The pluripotency factor Nanog regulates pericentromeric ...genesdev.cshlp.org/content/30/9/1101.full.pdf · The pluripotency factor Nanog regulates pericentromeric heterochromatin

The pluripotency factor Nanog regulatespericentromeric heterochromatinorganization in mouse embryonic stemcellsClara Lopes Novo,1 Calvin Tang,2,3 Kashif Ahmed,2 Ugljesa Djuric,4,5 Eden Fussner,2,3

Nicholas P. Mullin,6 Natasha P. Morgan,1 Jasvinder Hayre,1 Arnold R. Sienerth,1

Sarah Elderkin,7 Ryuichi Nishinakamura,8 Ian Chambers,6 James Ellis,4,5 David P. Bazett-Jones,2,3

and Peter J. Rugg-Gunn1,9,10

1Epigenetics Programme, The Babraham Institute, Cambridge CB22 3AT, United Kingdom; 2Program in Genetics and GenomeBiology, Hospital for Sick Children, Toronto, Ontario MSG 1L7, Canada; 3Department of Biochemistry, University of Toronto,Toronto, Ontario M5S 1A8, Canada; 4Program in Developmental and Stem Cell Biology, Hospital for Sick Children, Toronto,OntarioM5G1L7, Canada; 5Department ofMolecularGenetics, University of Toronto, Toronto,OntarioM5S 1A8, Canada; 6MRCCentre for Regenerative Medicine, Institute for Stem Cell Research, School of Biological Sciences, University of Edinburgh,Edinburgh EH16 4UU, United Kingdom; 7Nuclear Dynamics Programme, The Babraham Institute, Cambridge, CB22 3AT, UnitedKingdom; 8Department of Kidney Development, Institute of Molecular Embryology and Genetics, Kumamoto University,Kumamoto 860-0811, Japan; 9Centre for Trophoblast Research, University of Cambridge, Cambridge CB2 3EG, United Kingdom;10Wellcome Trust-Medical Research Council Cambridge Stem Cell Institute, University of Cambridge, Cambridge CB2 1QR,United Kingdom

An open and decondensed chromatin organization is a defining property of pluripotency. Several epigenetic regu-lators have been implicated in maintaining an open chromatin organization, but how these processes are connectedto the pluripotency network is unknown. Here, we identified a new role for the transcription factor NANOGas a keyregulator connecting the pluripotency network with constitutive heterochromatin organization in mouse embry-onic stem cells. Deletion ofNanog leads to chromatin compaction and the remodeling of heterochromatin domains.Forced expression of NANOG in epiblast stem cells is sufficient to decompact chromatin. NANOG associates withsatellite repeats within heterochromatin domains, contributing to an architecture characterized by highly dispersedchromatin fibers, low levels of H3K9me3, and high major satellite transcription, and the strong transactivationdomain of NANOG is required for this organization. The heterochromatin-associated protein SALL1 is a directcofactor for NANOG, and loss of Sall1 recapitulates the Nanog-null phenotype, but the loss of Sall1 can be cir-cumvented through direct recruitment of the NANOG transactivation domain to major satellites. These resultsestablish a direct connection between the pluripotency network and chromatin organization and emphasize thatmaintaining an open heterochromatin architecture is a highly regulated process in embryonic stem cells.

[Keywords: embryonic stem cells; pluripotency; heterochromatin; nuclear organization]

Supplemental material is available for this article.

Received November 29, 2015; revised version accepted March 23, 2016.

The genome of eukaryotic cells is organized into euchro-matin, which is generally permissive for gene transcrip-tion and activation, and heterochromatin, which islargely gene-poor. This form of nuclear compartmentali-zation is thought to impact genome regulation and stabil-ity, thereby contributing to cell identity (Fraser andBickmore 2007; Misteli 2007; Bickmore and van Steensel

2013). Pluripotentmouse embryonic stem cell (ESC) chro-matin exists in an unusual configuration with widely dis-persed open chromatin throughout the nucleoplasm,including within constitutive heterochromatin domainssuch as pericentromeric satellite repeats (Meshorer et al.2006; Efroni et al. 2008; Fussner et al. 2011; de Wit et al.2013). A similar form of highly dispersed chromatin archi-tecture also characterizes pluripotent epiblast cellswithin

Corresponding author: [email protected] published online ahead of print. Article and publication dateare online at http://www.genesdev.org/cgi/doi/10.1101/gad.275685.115.Freely available online through the Genes & Development Open Accessoption.

© 2016 Novo et al. This article, published in Genes & Development, isavailable under a Creative Commons License (Attribution 4.0 Internation-al), as described at http://creativecommons.org/licenses/by/4.0/.

GENES & DEVELOPMENT 30:1101–1115 Published by Cold Spring Harbor Laboratory Press; ISSN 0890-9369/16; www.genesdev.org 1101

Cold Spring Harbor Laboratory Press on November 2, 2020 - Published by genesdev.cshlp.orgDownloaded from

mailto:[email protected]




http://www.genesdev.org/cgi/doi/10.1101/gad.275685.115

http://www.genesdev.org/cgi/doi/10.1101/gad.275685.115

http://genesdev.cshlp.org/site/misc/terms.xhtml

http://creativecommons.org/licenses/by/4.0/


http://genesdev.cshlp.org/site/misc/terms.xhtml

http://genesdev.cshlp.org/

http://www.cshlpress.com

the mouse blastocyst (Ahmed et al. 2010; Boskovic et al.2014). Upon cell differentiation, there is extensive nuclearreorganization that is associated with chromatin compac-tion and the formation of condensed heterochromatin do-mains that form a repressive environment (Meshorer et al.2006; Efroni et al. 2008; Wen et al. 2009; Ahmed et al.2010; Wijchers et al. 2015). Therefore, remodeling of het-erochromatin architecture during stem cell and develop-mental fate transitions can provide an important modelfor investigating chromatin domain organization.

An open chromatin structure may contribute to cellpluripotency, potentially by creating a transcriptionallypermissive and accessible genome (Gaspar-Maia et al.2011; Cavalli and Misteli 2013). Reducing the expressionof several epigenetic regulators (such as Chd1, membersof the esBAF complex, and Padi4) in ESCs results inthe accumulation of compact heterochromatin domains,disrupted self-renewal, and altered ESC differentiationpotential (Meshorer et al. 2006; Gaspar-Maia et al.2009; Lessard and Crabtree 2010; Christophorou et al.2014). Furthermore, forced heterochromatin decompac-tion using DNA methyltransferase and histone deacety-lase inhibitors or genetic depletion of histone H3 Lys9methyltransferases increases the efficiency with whichsomatic cells can be reprogrammed to a pluripotent state(Huangfu et al. 2008; Mikkelsen et al. 2008; Soufi et al.2012; Sridharan et al. 2013). These findings have led tothe conclusion that heterochromatin regions act as imped-iments to the reprogramming processes and may restrictthe establishment and/or maintenance of pluripotency.

In addition to influencing genome plasticity, hetero-chromatin organization could also have unexplored andimportant functions in regulating other aspects of genomefunction and stability in pluripotent cells. The chromatinenvironment of constitutive pericentromeric hetero-chromatin (PCH) has been well characterized in somaticcells and shown to contain condensed chromatin fibersand high levels of histone H3 Lys9 trimethylation(H3K9me3) that is mediated by Suv39h1/2 methyltrans-ferases (Peters et al. 2001; Lehnertz et al. 2003). Themajorsatellite DNA repeats within PCH are typically transcrip-tionally repressed yet remain accessible to DNA-bindingfactors and are responsive to transcriptional regulation(Bulut-Karslioglu et al. 2012). Deletion of epigenetic regu-lators (including Suv39h1/2 and Dicer) in mouse somaticcells perturbs PCH identity, causes the transcriptional up-regulation of major satellite sequences, and is associatedwith severe chromosome missegregation phenotypes (Pe-ters et al. 2001; Kanellopoulou et al. 2005). Interestingly,the chromatin environment of PCH in ESCs appears tobe distinct, with open and decondensed chromatin fibersand lower levels of H3K9me3 compared with somaticcells (Meshorer et al. 2006; Efroni et al. 2008; Fussneret al. 2011). The key drivers of this unusual architectureremain largely unknown, in part because the repetitivenature of heterochromatin sequences makes them chal-lenging to study. Importantly, deletion of Suv39h1/2and Dicer in ESCs can lead to increased major satellitetranscription, as in somatic cells; however, the down-stream response is different because the transcriptional

up-regulation does not cause chromosomemissegregationin ESCs (Peters et al. 2001; Kanellopoulou et al. 2005).These findings raise the possibility that ESCs can tolerateor perhaps even require a unique PCH identity and suggestthe existence of key functional differences in heterochro-matin regulation between pluripotent and somatic cells.

In order to better understand how an open PCH organi-zation is established and maintained in pluripotent cells,it is essential to dissect the functional links between plu-ripotency networks and nuclear architecture. One keymember of the stem cell pluripotency network is the tran-scription factorNanog (Chambers et al. 2003;Mitsui et al.2003). Despite the central position of Nanog within thenetwork, Nanog–/– ESCs and Nanog–/–-induced pluripo-tent stem cells are able to undergo self-renewal and arepluripotent, suggesting that Nanog may have additionalroles in pluripotent cells outside of controlling the tran-scriptional network (Chambers et al. 2007; Carter et al.2014; Schwarz et al. 2014). We reasoned that Nanog is apotential candidate for regulating PCH organization inESCs because it is expressed in cells that are associatedwith an open PCHarchitecture, such as early embryo cellsand germ cells (Chambers et al. 2003; Mitsui et al. 2003;Hart et al. 2004), and we and others have shown previous-ly that Nanog levels inversely correlate with several indi-cators of heterochromatin compaction in ESCs andembryos (Ahmed et al. 2010; Fussner et al. 2011; Mattoutet al. 2011). Here, we show that Nanog is necessary andsufficient for PCH organization in ESCs. Deletion ofNanog leads to compaction and reorganization of consti-tutive heterochromatin domains, and forced expressionof NANOG in epiblast stem cells (EpiSCs) is sufficientto decondense PCH organization and redistribute consti-tutive heterochromatin domains. We found that NANOGassociates with satellite repeats within PCH domains,contributing to an overall heterochromatin architecturein ESCs that is characterized by highly dispersed chro-matin fibers, low levels of H3K9me3, and high major sat-ellite transcription. Importantly, tethering the NANOGtransactivator domain directly to major satellite DNA issufficient to remodel PCH organization, thereby defininga direct and active role forNanog in regulating heterochro-matin. Through a proteomic approach, we identified thezinc finger-containing transcription factor SALL1 as a di-rect NANOG-interacting protein during heterochromatinremodeling. SALL1 has a prominent heterochromatinlocalization in ESCs (Sakaki-Yumoto et al. 2006), andSALL1–NANOG interactions have been detected inESCs previously (Karantzali et al. 2011); however, a func-tional role for Sall1 in ESC heterochromatin regulationhas not been reported. Here, we show that Sall1, likeNanog, is necessary tomaintain an open heterochromatinorganization in ESCs and is required for NANOG to as-sociate with PCH in order to mediate heterochromatinremodeling. Together, these results establish the first di-rect molecular connection between a key member of thepluripotency network and higher-order chromatin organi-zation in pluripotent cells and lead to the conclusion thatmaintaining an open and dispersed PCH architecture is ahighly regulated and integrated process in ESCs.

Novo et al.

1102 GENES & DEVELOPMENT




Results

Nanog is necessary for an open heterochromatinorganization in ESCs

To test whether Nanog has a direct role in the mainte-nance of decondensed constitutive heterochromatin do-mains, we compared chromatin organization betweenwild-type ESCs and Nanog–/– ESCs (Chambers et al.2007). Electron spectroscopic imaging (ESI), a direct andquantitative technique to examine nuclear ultrastructure,confirmed that chromatin in wild-type ESCs is largelydecondensed and homogenous throughout the nucleo-plasm (Fig. 1A; Efroni et al. 2008). In contrast, chromatinin Nanog–/– ESCs was less uniformly distributed, tendingto compact at the nuclear envelope and nucleolar periph-

ery into distinct heterochromatin domains (Fig. 1A).These datawere supported by the increased density of het-erochromatin fibers in Nanog–/– ESCs compared withwild-type ESCs (Fig. 1B). We extended these experimentsto several transgenic ESC lines representing a Nanog ex-pression gradient (Chambers et al. 2007) and found astrong correlation between Nanog levels and heterochro-matin dispersion (Fig. 1A,B).Immunofluorescent microscopy of the heterochro-

matin marker H3K9me3 revealed major changes in PCHorganization in Nanog–/– ESCs. In contrast to wild-typeESCs, H3K9me3-positive chromocenters were detectedas small, discrete foci in Nanog–/– ESCs, and the mediannumber and total area of H3K9me3-labeled foci per nucle-us was significantly higher (Fig. 1C). Nuclear area was

Figure 1. Nanog is required for open het-erochromatin organization in ESCs. (A)ESI analysis of wild-type (WT), Nanog–/–,Nanog+/–, and Nanog-overexpressing ESCs.Quantitative phosphorus and nitrogen ratioimages were segmented to show chromatinin yellow and protein-based structures inblue. The nuclear membrane is indicatedwith arrowheads. The regions imaged con-tain H3K9me3-positive PCH as determinedby correlative immunofluorescent micros-copy. Bar, 0.5 μm. (B) Box and whisker plotsshow the distribution of heterochromatinfiber density as revealed by phosphorus im-ages. Data were compared using a one-wayANOVA followed by Bonferroni’s multiplecomparison test. (C ) Chromocenter organi-zation revealed by immunofluorescentanalysis of H3K9me3 in ESCs expressingdifferent levels of Nanog. Note thatH3K9me3 foci are formed from PCH anddo not overlap with other heterochro-matin compartments, including telomeres.OCT4 labeling confirms the undifferentiat-ed status of the cell type. Bar, 2 µm. Box andwhisker plots show the number (top left),size (top right), and total area (bottom left)of H3K9me3 foci per nucleus. (Bottomright) Nuclear area was unchanged. Datawere compared using a one-way ANOVAfollowed by Bonferroni’s multiple compari-son test. Data were collected from at leasttwo independent experiments.

NANOG regulates heterochromatin in ESCs

GENES & DEVELOPMENT 1103




unchanged (Fig. 1C). Compared with wild-type ESCs,chromocenter number was also significantly higher inNanog+/– ESCs and significantly lower in Nanog-over-expressing ESCs, further reinforcing a correlation betweenNanog levels and heterochromatin organization (Fig. 1C).DAPI line scan analyses demonstrated that NANOG–/–

ESCs chromocenters appear as distinct, bright foci andare well compartmentalized, while those of wild-typeESCs are more disrupted and dispersed with lower DAPIsignal relative to nucleoplasmic background (Supple-mental Fig. 1A). Differences in heterochromatin organiza-tion were confirmed using alternative wild-type andNanog–/– ESC lines (Supplemental Fig. 1B; Chamberset al. 2007). We also assessed whether chromocenter or-ganization is correlated with the variegated NANOGexpression that is typically observed within a colony ofwild-type ESCs (Chambers et al. 2007). In agreement withour previous findings (Fussner et al. 2011), high NANOG-expressing cells exhibited larger, fewer, and more dis-rupted chromocenters compared with low NANOG-expressing cells (Supplemental Fig. 1C). Finally, directvisualization of PCH distribution by major satellite DNAfluorescence in situ hybridization (FISH) also revealeddifferences in organization between wild-type ESCsand Nanog–/– ESCs (Supplemental Fig. 1D). Importantly,the altered heterochromatin organization observed inNanog–/– ESCs could be rescued by restoringNANOG lev-els with a transgene (Fig. 1B,C; Supplemental Fig. S1D–F).

The increased chromatin compaction and redistribu-tion of heterochromatin domains in Nanog–/– ESCs aresimilar to changes that occur upon ESC differentiation(Meshorer et al. 2006), raising the possibility that the chro-matin phenotype may be caused indirectly by changes incell state. Transcriptional and functional analyses, how-ever, showed thatNanog–/– ESCs retain the defining prop-erties of wild-type ESCs. The presence of Klf4,Nr0b1, andZfp42 transcripts and the low level of early differentiationmarkers such as T, Lefty1, and Eomes indicate thatNanog–/– ESCs have not initiated differentiation (Supple-mental Fig. 2A), andNanog–/– ESCs express ESC-associat-ed transcripts (ECATs) (Mitsui et al. 2003; Chambers et al.2007) and known H3K9me3 methyltransferases and his-tone demethylases at levels similar to wild-type ESCs(Supplemental Fig. 2B). Nanog–/– ESCs are also alkalinephosphatase-positive in a LIF-dependent manner (Cham-bers et al. 2007) and reveal a similar distribution ofOCT4 and SOX2 protein levels within the cell populationcompared with wild-type ESCs (Supplemental Fig. 2C).Importantly, differences in chromocenter organization be-tween wild-type and Nanog–/– ESCs were retained whenthe analysis was restricted to KLF4-positive cells, whichis a sensitive indicator of naïve pluripotency (Supplemen-tal Fig. 2D; Guo et al. 2009), and also when cultured inmore stringent 2i/LIF conditions that hold ESCs in a naïvestate (Supplemental Fig. 2E; Ying et al. 2008). Together,these data show that heterochromatin compaction andredistribution occur in Nanog–/– ESCs independently ofsubstantial changes in cell state, thereby identifying anessential role for Nanog in maintaining an open hetero-chromatin organization in ESCs.

Down-regulation of Nanog during ESC differentiation isrequired for heterochromatin remodeling

Nanog is rapidly down-regulated upon ESC differentiation(Chambers et al. 2003), potentially providing a cue to con-dense and remodel heterochromatin architecture. To in-vestigate whether loss of Nanog expression could beresponsible for driving chromatin reorganization, we ex-amined the timing of heterochromatin remodeling thatoccurs upon ESC differentiation. ESCs were treated withretinoic acid for 5 d, and chromocenter organization wasexamined every 24 h (Fig. 2A). Consistent with previousstudies (Meshorer et al. 2006), PCH foci, as revealed byH3K9me3 immunofluorescent signals, became more nu-merous, smaller, and more intense upon ESC differentia-tion (Fig. 2B,C). Importantly, a major change in theseparameters occurs within the first 48 h of retinoic acid in-duction, coinciding with loss of pluripotency factors, in-cluding NANOG (Fig. 2A,C). Therefore, the timing ofheterochromatin remodeling upon the early stages ofESC differentiation is consistent with a role for Nanogin orchestrating these nuclear organization events.

We next assessed the impact of altering NANOG levelson chromocenter remodeling during ESC differentiation.At day 0, Nanog–/– ESCs already displayed well-definedand discrete chromocenters, and this distribution didnot significantly change over the first 3 d of ESC differen-tiation, suggesting that there is little PCH remodeling dur-ing this period in the absence of NANOG (Fig. 2B,C).Interestingly, a subsequent phase of chromocenter remod-eling occurred after day 3, pointing to the existence of alater stage NANOG-independent process. Conversely,continuous ectopic expression of NANOG in wild-typeESCs prevented the typical remodeling in chromocenterorganization, instead maintaining the highly disruptedPCH organization that is characteristic of undifferentiat-ed ESCs (Fig. 2B,C). This phenotype could be direct or in-direct because it coincided with the failure to down-regulate pluripotency factors such as OCT4, a findingthat is consistent with our previously published differen-tiation experiments (Chambers et al. 2003). Overall, theseresults identify a dependency between NANOG levelsand PCH remodeling during ESC differentiation.

Nanog is sufficient to remodel heterochromatin state

To further explore the impact of early differentiationevents on heterochromatin organization, we examinedchromatin organization in EpiSCs. EpiSCs are capable ofdifferentiating into all three germ layers and express sever-al pluripotency factors, such as OCT4, at levels similar toESCs but importantly express NANOG at lower levelscompared with ESCs (Fig. 3A; Supplemental Fig. 2A;Brons et al. 2007; Tesar et al. 2007; Osorno and Chambers2011; Osorno et al. 2012). We hypothesized that if Nanogwas instructive in maintaining a decondensed hetero-chromatin organization, then EpiSCs may reveal a morecompacted chromatin architecture. Indeed, ultrastructur-al analysis using ESI showed that chromatin in EpiSCswas organized into distinct compacted chromatin

Novo et al.



http://genesdev.cshlp.org/lookup/suppl/doi:10.1101/gad.275685.115/-/DC1
















domains and generally was less uniformly distributedcompared with ESCs (Fig. 3B). Differences in chromatinorganization between EpiSCs and ESCs was confirmedby immunofluorescent microscopy of heterochromatinfoci identified by H3K9me3 and by DNA FISH for majorsatellites, revealing that chromocenters are organizedinto small discrete foci in EpiSCs (Fig. 3C; SupplementalFig. 3A). Together, these data reveal that chromatin inEpiSCs is organized similarly to Nanog–/– ESCs, therebyreinforcing a functional link between Nanog and hetero-chromatin organization.As forcedNANOGexpression could restore typical ESC

chromatin architecture in Nanog–/– ESCs, we reasonedthat elevated expression of NANOG could also be suffi-cient to remodel heterochromatin organization in EpiSCs.Prolonged NANOG expression has been shown previous-ly to enable EpiSC reprogramming (Silva et al. 2009), po-tentially confounding analysis of chromatin remodeling.We therefore designed experiments to investigate theeffects of short-term NANOG induction in EpiSCs. Weengineered EpiSC lines that expressed Nanog upon doxy-cycline (DOX) treatment, thereby allowing precise controlof the timing of NANOG induction (Nanog-EpiSCs) (Fig.3A; Supplemental Fig. 3B). Remarkably, direct analysisof chromatin organization using ESI and major satelliteDNA FISH in addition to indirect indicators, includingH3K9me3 immunofluorescent signals and DAPI line

scan analyses, revealed that heterochromatin was remod-eled and dispersed within 24 h of NANOG induction (Fig.3B,C; Supplemental Fig. 3C,D). At this time point, themajority of chromatinwas uniformly distributed through-out the nucleoplasm, with an overall chromatin architec-ture indistinguishable from ESCs. As expected, controlcells, including noninduced Nanog-EpiSCs and inducedGFP-EpiSCs, revealed chromatin organization typical ofEpiSCs, with domains of compacted chromatin and thepresence of small heterochromatin foci (Fig. 3B,C; Sup-plemental Fig. 3C,D). Induced expression of a NANOGhomeodomain point mutant that has substantially re-duced DNA-binding affinity (N51E) (Jauch et al. 2008)was unable to remodel heterochromatin architecture,suggesting that a functional homeodomain is required(Supplemental Fig. 3E,F). We confirmed that Nanog ex-pression driven from a constitutive promoter in EpiSCswas also sufficient to remodel heterochromatin (datanot shown). Importantly, short-term forced expression ofalternative pluripotency factors, including Klf2, Klf4,Esrrb, Nr0b1, Prdm14, Dppa3, and Nr5a2, did not causedetectable changes in heterochromatin organization, un-derscoring the specific role forNanog in remodeling chro-matin organization (Fig. 3D; Supplemental Fig. 4A,B). Inaddition, NANOGwas unable to access and remodel het-erochromatin when overexpressed in fibroblasts, indicat-ing that the function may be restricted to early embryo

Figure 2. The timing of chromatin remod-eling upon ESC differentiation is consistentwith a role forNanog in orchestrating thesenuclear organization events. (A) Westernblot of Nanog–/–, wild-type (WT), andNanog-overexpressing ESCs over 5 d ofdifferentiation with NANOG and OCT4antibodies. (B) Chromocenter organizationrevealed by immunofluorescent analysisof H3K9me3 during ESC differentiation.(Dashed line) Nuclear periphery. Bar, 2µm. (C ) Box and whisker plots show thenumber (left) and size (right) of H3K9me3foci per nucleus. Data were compared usinga one-wayANOVA followed by Bonferroni’smultiple comparison test. (n.s.) P > 0.1; (∗)P < 0.01. n > 50 per time point.














or stem cell types (Supplemental Fig. 4C,D; data notshown). Last, investigation of EpiSC status upon chro-matin remodeling revealed unchanged epigenetic and

transcriptomic profiles after 24 h of NANOG expres-sion (Supplemental Fig. 4E,F), thereby indicating thatNANOG-induced heterochromatin reorganization occurs

Figure 3. Nanog is sufficient to remodel heterochromatin in EpiSCs. (A) Western blot of NANOG and OCT4 in ESCs, EpiSCs, and dox-ycycline (DOX)-inducibleNanog-EpiSCs. DOXwas applied for 24 h. (B) ESI analysis of ESCs, EpiSCs, DOX-inducibleNanog-EpiSCs, andDOX-inducibleGFP-EpiSCs. DOX was applied for 24 h. The nuclear membrane is indicated by arrowheads. Bar, 0.5 μm. Box and whiskerplots reveal the distribution in size of the chromatin clusters. P-values were calculated using Student’s t-test. (C ) Chromocenter organi-zation revealed by immunofluorescent analysis of H3K9me3. OCT4 labeling confirmed the undifferentiated status of the cell type. DOXwas applied for 24 h. Bar, 2 µm. Box and whisker plots show the number (left) and size (right) of H3K9me3 foci per nucleus. P-values werecalculated using Student’s t-test. (n.s.) P > 0.1. Data were collected from at least two independent experiments. (D) Several pluripotencyfactors were overexpressed in EpiSCs for 24 h; only Nanog was able to remodel chromocenter organization. Box and whisker plots showthe number of H3K9me3 foci per nucleus. n = 50 per cell line (images are shown in Supplemental Fig. 4A.)

Novo et al.








independently of EpiSC-to-ESC reprogramming. Collec-tively, these results demonstrate that NANOG is suffi-cient to remodel heterochromatin in EpiSCs, resultingin an open chromatin architecture that is indistinguish-able from ESCs. Importantly, these remodeling eventscan precede changes in other epigenetic and transcription-al events, implying that heterochromatin organizationcan be decoupled from cell state.

Nanog-dependent pericentromeric satelliteorganization in ESCs

Our DNA FISH experiments revealed that pericentro-meric major satellite sequences that cluster within chro-mocenters undergo substantial remodeling in Nanog–/–

ESCs (Supplemental Fig. 1D). Given that transcription fac-tors can directly control major satellite DNA in othercell types (Bulut-Karslioglu et al. 2012), we hypothesizedthat NANOG could regulate the chromatin state of majorsatellite repeats in ESCs, thereby contributing to PCH or-ganization. To test this hypothesis, we performed electro-phoretic mobility shift assays to assess direct binding ofthe NANOG homeodomain to major satellite repeats.We found that the recombinant NANOG homeodomainwas able to substantially reduce the mobility of the full-length major satellite probe, and the shift was even morepronounced than that of a probe containing a well-charac-terized NANOG-binding site within the Tcf3 promoter(Fig. 4A; Jauch et al. 2008). A point mutation in the recog-nition helix of the NANOG homeodomain (N51A) (Jauchet al. 2008) abolishedDNA interactionwith themajor sat-ellite and Tcf3 probes (Fig. 4A), demonstrating that a func-tional homeodomain is required for major satellite DNAbinding. We next examined NANOG occupancy at PCHin ESCs using chromatin immunoprecipitation (ChIP).ChIP experiments revealed that NANOG bound to majorsatellite repeats in wild-type ESCs but not to other repeatclasses (Fig. 4B). The association of NANOG with majorsatellite repeats corresponded with several hallmarks ofan open PCH organization. First, we examined RNA out-put using RT-qPCR with a primer pair that amplifies oneunit of the 234-base-pair (bp) mouse major satellite repeat(Lehnertz et al. 2003). Major satellite transcripts were sig-nificantly decreased (approximately twofold) in Nanog–/–

ESCs compared with wild-type ESCs (Fig. 4C). Second,ChIP analyses revealed that H3K9me3 levels were approx-imately twofold increased and that H3K9 acetylation(H3K9ac) levels were approximately twofold decreasedat major satellite repeats in Nanog–/– ESCs comparedwith wild-type ESCs, and this difference was associatedwith increased occupancy of the H3K9 methyltransferaseSUV39H1 (Fig. 4D), together indicating an accumulationof heterochromatinization at major satellite repeats inNanog–/– ESCs. Other repeat sequences such as LINE,SINE, and IAPwereunaffected (Supplemental Fig. 5A). Im-munofluorescent microscopy of NANOG localization re-vealed a strong pan-nuclear signal that was not enrichedor depleted at chromocenters (Supplemental Fig. 5B).Last, separation of ESCs based on variegatedNanog levelsrevealed that high Nanog-expressing ESCs transcribed

higher levels of major satellite RNA compared with lowNanog-expressing ESCs, further reinforcing the connec-tion between NANOG protein levels and a more openPCH organization (Fig. 4E).Wenext investigatedNANOGbinding andRNAoutput

in EpiSCs upon chromatin remodeling. ChIP experimentsshowed increased NANOG occupancy at major satelliterepeats, but not at LINE and SINE sequences, uponNanoginduction in EpiSCs (Fig. 4F). The binding events correlat-ed with RNA output from major satellite repeats, whichsignificantly increased (approximately twofold) after 24 hofNanog induction in EpiSCs (Fig. 4G). At this time point,major satellite transcripts reached the same level as wild-type ESCs. Consistent with these changes, ChIP analysesrevealed that H3K9me3 levels at major satellite DNAsignificantly decreased and that H3K9ac levels increasedin Nanog-overexpressing EpiSCs compared with nonin-ducedEpiSCs (Fig. 4H). Importantly, expressionof aNanoghomeodomain point mutant that has substantially re-duced DNA-binding affinity (N51E) (Jauch et al. 2008)was unable to induce changes in major satellite transcrip-tion and H3K9me3 levels in EpiSCs, demonstrating that afunctional homeodomain is required (Supplemental Fig.5C). Last, as major satellite transcription could be cell cy-cle-regulated (Lu andGilbert 2007), we examinedwhetherchanges in cell cycle timing may contribute to the ob-served changes in RNA output from major satellite re-peats. Flow cytometry analysis revealed that cell cycleparameters were unaltered in Nanog–/– ESCs and alsoupon Nanog induction in EpiSCs (Supplemental Fig. 5D).Together, these results establish that NANOG binding isassociated with increased major satellite transcriptionand decreased heterochromatinization of major satelliterepeats, underlying the role of Nanog in maintaining anopen PCH organization in pluripotent cells.

The NANOG transactivation domain is criticalfor heterochromatin remodeling

We next examined the molecular basis for Nanog-depen-dent PHC organization. Transactivation activity ofNANOG can be mediated via the C-terminal WR andCD2 domains (Supplemental Fig. 6A; Pan and Pei 2003;Oh et al. 2005). We hypothesized that the recruitmentof the transactivation domains to major satellite repeatscould underlie the open PCH organization typical ofESCs. To test this hypothesis, we first expressed a Nanogtransgene that lacked theWRandCD2 transactivation do-mains (NanogΔC) in EpiSCs (Supplemental Fig. 6A–C). ESIanalysis and H3K9me3 immunofluorescent microscopyrevealed thatNanogΔCwas unable to decompact chroma-tin, remodel chromocenter organization, or up-regulatemajor satellite transcription (Fig. 5A,B; SupplementalFig. 6D). These findings demonstrate the requirement forthe transactivation domain in Nanog-mediated hetero-chromatin remodeling.To investigate whether recruitment of NANOG

transactivation domains to PCH was sufficient to initiatechromatin remodeling, we constructed a fusion pro-tein between the CD2 transactivation domain and a
















transcription activator-like effector (TALE) that is knownto specifically bind mouse major satellite DNA (TALE-CD2) (Miyanari et al. 2013). A luciferase-based reporterassay confirmed the activity of the fusion protein (Supple-mental Fig. 6E). As a control, we used a previously pub-lished and characterized TALE-mClover protein, whichbinds to major satellite DNA but does not alter the tran-scriptional or epigenetic properties of the target sequences(Supplemental Fig. 6E; Miyanari et al. 2013). We engi-

neered EpiSC lines with DOX-inducible TALE-CD2 orTALE-mClover transgenes and confirmed that additionof DOX to the culture medium caused up-regulation ofthe transgenes and localization of the fusion proteins toPCH (Fig. 5C). After 24 h of DOX induction, TALE-CD2,but not TALE-mClover, caused major satellite repeatsto adopt a more open and active state, as shown by a sig-nificant transcriptional up-regulation and correspondingchanges inH3K9me3 andH3K9ac levels (Fig. 5D,E). Other

Figure 4. NANOG associates with major satellite repeats in ESCs. (A) His-tagged recombinant wild-type and N51A mutant NANOGhomeodomains were used for electrophoretic mobility shift assays with a full-length major satellite probe (234 bp) (Bulut-Karsliogluet al. 2012) and a Tcf3 probe (14 bp) (Jauch et al. 2008). (B) ChIP-qPCR analysis of NANOG at major satellite, LINE, SINE, and IAPDNA in wild-type andNanog–/– ESCs. (C ) RT-qPCR for major satellite transcripts in wild-type andNanog–/– ESCs. Values were normal-ized toHmbs and are shown relative towild type. (D) ChIP-qPCR for H3K9me3, H3K9ac, and IgG (normalized to unmodifiedH3) (left) andSUV39H1 and IgG (normalized to IgG) (right) atmajor satellite DNA inwild-type andNanog–/– ESCs. (E) RT-qPCR formajor satellite tran-scripts in ESCs that were separated by flow cytometry forNanog low-expressing andNanog high-expressing cells using an ESC line witheGFP inserted into one Nanog allele (TNGA). (F ) ChIP-qPCR analysis of NANOG at major satellite DNA, LINE, and SINE in Nanog-EpiSCs with and without 24 h of DOX induction. (G) RT-qPCR for major satellite transcripts in Nanog-EpiSCs with and without 24 hof DOX induction. Values were normalized to Hmbs. (H) ChIP-qPCR for H3K9me3, H3K9ac, and IgG at major satellite DNA inNanog-EpiSCs with and without 24 h of DOX induction. Values were normalized to unmodified H3. All data represent mean ± SDfrom three biological experiments.

Novo et al.








repeat classes such as SINE and LINEwere unaffected. Im-portantly, induction of TALE-CD2 in EpiSCs was also suf-ficient to remodel chromocenter organization such that

chromocenters adopted a highly disrupted and dispersedpattern that is typical of undifferentiated ESCs (Fig. 5F).Of note is that induction of a substantially stronger

Figure 5. The NANOG transactivation domain is necessary and sufficient for heterochromatin remodeling. (A) ESI analysis of DOX-in-ducible NanogΔC-EpiSCs. DOX was applied for 24 h. The nuclear membrane is indicated by arrowheads. Bar, 0.5 μm. Box and whiskerplots reveal the distribution in size of chromatin clusters. P-value was calculated using Student’s t-test. (n.s.) P > 0.1. (B) Chromocenterorganization revealed by immunofluorescent analysis of H3K9me3. OCT4 labeling confirmed the undifferentiated status of the celltype. DOX was applied for 24 h. Bar, 2 μm. Box and whisker plots show the number (left) and size (right) of H3K9me3 foci per nucleus.P-values were calculated using Student’s t-test. (n.s.) P > 0.1. (C ) Diagram of transcription activator-like effector (TALE)-CD2 andTALE-mClover fusion proteins and a fluorescent microscopy image demonstrating localization of TALE-mClover to chromocenters after24 h of DOX induction in EpiSCs. (D) RT-qPCR for major satellite transcripts, LINE, and SINE in TALE-CD2-EpiSCs and TALE-mClover-EpiSCswith andwithout 24 h of DOX induction. Values were normalized toHmbs. (E) ChIP-qPCR for H3K9me3 atmajor satellite, LINE,and SINE DNA in TALE-CD2-EpiSCs and TALE-mClover-EpiSCs with and without 24 h of DOX induction. Values were normalized tounmodified H3. All data represent mean ± SD from at least three biological experiments. (F ) Chromocenter organization revealed by im-munofluorescent analysis of H3K9me3. Box and whisker plots show the number (left) and size (right) of H3K9me3 foci per nucleus.






transactivator (TALE-VP64) inEpiSCscaused a severe phe-notype with irregular nuclear morphology (data notshown), suggesting that transactivator strength is impor-tant. Together, these results show that recruitment of theNANOGtransactivatordomainspecifically tomajor satel-lite sequences is able to recapitulate the phenotype in-duced by overexpression of full-length Nanog, therebyidentifying a direct and active role forNanog in regulatingPCH organization.

Sall1 is required for Nanog-mediated remodeling

To uncover the molecular mechanisms through whichNANOG can associate with PCH in order to actively reg-ulate major satellite repeats, we identified proteins thatinteract with NANOG during PCH remodeling. We gen-erated EpiSCs containing a DOX-inducible 2xFlag-Nanogtransgene (Supplemental Fig. 6B,C). The tagged proteinwas functional, as it was able to remodel heterochromatinwhen overexpressed in EpiSCs, and was sufficient to en-able LIF-independent ESC proliferation (data not shown).We expressed the transgene for 24 h in EpiSCs, immuno-purified Flag-containing protein complexes, and identifiedassociated proteins by mass spectrometry. As a control,we examined the same EpiSC line without DOX induc-tion. Out of the proteins detected, we focused on thosepreviously shown to interact with NANOG in ESCs(Wang et al. 2006; Costa et al. 2013; Gagliardi et al.2013), as they are the most likely candidates for establish-ing and maintaining heterochromatin identity in pluripo-tent cells (Fig. 6A). In particular, the interaction partnerSALL1 was of interest because it has been shown previ-ously to bind heterochromatin domains in ESCs (Sakaki-Yumoto et al. 2006) and could therefore provide a linkbetween NANOG and recruitment to PCH. We used en-dogenous coimmunoprecipitation to confirm the asso-ciation in 2xFlag-Nanog-EpiSCs (data not shown) andwild-type ESCs (Fig. 6B) as well as the direct interactionof recombinant NANOG and SALL1 proteins (Fig. 6B).We also verified that SALL1 is present at similar levelsin ESCs and EpiSCs (Supplemental Fig. 7A).

Despite the prominent heterochromatin localization ofSALL1 in ESCs (Sakaki-Yumoto et al. 2006), a functionalrole for Sall1 in heterochromatin regulation has not beenreported. We therefore examined chromatin organizationin Sall1–/– ESCs (Yuri et al. 2009) in order to establishwhether Sall1, like Nanog, is required to maintain openheterochromatin. Ultrastructural analysis using ESIrevealed that chromatin in Sall1–/– ESCs was highly het-erogeneous, frequently forming regions of compact chro-matin at the nuclear envelope and nucleolar periphery(Fig. 6C). Chromatin cluster size and heterochromatin fi-ber densitywere significantly higher in Sall1–/–ESCs com-paredwithwild-typeESCs (Fig. 6C).Thealtered chromatinarchitecture in Sall1–/– ESCs was confirmed by H3K9me3and DAPI line scan immunofluorescence microscopy aswell as major satellite DNA FISH (Fig. 6D; SupplementalFig. 7B,C) and also when cultured in 2i/LIF conditions(Supplemental Fig. 7D). The alteration in chromatinorganization observed in Sall1–/– ESCs could be rescued

by restoring Sall1 levels with a transgene (Fig. 6D; Supple-mental Fig. 7C,E,F). Therefore, the inactivation of Sall1phenocopies the defects in PCH organization observed inNanog–/– ESCs. Importantly, NANOG levels are un-changed in Sall1–/– ESCs and therefore remain highly ex-pressed, and SALL1 levels are unchanged in Nanog–/–

ESCs and Nanog-overexpressing EpiSCs (SupplementalFigs. 1F, 6B, 7E). Furthermore, transcripts that charac-terize wild-type ESCs are unaltered in Sall1–/– ESCs,demonstrating that loss of Sall1 perturbs chromatin orga-nization without alteration of ESCs’ identities (Supple-mental Fig. 7G,H).

At the molecular level, ChIP analysis confirmed thatSALL1 binds to major satellite DNA in wild-type ESCs(Fig. 6E). Moreover, re-ChIP demonstrated that NANOGand SALL1 co-occupy major satellite DNA in wild-typeESCs, as expected given their direct interaction (Fig. 6F).Importantly, deletion of Sall1 in ESCs leads to loss ofNANOG binding tomajor satellite DNA, thereby demon-strating a requirement for SALL1 in enabling NANOGoccupancy at PCH repeats (Fig. 6E). SALL1 binding wasalso reduced at major satellite DNA in Nanog–/– ESCs(Fig. 6E). Deletion of Sall1 leads to a reorganization ofthe PCH state that is characterized by increased levelsof H3K9me3, decreased levels of H3K9ac, and down-regu-lated major satellite transcription (Fig. 6G; SupplementalFig. 7I). Other repeat sequences such as LINE, SINE, andIAP were unaffected (Supplemental Fig. 7J). Together,these findings identify a role for SALL1 in regulatingPCH organization in ESCs.

Based on the above results, we propose that NANOGand SALL1 codependently maintain heterochromatinorganization in ESCs. To further test this model, wenext addressed whether SALL1 is required for NANOG-mediated heterochromatin reorganization in EpiSCs. Wegenerated Sall1–/– EpiSCs that contained aDOX-inducibleNanog transgene (Supplemental Fig. 7K). After NANOGinduction in these cells, there was no difference in thenumber and appearance of chromocenters (Fig. 7A). Incontrast, NANOG induction together with restorationof SALL1 levels resulted in chromocenter reorganizationto levels typical of ESCs (Fig. 7A), demonstrating thatSALL1 is a necessary cofactor for NANOG-mediated het-erochromatin remodeling. Importantly, the requirementfor SALL1 in heterochromatin remodeling could be by-passed through expression of TALE-CD2, which wasable to remodel chromocenter organization in Sall1–/–

EpiSCs (Fig. 7A). We therefore propose a model in whichSALL1 is required for NANOG binding at major satellitepericentromeric repeats (Fig. 7B). Once at the repeats,the strong transactivation domains of NANOG are ableto promote a more open and active chromatin state atPHC domains.

Discussion

An open and highly dispersed chromatin architecture is adefining property of naïve pluripotency in vitro and invivo (Meshorer et al. 2006; Efroni et al. 2008; Ahmed

Novo et al.




















Figure 6. SALL1 binds NANOG directly and is required for open heterochromatin organization in ESCs. (A) Table showing a subset ofproteins copurifying with 2xFlag-Nanog in EpiSCs, as identified by mass spectrometry. (B, top) Coimmunoprecipitation of endogenousNANOG from wild-type ESC nuclear extracts, analyzed by Western blot (WB). Benzonase (Benzo) and ethidium bromide (Et. Br.) wereaddedwhere indicated. (Bottom) Coimmunoprecipitation of recombinantNANOGand SALL1, analyzed byWestern blot. (C ) ESI analysisof wild-type (WT) and Sall1–/– ESCs. The nuclear membrane is indicated by an arrowhead. Bar, 0.5 μm. Box and whisker plots reveal thedistribution in size of chromatin clusters. P-value was calculated using Student’s t-test. Heterochromatin fiber density was also signifi-cantly increased in Sall1–/– ESCs (data not shown). (D) Chromocenter organization revealed by immunofluorescent analysis ofH3K9me3. OCT4 labeling confirmed the undifferentiated status of the cell type. Bar, 2 µm. Box and whisker plots show the number(left) and size (right) of H3K9me3 foci per nucleus. Data were compared using a one-way ANOVA followed by Bonferroni’s multiple com-parison test. Data were collected from at least two independent experiments. (E) ChIP-qPCR analysis of SALL1 and NANOG at majorsatellite DNA in wild-type, Sall1–/–, and Nanog–/– ESCs. (F ) Re-ChIP-qPCR analysis of NANOG and SALL1 co-occupancy at major sat-ellite, LINE, and SINE DNA in wild-type and Nanog–/– ESCs. (G) ChIP-qPCR for H3K9me3, H3K9ac, and IgG at major satellite DNAin wild-type and Sall1–/– ESCs. Values were normalized to unmodified H3. All qPCR data represent mean ± SD from three biologicalexperiments.






et al. 2010; Boskovic et al. 2014). We identified a criticalnew role for the transcription factor Nanog in the estab-lishment and maintenance of open heterochromatin,thereby forming a direct link between the ESC regu-latory network and nuclear organization in pluripotentcells. Characterization of the mechanism revealed a re-quirement for the C-terminal transactivation domainsof NANOG to be recruited to PCH and the associated reg-ulation of satellite repeats. An alternative set of transcrip-tion factors has been shown to maintain heterochromatinstate through satellite regulation in fibroblast cells (Bulut-Karslioglu et al. 2012), an indication that this mode ofchromatin regulation may be common but involves celltype-specific transcription factors. In ESCs,Nanog is like-ly to function together with key chromatin regulators,such as Chd1, to orchestrate an open higher-order chro-matin structure (Gaspar-Maia et al. 2011). Our analysisof NANOG-interacting proteins now provides a set of ad-ditional factors that may have functional roles in control-ling chromatin organization in ESCs and will be a focus offuture research.

Constitutive heterochromatin is rapidly compactedupon ESC differentiation and embryo development,implying close coordination of chromatin with cell status(Meshorer et al. 2006; Ahmed et al. 2010; Boskovic et al.2014). Our findings suggest that down-regulation ofNanog, one of the earliest events in ESC differentiation,is a key driver of heterochromatin compaction. Converse-ly, heterochromatin decompaction and Nanog inductionare critical events that co-occur at a late stage in cell re-programming (Brambrink et al. 2008; Silva et al. 2009;Fussner et al. 2011; Schwarz et al. 2014). Therefore, ourfindings also have important consequences for controllingheterochromatin organization during reprogramming, anevent that has been shown previously to be a barrier toreprogramming efficiency (Soufi et al. 2012). The precisetiming of themolecular events that lead to heterochroma-tin remodeling during reprogramming will be importantto investigate further (Mattout et al. 2011).

Given that PCH organization is a highly regulated pro-cess in ESCs and, as we show here, has been integratedwithin the pluripotency network, what could be the

Figure 7. Sall1 is required for Nanog-mediated het-erochromatin remodeling. (A) Nanog is unable to re-model chromocenter organization in the absence ofSall1, but recruitment of NANOG-CD2 directly tomajor satellites can bypass the requirement forSall1. Chromocenter organization revealed by immu-nofluorescent analysis of H3K9me3. OCT4 labelingconfirmed the undifferentiated status of the celltype. DOX was applied for 24 h. Bar, 2 µm. Box andwhisker plots show the number of H3K9me3 fociper nucleus. P-values were calculated using Student’st-test. Data were collected from at least two indepen-dent experiments. (B) Model illustrating the role ofNanog inmaintaining an open heterochromatin orga-nization in pluripotent cells.

Novo et al.





role of an open chromatin architecture in ESCs? So far, theprevailing model to explain the function of an open chro-matin configuration in ESCs proposes that it helps main-tain genome plasticity (Gaspar-Maia et al. 2011; Cavalliand Misteli 2013). Our results reinforce the concept thatthe chromatin state of PCH domains is maintained in anunusually open and active form. Interestingly, a recentstudy identified genomic regions that loop and physicallyinteract with PCH (Wijchers et al. 2015). Transcriptionaland epigenetic approaches demonstrated that PCH do-mains are not a strong repressive environment in ESCs,but, instead, this property is acquired upon ESC differenti-ation (Wijchers et al. 2015). These findings are consistentwith a corresponding accumulation of repressivemarks atPCH domains upon ESC differentiation and in somaticcells. It is therefore possible that PCH domains are orga-nized and controlled in ESCs to prevent the unwantedstrong repression that could potentially restrict genomeregulation. Importantly, our results show that Nanog–/–

ESCs can tolerate compaction of heterochromatin do-mains without substantial changes in cell state or theability of the cells to differentiate into all three germ layers(Chambers et al. 2007). However,Nanog–/– ESCs are com-promised, as they do exhibit diminished colony formationand are more prone to spontaneous differentiation thanwild-type ESCs (Mitsui et al. 2003; Chambers et al. 2007).In addition, Sall1-deficient embryos and ESCs have noapparent defects in pluripotency or early development(Nishinakamura et al. 2001; Yuri et al. 2009) despite thedemonstration here that Sall1 is required for open hetero-chromatin organization in ESCs. Therefore, it is possiblethat compaction of heterochromatin domains may desta-bilize and restrict ESCs but that additional events are re-quired to trigger functional changes in ESC state. Thesame model may be true of pluripotent cells in vivo andthereby account for the absence of an early developmentalphenotype in Sall1mutant embryos, although the regula-tive nature of early development may also result in com-promised or unfit cells being excluded from the embryo.An open and active PCH configuration with relatively

low levels of heterochromatin modifications could befunctionally linked with the observation that pericentro-meric-associated proteins bind more loosely or are absentin ESCs (Meshorer et al. 2006; Melcer et al. 2012; Mattoutet al. 2015). Potentially, this class of protein is not able toengage or be retained at PCH in cell types with lower lev-els of H3K9me3 such as ESCs. As PCH regulation andbinding of pericentromeric-associated proteins are criticalfor centromere function in other cell types (Hall et al.2012; Saksouk et al. 2015), it is possible that maintaininga particular PCH architecture may also be linked to pre-serving centromere function in ESCs. ESCs could have ac-quired a unique form of centromere organization alongwith other unusual properties of pluripotent cells, suchas their cell cycle parameters, DNA damage checkpoints,or prolonged maintenance undergoing self-renewal (Bur-don et al. 2002; Weissbein et al. 2014). Therefore, it willbe important in future research to examine centromerefunction more closely in ESCs that have an experimental-ly perturbed heterochromatin organization. Last, future

studies should also investigate how higher-order chroma-tin structure can influence nuclear organization and ge-nome interactions in regions outside of heterochromatinin pluripotent cells; for instance, in coordinating move-ments of chromosome territories upon cell differentiationand reprogramming (Politz et al. 2013).

Materials and methods

Cell lines

E14Tg2a (129P2/OlaHsd; passages 19–28) (Hooper et al. 1987), J1(129S4/SvJae; passages 20–24), EF1 (E14Tg2a-derived Nanog-overexpressing cells; passages 22–26) (Chambers et al. 2003),RCNβH (E14Tg2a-derived Nanog+/–; passages 40–44) (Chamberset al. 2007), RCNβH-B(t) (E14Tg2a-derived Nanog–/–; passages20–30) (Chambers et al. 2007), TβC44cre6 (E14Tg2a-derivedNanog–/–; passages 32–36) (Chambers et al. 2007), TNGA(E14Tg2a-derived Nanog-GFP knock-in) (Chambers et al. 2007),and Sall1-del (E14.1-derived Sall1–/–; passages 20–30) (Nishinaka-mura et al. 2001; Yuri et al. 2009) ESCs were cultured on gelatin-coated surfaces in standard ESC medium (DMEM supplementedwith 15% FBS, 1 mM sodium pyruvate, 0.1 mM 2-mercaptoetha-nol, 0.1 mM nonessential amino acids, 2 mM glutamax, 1000 U/mL LIF). All ESC lines weremale. During expansion of RCNβH-B(t) ESCs, 25 µg/mL hygromycin was added to select for Nanog–/–

cells. Where indicated as 2i conditions, ESCs were cultured formore than five passages on gelatin-coated surfaces in N2B27(1:1 DMEM/F-12:neurobasal, 2 mM glutamax, 0.1 mM 2-mer-captoethanol, 1% B27, 0.5% N2) supplemented with 1 µMPD0325901, 3 µMCHIR99021, and 1000U/mL LIF. ESC differen-tiation was achieved by plating 300,000 cells onto a gelatin-coat-ed 10-cmplate in ESCmedium.After 24 h,mediumwas switchedto ESC medium without LIF (supplemented with 5 µM all-transretinoic acid) and changed daily.Embryo-derived 129S2 (passages 14–28) (Brons et al. 2007) and

B2 (ICR; passages 8–14) (Rugg-Gunn et al. 2012) EpiSCs were cul-tured on 10 µg/mL fibronectin or γ-irradiated mouse embryonicfibroblasts in N2B27 supplemented with 20 ng/mL Activin Aand 12 ng/mL bFGF. Both EpiSC lines were female. Sall1–/–

EpiSCs were generated by converting Sall1–/– ESCs into EpiSCsas described (Rugg-Gunn et al. 2012). Sall1–/– EpiSCs were main-tained for at least 10 passages in EpiSC conditions before use. Seethe Supplemental Material for a detailed description of the trans-genic cell lines used.

Imaging and analysis

Cells were cultured on glass coverslips precoated with gelatin,fibronectin, or γ-irradiatedmouse embryonic fibroblasts. Samplesfor ESI were processed and analyzed as described (Ahmed et al.2010). For the majority of immunofluorescent experiments, cellswere fixed with 2% paraformaldehyde in PBS for 10 min at roomtemperature, washed three timeswith PBS for 5min, and blockedwith 5% FBS and 0.1% Triton X-100 in PBS for 1 h. Cells were in-cubatedwith primary antibody (SupplementalMaterial) in block-ing buffer overnight at 4°C, washed three times with PBS for 5min, and incubated with secondary antibodies for 2 h at roomtemperature. Nuclei were counterstained with DAPI. Imageswere collected on an Olympus FV1000 confocal microscope. Op-tical section thickness ranged from 0.5 to 2 µm. ImageJ softwarewas used to quantify H3K9me3 foci size and intensity using the“analyze particles” tool. Line scan analysis was performed as de-scribed (Fussner et al. 2011).








Acknowledgments

We thank Ludovic Vallier for constitutive Nanog-EpiSCs, Gabri-elle Brons for 129S2 EpiSCs, Prim Singh for H3K9me3 antibody,Maria Elena Torres Padilla for TALE-mClover and luciferase plas-mids, Wellcome Trust Sanger Institute for pCyL43 plasmid, andAndras Nagy for PB-TET and rtTA plasmids. We are grateful toDavid Oxley and Judith Webster for mass spectrometry support,Simon Walker for imaging support, and Anne Segonds-Pichonfor statistical advice. We thank Wolf Reik and Jon Houseley forcomments on themanuscript, andmembers ofWolf Reik’s groupfor helpful discussions. P.J.R.-G. is supported by the WellcomeTrust (WT093736), Biotechnology and Biological Sciences Re-search Council (M022285), and the European Commission Net-work of Excellence EpiGeneSys (HEALTH-F4-2010-257082).This work was also supported with funds from the Canadian In-stitutes of Health Research to J.E. (Team Grant EPS-129129)and D.P.B.-J. D.P.B.-J. holds the Canada Research Chair inMolec-ular and Cellular Imaging. I.C. is supported by the Medical Re-search Council.

References

Ahmed K, Dehghani H, Rugg-Gunn P, Fussner E, Rossant J,Bazett-Jones DP. 2010. Global chromatin architecture reflectspluripotency and lineage commitment in the earlymouse em-bryo. PLoS One 5: e10531.

Bickmore WA, van Steensel B. 2013. Genome architecture:domain organization of interphase chromosomes. Cell 152:1270–1284.

Boskovic A, Eid A, Pontabry J, Ishiuchi T, Spiegelhalter C, RaghuRam EV, Meshorer E, Torres-Padilla ME. 2014. Higher chro-matin mobility supports totipotency and precedes pluripo-tency in vivo. Genes Dev 28: 1042–1047.

Brambrink T, Foreman R, Welstead GG, Lengner CJ, Wernig M,Suh H, Jaenisch R. 2008. Sequential expression of pluripo-tencymarkers during direct reprogramming ofmouse somaticcells. Cell Stem Cell 2: 151–159.

Brons IG, Smithers LE, Trotter MW, Rugg-Gunn P, Sun B,Chuva de Sousa Lopes SM, Howlett SK, Clarkson A,Ahrlund-Richter L, Pedersen RA, et al. 2007. Derivationof pluripotent epiblast stem cells from mammalian embry-os. Nature 448: 191–195.

Bulut-Karslioglu A, Perrera V, ScaranaroM, de la Rosa-VelazquezIA, van de Nobelen S, Shukeir N, Popow J, Gerle B, Opravil S,PaganiM, et al. 2012. A transcription factor-basedmechanismfor mouse heterochromatin formation. Nat Struct Mol Biol19: 1023–1030.

Burdon T, Smith A, Savatier P. 2002. Signalling, cell cycle andpluripotency in embryonic stem cells. Trends Cell Biol 12:432–438.

Carter AC, Davis-Dusenbery BN, Koszka K, Ichida JK, Eggan K.2014. Nanog-independent reprogramming to iPSCs with ca-nonical factors. Stem Cell Rep 2: 119–126.

Cavalli G,Misteli T. 2013. Functional implications of genome to-pology. Nat Struct Mol Biol 20: 290–299.

Chambers I, Colby D, Robertson M, Nichols J, Lee S, Tweedie S,Smith A. 2003. Functional expression cloning of Nanog, a plu-ripotency sustaining factor in embryonic stem cells.Cell 113:643–655.

Chambers I, Silva J, Colby D, Nichols J, Nijmeijer B, RobertsonM, Vrana J, Jones K, Grotewold L, Smith A. 2007. Nanog safe-guards pluripotency andmediates germline development.Na-ture 450: 1230–1234.

Christophorou MA, Castelo-Branco G, Halley-Stott RP, OliveiraCS, Loos R, Radzisheuskaya A, Mowen KA, Bertone P, SilvaJC, Zernicka-Goetz M, et al. 2014. Citrullination regulatespluripotency and histone H1 binding to chromatin. Nature507: 104–108.

Costa Y, Ding J, Theunissen TW, Faiola F, Hore TA, Shliaha PV,Fidalgo M, Saunders A, Lawrence M, Dietmann S, et al.2013. NANOG-dependent function of TET1 and TET2 in es-tablishment of pluripotency. Nature 495: 370–374.

deWit E, Bouwman BA, Zhu Y, Klous P, Splinter E, VerstegenMJ,Krijger PH, Festuccia N, Nora EP, Welling M, et al. 2013. Thepluripotent genome in three dimensions is shaped around plu-ripotency factors. Nature 501: 227–231.

Efroni S, Duttagupta R, Cheng J, Dehghani H, Hoeppner DJ, DashC, Bazett-Jones DP, Le Grice S, McKay RD, Buetow KH, et al.2008. Global transcription in pluripotent embryonic stemcells. Cell Stem Cell 2: 437–447.

Fraser P, Bickmore W. 2007. Nuclear organization of the genomeand the potential for gene regulation. Nature 447: 413–417.

Fussner E, Djuric U, StraussM, Hotta A, Perez-Iratxeta C, LannerF, Dilworth FJ, Ellis J, Bazett-JonesDP. 2011. Constitutive het-erochromatin reorganization during somatic cell reprogram-ming. EMBO J 30: 1778–1789.

Gagliardi A, Mullin NP, Ying Tan Z, Colby D, Kousa AI, Halbrit-ter F, Weiss JT, Felker A, Bezstarosti K, Favaro R, et al. 2013. Adirect physical interaction betweenNanog and Sox2 regulatesembryonic stem cell self-renewal. EMBO J 32: 2231–2247.

Gaspar-MaiaA, AlajemA, Polesso F, SridharanR,MasonMJ,Hei-dersbach A, Ramalho-Santos J, McManus MT, Plath K,Meshorer E, et al. 2009. Chd1 regulates open chromatin andpluripotency of embryonic stem cells. Nature 460: 863–868.

Gaspar-Maia A, AlajemA,Meshorer E, Ramalho-SantosM. 2011.Open chromatin in pluripotency and reprogramming.Nat RevMol Cell Biol 12: 36–47.

Guo G, Yang J, Nichols J, Hall JS, Eyres I, Mansfield W, Smith A.2009. Klf4 reverts developmentally programmed restriction ofground state pluripotency. Development 136: 1063–1069.

Hall LE, Mitchell SE, O’Neill RJ. 2012. Pericentric and centro-meric transcription: a perfect balance required. ChromosomeRes 20: 535–546.

Hart AH, Hartley L, Ibrahim M, Robb L. 2004. Identification,cloning and expression analysis of the pluripotency promotingNanog genes in mouse and human. Dev Dyn 230: 187–198.

Hooper M, Hardy K, Handyside A, Hunter S, Monk M. 1987.HPRT-deficient (Lesch-Nyhan) mouse embryos derived fromgermline colonization by cultured cells.Nature 326: 292–295.

Huangfu D, Maehr R, Guo W, Eijkelenboom A, Snitow M, ChenAE, Melton DA. 2008. Induction of pluripotent stem cells bydefined factors is greatly improved by small-molecule com-pounds. Nat Biotechnol 26: 795–797.

Jauch R, Ng CK, Saikatendu KS, Stevens RC, Kolatkar PR. 2008.Crystal structure and DNA binding of the homeodomain ofthe stem cell transcription factor Nanog. J Mol Biol 376:758–770.

Kanellopoulou C, Muljo SA, Kung AL, Ganesan S, Drapkin R,Jenuwein T, Livingston DM, Rajewsky K. 2005. Dicer-defi-cient mouse embryonic stem cells are defective in differentia-tion and centromeric silencing. Genes Dev 19: 489–501.

Karantzali E, Lekakis V, Ioannou M, Hadjimichael C, Papama-theakis J, Kretsovali A. 2011. Sall1 regulates embryonicstem cell differentiation in association with nanog. J BiolChem 286: 1037–1045.

Lehnertz B, Ueda Y, Derijck AA, Braunschweig U, Perez-BurgosL, Kubicek S, Chen T, Li E, Jenuwein T, Peters AH. 2003.Suv39h-mediated histone H3 lysine 9 methylation directs

Novo et al.





DNAmethylation tomajor satellite repeats at pericentric het-erochromatin. Curr Biol 13: 1192–1200.

Lessard JA, Crabtree GR. 2010. Chromatin regulatory mecha-nisms in pluripotency. Ann Rev Cell Dev Biol 26: 503–532.

Lu J, Gilbert DM. 2007. Proliferation-dependent and cell cycleregulated transcription of mouse pericentric heterochroma-tin. J Cell Biol 179: 411–421.

Mattout A, Biran A,Meshorer E. 2011. Global epigenetic changesduring somatic cell reprogramming to iPS cells. JMolCell Biol3: 341–350.

Mattout A, AaronsonY, Sailaja BS, RaghuRamEV, Harikumar A,Mallm JP, Sim KH, Nissim-Rafinia M, Supper E, Singh PB,et al. 2015. Heterochromatin Protein 1β (HP1β) has distinctfunctions and distinct nuclear distribution in pluripotent ver-sus differentiated cells. Genome Biol 16: 213.

Melcer S, Hezroni H, Rand E, Nissim-Rafinia M, Skoultchi A,Stewart CL, Bustin M, Meshorer E. 2012. Histone modifica-tions and laminA regulate chromatin protein dynamics in ear-ly embryonic stem cell differentiation. Nat Commun 3: 910.

Meshorer E, Yellajoshula D, George E, Scambler PJ, Brown DT,Misteli T. 2006. Hyperdynamic plasticity of chromatin pro-teins in pluripotent embryonic stem cells. Dev Cell 10:105–116.

Mikkelsen TS, Hanna J, Zhang X, KuM,WernigM, Schorderet P,Bernstein BE, Jaenisch R, Lander ES, Meissner A. 2008. Dis-secting direct reprogramming through integrative genomicanalysis. Nature 454: 49–55.

Misteli T. 2007. Beyond the sequence: cellular organization of ge-nome function. Cell 128: 787–800.

Mitsui K, Tokuzawa Y, Itoh H, Segawa K, Murakami M, Takaha-shi K,MaruyamaM,MaedaM, Yamanaka S. 2003. The home-oproteinNanog is required formaintenance of pluripotency inmouse epiblast and ES cells. Cell 113: 631–642.

Miyanari Y, Ziegler-BirlingC, Torres-PadillaME. 2013. Live visu-alization of chromatin dynamicswith fluorescent TALEs.NatStruct Mol Biol 20: 1321–1324.

Nishinakamura R,Matsumoto Y, Nakao K, Nakamura K, Sato A,Copeland NG, Gilbert DJ, Jenkins NA, Scully S, Lacey DL,et al. 2001. Murine homolog of SALL1 is essential for uretericbud invasion in kidney development. Development 128:3105–3115.

Oh JH, Do HJ, Yang HM,Moon SY, Cha KY, Chung HM, Kim JH.2005. Identification of a putative transactivation domain inhuman Nanog. Exp Mol Med 37: 250–254.

Osorno R, Chambers I. 2011. Transcription factor heterogeneityand epiblast pluripotency. Philos Trans R Soc Lond B BiolSci 366: 2230–2237.

Osorno R, Tsakiridis A, Wong F, Cambray N, Economou C,Wilkie R, Blin G, Scotting PJ, Chambers I, Wilson V. 2012.The developmental dismantling of pluripotency is reversedby ectopic Oct4 expression. Development 139: 2288–2298.

Pan GJ, Pei DQ. 2003. Identification of two distinct transactiva-tion domains in the pluripotency sustaining factor nanog.Cell Res 13: 499–502.

Peters AH, O’Carroll D, Scherthan H, Mechtler K, Sauer S,Schofer C, Weipoltshammer K, Pagani M, Lachner M, Kohl-maier A, et al. 2001. Loss of the Suv39h histone methyltrans-ferases impairs mammalian heterochromatin and genomestability. Cell 107: 323–337.

Politz JC, Scalzo D, GroudineM. 2013. Something silent this wayforms: the functional organization of the repressive nuclearcompartment. Ann Rev Cell Dev Biol 29: 241–270.

Rugg-Gunn PJ, Cox BJ, Lanner F, Sharma P, Ignatchenko V,McDonald AC, Garner J, Gramolini AO, Rossant J, KislingerT. 2012. Cell-surface proteomics identifies lineage-specificmarkers of embryo-derived stem cells. Dev Cell 22: 887–901.

Sakaki-YumotoM, Kobayashi C, Sato A, Fujimura S, MatsumotoY, Takasato M, Kodama T, Aburatani H, Asashima M, Yosh-ida N, et al. 2006. The murine homolog of SALL4, a causativegene in Okihiro syndrome, is essential for embryonic stemcell proliferation, and cooperates with Sall1 in anorectal,heart, brain and kidney development. Development 133:3005–3013.

Saksouk N, Simboeck E, Dejardin J. 2015. Constitutive hetero-chromatin formation and transcription inmammals. EpigenetChrom 8: 3.

Schwarz BA, Bar-NurO, Silva JC, Hochedlinger K. 2014.Nanog isdispensable for the generation of induced pluripotent stemcells. Curr Biol 24: 347–350.

Silva J,Nichols J, TheunissenTW,GuoG, vanOostenAL, Barran-don O, Wray J, Yamanaka S, Chambers I, Smith A. 2009.Nanog is the gateway to the pluripotent ground state. Cell138: 722–737.

Soufi A, Donahue G, Zaret KS. 2012. Facilitators and impedi-ments of the pluripotency reprogramming factors’ initial en-gagement with the genome. Cell 151: 994–1004.

Sridharan R, Gonzales-Cope M, Chronis C, Bonora G, McKee R,Huang C, Patel S, Lopez D, Mishra N, Pellegrini M, et al.2013. Proteomic and genomic approaches reveal critical func-tions of H3K9methylation and heterochromatin protein-1γ inreprogramming to pluripotency. Nat Cell Biol 15: 872–882.

Tesar PJ, Chenoweth JG, Brook FA, Davies TJ, Evans EP, MackDL, Gardner RL,McKay RD. 2007.New cell lines frommouseepiblast share defining features with human embryonic stemcells. Nature 448: 196–199.

Wang J, Rao S, Chu J, Shen X, Levasseur DN, Theunissen TW,Orkin SH. 2006. A protein interaction network for pluripo-tency of embryonic stem cells. Nature 444: 364–368.

Weissbein U, Benvenisty N, Ben-David U. 2014. Quality control:genome maintenance in pluripotent stem cells. J Cell Biol204: 153–163.

Wen B, Wu H, Shinkai Y, Irizarry RA, Feinberg AP. 2009. Largehistone H3 lysine 9 dimethylated chromatin blocks distin-guish differentiated from embryonic stem cells. Nat Genet41: 246–250.

Wijchers PJ, Geeven G, Eyres M, Bergsma AJ, Janssen M, Verste-gen M, Zhu Y, Schell Y, Vermeulen C, de Wit E, et al. 2015.Characterization and dynamics of pericentromere-associateddomains in mice. Genome Res 25: 958–969.

Ying QL, Wray J, Nichols J, Batlle-Morera L, Doble B, Woodgett J,Cohen P, Smith A. 2008. The ground state of embryonic stemcell self-renewal. Nature 453: 519–523.

Yuri S, Fujimura S, Nimura K, TakedaN, Toyooka Y, Fujimura Y,Aburatani H, Ura K, Koseki H, Niwa H, et al. 2009. Sall4 is es-sential for stabilization, but not for pluripotency, of embryon-ic stem cells by repressing aberrant trophectoderm geneexpression. Stem Cells 27: 796–805.






10.1101/gad.275685.115Access the most recent version at doi: originally published online April 28, 201630:2016, Genes Dev.

Clara Lopes Novo, Calvin Tang, Kashif Ahmed, et al. heterochromatin organization in mouse embryonic stem cells

regulates pericentromericNanogThe pluripotency factor

Material

Supplemental

http://genesdev.cshlp.org/content/suppl/2016/04/28/gad.275685.115.DC1

References

http://genesdev.cshlp.org/content/30/9/1101.full.html#ref-list-1

This article cites 60 articles, 12 of which can be accessed free at:

License

Commons Creative

.http://creativecommons.org/licenses/by/4.0/License (Attribution 4.0 International), as described at

, is available under a Creative CommonsGenes & DevelopmentThis article, published in

ServiceEmail Alerting

click here.right corner of the article or

Receive free email alerts when new articles cite this article - sign up in the box at the top

© 2016 Novo et al.; Published by Cold Spring Harbor Laboratory Press


http://genesdev.cshlp.org/lookup/doi/10.1101/gad.275685.115

http://genesdev.cshlp.org/content/suppl/2016/04/28/gad.275685.115.DC1

http://genesdev.cshlp.org/content/30/9/1101.full.html#ref-list-1


http://genesdev.cshlp.org/cgi/alerts/ctalert?alertType=citedby&addAlert=cited_by&saveAlert=no&cited_by_criteria_resid=protocols;10.1101/gad.275685.115&return_type=article&return_url=http://genesdev.cshlp.org/content/10.1101/gad.275685.115.full.pdf



Documents

The pluripotency factor Nanog regulates pericentromeric ...genesdev.cshlp.org/content/30/9/1101.full.pdf · The pluripotency factor Nanog regulates pericentromeric heterochromatin