple genes during embryogenesis, and NANOG activates the expressionofgermcell-specific regulates neuronal gene expression and neural stem cell transcription factors such as PRDM1 and PRDM14 as (NSC) fate in vertebrates [40–42]. In mouse, Rest has dif- an enhancer in mouse [22], suggesting a crucial role for ferential regulatory roles in ESCs and NSCs [43]. Even, Rest NANOG in germ cells. In chicken, NANOG is expressed in shared common targets with pluripotency-related factors, in- both ESCs and PGCs [18,21]; however, epigenetic regula- cluding Nanog [43], indicating the possible interaction be- tion of NANOG expression by the HDAC complex in PGCs tween REST and NANOG in other cell types. In the case of has not been investigated. chicken PGCs, to moderately regulate gene expression, Generally, acetylation of histone tails, including H3K9ac, especially NANOG, such as REST seem to promote activates gene expression. In mouse, H3K9ac was increasing histone deacetylation of gene promoters. According to our sharply in PGCs at embryonic day (E) 10.5 when entering results, REST is involved in repression of NANOG expres- into genital ridges, but decreasing after settlement [3]. In sion and reduction of H3K9ac in promoter. migrating chicken PGCs, H3K9ac was restricted in foci of Furthermore, among the members of HDAC corepressor loose chromatin structures, which could be involved in germ complexes, such as CoREST [23], MTA [24], and SIN3 [25], cell-specific gene expression [10]. Treatment with an HDAC RCOR3 is a major upstream regulator of histone acetylation in inhibitor resulted in global elevation of the H3K9ac level. NANOG promoter in chicken PGCs. Mouse ESCs showed Indeed, treatment with the HDAC1/2-specific inhibitor ro- varying numbers of recruited corepressors with Rest in target midepsin resulted in a greater elevation of the H3K9ac level, sites, indicating different interactions among their complexes implying that HDAC1 or 2 could be the active deacetylase in diverse regions [44]. In terms of early neuronal differenti- in chicken PGCs. Surprisingly, although NANOG, as well as ation in mouse ESCs, the REST-CoREST complex modulates POUV and VASA, is expressed at high levels in chicken gene expression [45]. In this study, our results demonstrated PGCs, the expression was increased following the increase that the REST/CoREST/HDAC complex modulates the in H3K9ac or indirect effects induced by treatment with H3K9ac level in upstream of NANOG to regulate its moderate HDAC inhibitor. NANOG expression was also increased in expression in chicken PGCs. romidepsin-treated chicken DF-1 cells, suggesting that the In conclusion, we report that the CoREST/HDAC complex transcription of NANOG is epigenetically regulated in with REST, which binds to the NANOG promoter, functions chicken, regardless of cell type. In contrast, the expression as a transcriptional corepressor to regulate NANOG ex-

Downloaded by Seoul National Univ. Medical College from www.liebertpub.com at 11/01/18. For personal use only. levels of differentiation markers were unaffected. In mam- pression in chicken PGCs (Fig. 5). The CoREST/REST/ malian ESCs, treatment with low concentrations of HDAC HDAC repressor complex in PGCs is involved in maintaining inhibitors results in enrichment of H3K9ac to promote self- moderate expression of NANOG, which is a critical tran- renewal and pluripotency [33,34]. However, PGCs, unlike scription factor for germ cell characteristics. Therefore, ESCs, generate sperm and oocytes and transmit genetic in- regulation of NANOG expression by the REST/CoREST/ formation to the next generation; therefore, gene expression HDAC complex could be involved in maintaining the PGC in PGCs should be tightly regulated. In mammals, over- integrity. expression of NANOG causes germ cell tumors [35,36]; in contrast, suppression of NANOG induces apoptosis of germ Acknowledgments cells [14]. Correspondingly, germ cell tumor markers, such as AFP, MMP7, and NR6A1 [37], could be upregulated by This work was supported by the National Research NANOG overexpression due to romidepsin treatment in Foundation of Korea (NRF) grant funded by the Korea chicken PGCs. Also, we found that the migration-related government (MSIP) (no. NRF-2015R1A3A2033826). This gene in chicken, CXCR4 [38], was downregulated, indicat- research was also supported by the International Research & ing the possibility of aberrant migration of those treated Development Program of the National Research Foundation PGCs. The misallocated cells that could not undergo apo- of Korea (NRF) funded by the Ministry of Science, ICT and ptosis in extragonadal sites could give rise to germ cell tu- Future Planning of Korea (NRF-2016K1A3A1A21005676). 1224 JUNG ET AL.

Author Disclosure Statement 17. Zhang M, HG Leitch, WWC Tang, N Festuccia, E Hall- Ponsele, J Nichols, MA Surani, A Smith and I Chambers. No competing financial interests exist. (2018). Esrrb complementation rescues development of Nanog-null germ cells. Cell Rep 22:332–339. References 18. Lavial F, H Acloque, F Bertocchini, DJ Macleod, S Boast, E Bachelard, G Montillet, S Thenot, HM Sang, et al. 1. Tang WW, T Kobayashi, N Irie, S Dietmann and MA (2007). The Oct4 homologue PouV and Nanog regulate Surani. (2016). Specification and epigenetic programming pluripotency in chicken embryonic stem cells. Develop- of the human germ line. Nat Rev Genet 17:585–600. ment 134:3549–3563. 2. Strome S and D Updike. (2015). Specifying and protecting 19. Han JY, HG Lee, YH Park, YS Hwang, SK Kim, D Re- germ cell fate. Nat Rev Mol Cell Biol 16:406–416. ngaraj, BW Cho and JM Lim. (2018). Acquisition of 3. Seki Y, K Hayashi, K Itoh, M Mizugaki, M Saitou and Y pluripotency in the chick embryo occurs during intrauterine Matsui. (2005). Extensive and orderly reprogramming of embryonic development via a unique transcriptional net- genome-wide chromatin modifications associated with spec- work. J Anim Sci Biotechnol 9:31. ification and early development of germ cells in mice. Dev 20. Nakanoh S, N Fuse, R Tadokoro, Y Takahashi and K Biol 278:440–458. Agata. (2017). Jak1/Stat3 signaling acts as a positive reg- 4. Akiyama T, M Nagata and F Aoki. (2006). Inadequate ulator of pluripotency in chicken pre-gastrula embryos. Dev histone deacetylation during oocyte meiosis causes aneu- Biol 421:43–51. ploidy and embryo death in mice. Proc Natl Acad Sci U S A 21. Can˜o´n S, C Herranz and M Manzanares. (2006). Germ cell 103:7339–7344. restricted expression of chick Nanog. Dev Dyn 235:2889– 5. Steilmann C, A Paradowska, M Bartkuhn, M Vieweg, HC 2894. Schuppe, M Bergmann, S Kliesch, W Weidner and K Steger. 22. Murakami K, U Gu¨nesdogan, JJ Zylicz, WWC Tang, R (2011). Presence of histone H3 acetylated at lysine 9 in male Sengupta, T Kobayashi, S Kim, R Butler, S Dietmann and germ cells and its distribution pattern in the genome of hu- MA Surani. (2016). NANOG alone induces germ cells in man spermatozoa. Reprod Fertil Dev 23:997–1011. primed epiblast in vitro by activation of enhancers. Nature 6. Han JY. (2009). Germ cells and transgenesis in chickens. 529:403–407. Comp Immunol Microbiol Infect Dis 32:61–80. 23. Yang P, Y Wang, J Chen, H Li, L Kang, Y Zhang, S Chen, 7. Tsunekawa N, M Naito, Y Sakai, T Nishida and T Noce. B Zhu and S Gao. (2011). RCOR2 is a subunit of the LSD1 (2000). Isolation of chicken vasa homolog gene and tracing complex that regulates ESC property and substitutes for the origin of primordial germ cells. Development 127: SOX2 in reprogramming somatic cells to pluripotency. 2741–2750. Stem Cells 29:791–801. 8. Lee HC, HJ Choi, HG Lee, JM Lim, T Ono and JY Han. 24. dos Santos RL, L Tosti, A Radzisheuskaya, IM Caballero, (2016). DAZL expression explains origin and central for- K Kaji, B Hendrich and JC Silva. (2014). MBD3/NuRD mation of primordial germ cells in chickens. Stem Cells facilitates induction of pluripotency in a context-dependent Dev 25:68–79. manner. Cell Stem Cell 15:102–110. 9. Hamburger V and HL Hamilton. (1951). A series of normal 25. Saunders A, X Huang, M Fidalgo, MH Reimer, Jr., F stages in the development of the chick embryo. J Morphol Faiola, J Ding, C Sa´nchez-Priego, D Guallar, C Sa´enz, D Li 88:49–92. and J Wang. (2017). The SIN3A/HDAC corepressor com- 10. Kress C, G Montillet, C Jean, A Fuet and B Pain. (2016). plex functionally cooperates with NANOG to promote Chicken embryonic stem cells and primordial germ cells pluripotency. Cell Rep 18:1713–1726. display different heterochromatic histone marks than their 26. Park TS and JY Han. (2012). piggyBac transposition into mammalian counterparts. Epigenetics Chromatin 9:5. primordial germ cells is an efficient tool for transgenesis in 11. Yu J, MA Vodyanik, K Smuga-Otto, J Antosiewicz- chickens. Proc Natl Acad Sci U S A 109:9337–9341. Bourget, JL Frane, S Tian, J Nie, GA Jonsdottir, V Ruotti, 27. Park TS, HJ Lee, KH Kim, JS Kim and JY Han. (2014). et al. (2007). Induced pluripotent stem cell lines derived Targeted gene knockout in chickens mediated by TALENs. from human somatic cells. Science 318:1917–1920. Proc Natl Acad Sci U S A 111:12716–12721. 12. Yamaguchi S, H Kimura, M Tada, N Nakatsuji and T Tada. 28. Ko MH, YS Hwang, JS Rim, HJ Han and JY Han. (2017). (2005). Nanog expression in mouse germ cell development. Avian blastoderm dormancy arrests cells in G2 and sup- Downloaded by Seoul National Univ. Medical College from www.liebertpub.com at 11/01/18. For personal use only. Gene Expr Patterns 5:639–646. presses apoptosis. FASEB J 31:3240–3250. 13. Irie N, WW Tang and M Azim Surani. (2014). Germ cell 29. Kazantsev AG and LM Thompson. (2008). Therapeutic ap- specification and pluripotency in mammals: a perspective plication of histone deacetylase inhibitors for central nervous from early embryogenesis. Reprod Med Biol 13:203–215. system disorders. Nat Rev Drug Discov 7:854–868. 14. Yamaguchi S, K Kurimoto, Y Yabuta, H Sasaki, N Nakatsuji, 30. Kwon HJ, MS Kim, MJ Kim, H Nakajima and KW Kim. M Saitou and T Tada. (2009). Conditional knockdown of (2002). Histone deacetylase inhibitor FK228 inhibits tumor Nanog induces apoptotic cell death in mouse migrating pri- angiogenesis. Int J Cancer 97:290–296. mordial germ cells. Development 136:4011–4020. 31. Mitsui K, Y Tokuzawa, H Itoh, K Segawa, M Murakami, K 15. Chambers I, J Silva, D Colby, J Nichols, B Nijmeijer, M Takahashi, M Maruyama, M Maeda and S Yamanaka. Robertson, J Vrana, K Jones, L Grotewold and A Smith. (2003). The homeoprotein Nanog is required for mainte- (2007). Nanog safeguards pluripotency and mediates germ- nance of pluripotency in mouse epiblast and ES cells. Cell line development. Nature 450:1230–1234. 113:631–642. 16. Western PS, JA van den Bergen, DC Miles and AH Sin- 32. Loh YH, Q Wu, JL Chew, VB Vega, W Zhang, X Chen, G clair. (2010). Male fetal germ cell differentiation involves Bourque, J George, B Leong, et al. (2006). The Oct4 and complex repression of the regulatory network controlling Nanog transcription network regulates pluripotency in pluripotency. FASEB J 24:3026–3035. mouse embryonic stem cells. Nat Genet 38:431–440. THE MODERATE EXPRESSION OF NANOG IN CHICKEN PGCS 1225

33. Ware CB, L Wang, BH Mecham, L Shen, AM Nelson, M 41. Chen ZF, AJ Paquette and DJ Anderson. (1998). NRSF/ Bar, DA Lamba, DS Dauphin, B Buckingham, et al. (2009). REST is required in vivo for repression of multiple neuronal Histone deacetylase inhibition elicits an evolutionarily target genes during embryogenesis. Nat Genet 20:136–142. conserved self-renewal program in embryonic stem cells. 42. Grimes JA, SJ Nielsen, E Battaglioli, EA Miska, JC Speh, Cell Stem Cell 4:359–369. DL Berry, F Atouf, BC Holdener, G Mandel and T Kou- 34. Hezroni H, BS Sailaja and E Meshorer. (2011). Pluripotency- zarides. (2000). The co-repressor mSin3A is a functional related, valproic acid (VPA)-induced genome-wide histone component of the REST-CoREST repressor complex. J H3 lysine 9 (H3K9) acetylation patterns in embryonic stem Biol Chem 275:9461–9467. cells. J Biol Chem 286:35977–35988. 43. Johnson R, CH Teh, G Kunarso, KY Wong, G Srinivasan, 35. Hart AH, L Hartley, K Parker, M Ibrahim, LH Looijenga, M ML Cooper, M Volta, SS Chan, L Lipovich, et al. (2008). Pauchnik, CW Chow and L Robb. (2005). The pluripotency REST regulates distinct transcriptional networks in em- homeobox gene NANOG is expressed in human germ cell bryonic and neural stem cells. PLoS Biol 6:e256. tumors. Cancer 104:2092–2098. 44. Yu HB, R Johnson, G Kunarso and LW Stanton. (2011). 36. Heaney JD, EL Anderson, MV Michelson, JL Zechel, PA Coassembly of REST and its cofactors at sites of gene Conrad, DC Page and JH Nadeau. (2012). Germ cell repression in embryonic stem cells. Genome Res 21: pluripotency, premature differentiation and susceptibility to 1284–1293. testicular teratomas in mice. Development 139:1577–1586. 45. Lee M, H Ji, Y Furuta, JI Park and PD McCrea. (2014). 37. Juric D, S Sale, RA Hromas, R Yu, Y Wang, GE Duran, R p120-catenin regulates REST and CoREST, and modulates Tibshirani, LH Einhorn and BI Sikic. (2005). Gene ex- mouse embryonic stem cell differentiation. J Cell Sci 127: pression profiling differentiates germ cell tumors from 4037–4051. other cancers and defines subtype-specific signatures. Proc Natl Acad Sci U S A 102:17763–17768. Address correspondence to: 38. Stebler J, D Spieler, K Slanchev, KA Molyneaux, U Dr. Jae Yong Han Richter, V Cojocaru, V Tarabykin, C Wylie, M Kessel and Department of Agricultural Biotechnology E Raz. (2004). Primordial germ cell migration in the chick College of Agriculture and Life Sciences and mouse embryo: the role of the chemokine SDF-1/ Seoul National University CXCL12. Dev Biol 272:351–361. 1 Gwanak-ro 39. Schneider DT, AE Schuster, MK Fritsch, J Hu, T Olson, S Gwanak-gu Lauer, U Go¨bel and EJ Perlman. (2001). Multipoint im- Seoul 08826 printing analysis indicates a common precursor cell for Korea gonadal and nongonadal pediatric germ cell tumors. Cancer Res 61:7268–7276. E-mail: [email protected] 40. Chong JA, J Tapia-Ramı´rez, S Kim, JJ Toledo-Aral, Y Zheng, MC Boutros, YM Altshuller, MA Frohman, SD Kraner and G Mandel. (1995). REST: a mammalian si- Received for publication March 16, 2018 lencer protein that restricts sodium channel gene expression Accepted after revision June 26, 2018 to neurons. Cell 80:949–957. Prepublished on Liebert Instant Online July 21, 2018 Downloaded by Seoul National Univ. Medical College from www.liebertpub.com at 11/01/18. For personal use only.