Super-Enhancers at the Nanog Locus Differentially Regulate Neighboring Pluripotency-Associated Genes

Report

Super-Enhancers at the Nanog Locus Differentially

Regulate Neighboring Pluripotency-Associated

Genes

Graphical Abstract

Highlights

d

Genomic editing reveals functional differences in

super-enhancer regulation of Nanog

d

45 enhancer regulates both nearest neighbor genes, Nanog

and Dppa3

d

45 eRNAs specifically regulate Dppa3 by stabilizing

chromatin looping

Authors

Steven Blinka, Michael H. Reimer, Jr.,

Kirthi Pulakanti, Sridhar Rao

Correspondence

sridhar.rao@bcw.edu

In Brief

Super-enhancers are tissue specific

cis-regulatory elements that are transcribed

and regulate genes critical to cell identity.

Blinka et al. show that three neighboring

super-enhancers differentially regulate

the core pluripotency transcription factor

Nanog, and eRNAs produced at one

distal enhancer specifically activate a

different neighboring gene by stabilizing

chromatin looping.

Blinka et al., 2016, Cell Reports 17, 19–28 September 27, 2016ª 2016 The Author(s). http://dx.doi.org/10.1016/j.celrep.2016.09.002

Cell Reports

Report

Super-Enhancers at the Nanog Locus Differentially

Regulate Neighboring Pluripotency-Associated Genes

1_{Department of Cell Biology, Neurobiology, and Anatomy, Medical College of Wisconsin, Milwaukee, WI 53226, USA} 2_{Blood Research Institute, BloodCenter of Wisconsin, Milwaukee, WI 53226, USA}

3_{Department of Pediatrics, Medical College of Wisconsin, Milwaukee, WI 53226, USA} 4_{Lead Contact}

*Correspondence:sridhar.rao@bcw.edu http://dx.doi.org/10.1016/j.celrep.2016.09.002

SUMMARY

Super-enhancers are tissue-specific cis-regulatory

elements that drive expression of genes associated

with cell identity and malignancy. A cardinal feature

of super-enhancers is that they are transcribed to

produce enhancer-derived RNAs (eRNAs). It remains

unclear whether super-enhancers robustly activate

genes in situ and whether their functions are

attribut-able to eRNAs or the DNA element. CRISPR/Cas9

was used to systematically delete three discrete

super-enhancers at the Nanog locus in embryonic

stem cells, revealing functional differences in Nanog

transcriptional regulation. One distal super-enhancer

45 kb upstream of Nanog (

45 enhancer) regulates

both nearest neighbor genes, Nanog and Dppa3.

Interestingly, eRNAs produced at the

45 enhancer

specifically regulate Dppa3 expression by stabilizing

looping of the

45 enhancer and Dppa3. Our work

illustrates that genomic editing is required to

deter-mine enhancer function and points to a method to

selectively target a subset of

super-enhancer-regu-lated genes by depleting eRNAs.

INTRODUCTION

Embryonic stem cells (ESCs) are distinguished by two unique properties: the ability to undergo unlimited self-renewal while maintaining pluripotency and the capacity to differentiate into all embryonic germ layers. Transcription factors (TFs) form the core machinery that maintain pluripotency by binding ESC specific

cis-regulatory elements (CREs) that regulate transcription

inde-pendent of distance and orientation (Chen et al., 2008). While there have been great advances in identifying the genomic location of ESC-specific enhancers based on epigenetic signatures, much remains unknown about their exact mechanism and contribution to cell identity. Murine ESCs are an ideal model to study epigenetic regulation of cell identity, as the TF and chromatin interaction networks (e.g., chromatin immunoprecipitation sequencing [ChIP-seq] and high-throughput chromosome conformation cap-ture [Hi-C]) controlling pluripotency have been well characterized, resulting in identification of highly active cell-type-specific CREs,

termed super-enhancers or stretch enhancers, which associate with genes critical to cell identity (Whyte et al., 2013). In addition, super-enhancers play a critical role in driving disease processes such as cancer, and targeting them with small molecule inhibitors is an active area of research (Love´n et al., 2013). Murine ESCs have 231 super-enhancers, which are large (10 kb) CREs occu-pied by an extremely high density of the pluripotency-critical TFs Nanog, Oct4, and Sox2 (NOS) and coactivators such as Mediator (Whyte et al., 2013). Further, super-enhancers are defined by very high levels of the active enhancer epigenetic mark histone 3 Lysine 27 Acetylation (H3K27Ac), are bound by the initiating form of RNA polymerase II (serine 5 phosphorylated [Ser5P] RNAPII), and are bidirectionally transcribed to produce unspliced long non-coding RNAs (lncRNAs) termed enhancer RNAs (eRNAs;Pulakanti et al., 2013; Hnisz et al., 2013). eRNAs are a mark of highly active enhancers (Ørom et al., 2010; Kim et al., 2010; De Santa et al., 2010; Pulakanti et al., 2013) and have diverse roles in regulating transcription, including stabilizing enhancer-promoter looping by tethering cohesin and Mediator complexes at interacting CREs (Li et al., 2013; Lai et al., 2013; Hsieh et al., 2014; Feng et al., 2006;Figure S1A). How discrete super-enhancers in close proximity work together to regulate genes and the relative contri-bution of the DNA element and its transcribed product (eRNA) to gene expression are currently unknown.

The 160-kb extended Nanog locus on murine chromosome 6 contains four genes (Apobec1, Gdf3, Dppa3, and Nanog) that have critical roles early in embryonic development and are NOS regulated in ESCs. Nanog is a homeobox transcription fac-tor and is required for pluripotency and self-renewal in ESCs (Chambers et al., 2003; Mitsui et al., 2003). Enhancer-promoter interactions at the extended Nanog locus have been studied, revealing a complex network of putative CREs that interact and may co-regulate genes (Levasseur et al., 2008; Apostolou et al., 2013; de Wit et al., 2013). Interestingly, the Nanog locus contains three super-enhancers, but the relative contribution of each CRE to Nanog expression has not been studied in situ. We have previously demonstrated one distal super-enhancer 45 kb upstream of Nanog is transcribed and produces ESC-specific eRNAs (Pulakanti et al., 2013). Thus, the extended

Nanog locus is an ideal genomic region to study the function

of multiple super-enhancers and the regulatory capacity of the eRNAs they produce.

Here, we demonstrate striking differences in super-enhancer transcriptional regulation of Nanog. In addition, we show that a

Cell Reports 17, 19–28, September 27, 2016ª 2016 The Author(s). 19 This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

distal super-enhancer differentially regulates multiple neigh-boring genes and that eRNAs produced at this super-enhancer selectively activate expression of one neighboring gene. Thus, eRNAs are critical to enhancer function. This study provides insight into the mechanism of transcribed enhancers in a cluster of co-regulated genes, specifically parsing out genes that cluster with a highly active transcribed enhancer versus a subset of genes the enhancer regulates.

RESULTS

Super-Enhancers Cluster with Nanog

Whyte et al., (2013) identified 231 tissue-specific super-en-hancers in murine ESCs, which represent <5% of all ensuper-en-hancers. Subsequent studies have examined chromatin regions that interact with these highly active elements by genome-wide chromosome conformation capture (3C) approaches such as Hi-C or circular chromosome conformation capture (4C),

sug-gesting a functional role of long-range chromatin interactions in the maintenance of pluripotency (Schoenfelder et al., 2015; Apostolou et al., 2013; de Wit et al., 2013). Nanog is unique from other pluripotent loci, as it lies within 100 kb of three su-per-enhancers: 45 kb upstream (
45 enhancer), 17-kb region immediately upstream, and 60 kb downstream (+60 enhancer) of Nanog (Figure 1A). In this paper, we focus on a RNAPII-occu-pied enhancer 5 kb upstream of Nanog (
5 enhancer) as it is the most active element within the large proximal super-enhancer region (Figure 1A). Luciferase reporter assays show that all three super-enhancers increase transcriptional activity from the Nanog promoter in both orientations in ESCs but not in fibroblasts, indicating tissue-specific enhancer activity of the Nanog promoter (Pulakanti et al., 2013; Das et al., 2011; Figure 1B).

A common feature of all enhancers is that they interact with pro-moters by looping out the intervening chromatin segment (Guo et al., 2011; Nora et al., 2012; Sanyal et al., 2012; Rao et al.,

0 1 2 3 4 5 6 7 8 9 10 -45 Plus Enhancer -45 Minus Enhancer -5 Plus Enhancer -5 Minus Enhancer +60 Plus Enhancer +60 Minus Enhancer Fold C hange R elat ive to N anog Promoter Only WT ESCs Fibroblasts B A H3K4me1 H3K27Ac Nanog Oct4 Med1 Chr6:mm9 ** ** * ** ** * ** ** ** (0-150) (0-230) (0-60) (0-250) (0-120) Smc1a(0-20) 122,540 kb 122,620kb 122,700kb

Apobec1 Gdf3 Dppa3 -45 SE -5 SENanog Slc2a3 +60 SE

45 kb

35 kb 60 kb

5 kb

Nanog Promoter Luciferase Enhancer

Figure 1. Nanog Neighbors Three Putative Super-Enhancers

(A) Integrated Genome Viewer (IGV) image of genome-wide chromatin immunoprecipitation (ChIP-seq) datasets in wild-type (WT) ESCs at the extended Nanog locus. Super-enhancers are indicated below in red.

(B) Luciferase reporter plasmids containing the Nanog promoter and a fragment of the
45,
5, or +60 enhancer in both orientations were transfected into WT ESCs or a fibroblast cell line (NIH/3T3). All samples had a fold change (y axis) and statistical analysis calculated relative to a plasmid containing Nanog promoter only, which was set to one (dashed line). n = 3. Error bars indicate the SEM between experimental replicates. *p < 0.05, **p < 0.01 using a one-sample Student’s t test.

2014;Figure S1A). To detect interaction of each super-enhancer with Nanog, we used 3C. Previous work has shown that the
5 and
45 enhancers interact with Nanog (Levasseur et al., 2008; Apostolou et al., 2013), which we confirmed (Figures 2A and 2B). In addition, we identified an unappreciated interaction be-tween the +60 enhancer and Nanog in ESCs (Figure 2C). Interac-tions were increased above a control fibroblast cell line where

Nanog is not expressed. Further, published Hi-C data show

clus-tering of numerous CREs at the extended Nanog locus, similar to our 3C data (Figure S1B). To verify locus-wide clustering of

other CREs at this locus in ESCs, we demonstrate an interaction between Dppa3 and the +60 enhancer (Figure S1C) and an enhancer-enhancer interaction between the
45 and
5 en-hancers (Figure S1D). In sum, we confirm locus-wide clustering of co-regulated (NOS-occupied) CREs with Nanog that function as putative enhancers in reporter assays.

Functional Differences of Nanog Super-Enhancers Super-enhancers have been proposed to control genes critical to cell identity such as Nanog. Given that all three super-enhancers

B A C 0 0.2 0.4 0.6 0.8 1 1.2 50 55 60 65 No rma lize d Int er ac ti on R at io

Distance (kb) From Nanog Promoter WT ESCs Fibroblasts 0 0.2 0.4 0.6 0.8 1 1.2 0 10 20 30 40 50 60 No rm alize d Int er ac tio n R ati o

Distance (kb) From Nanog Promoter

WT ESCs Fibroblasts 0 0.2 0.4 0.6 0.8 1 1.2 0 1 2 3 4 5 6 7 No rm al ize d In te ra cti o n Ra ti o

Distance (kb) From Nanog Promoter

WT ESCs Fibroblasts HaeIII Nanog H3K27Ac Smc1a Primers -5 Enhancer Nanog (0-230) (0-50) (0-15) HaeIII Nanog H3K27Ac Smc1a Primers -45 Enhancer Nanog (0-50) (0-230) (0-15) HaeIII Nanog H3K27Ac Smc1a Primers +60 Enhancer Nanog (0-50) (0-280) (0-15) ##

*

*

Figure 2. A Looping Event Is Detected between Nanog and Three Super-Enhancers by Chromosome Conformation Capture

(A–C) The x axis indicates genomic distance (kilobases) from the Nanog promoter (anchor), and the y axis indicates the normalized interaction ratio. The peak interaction was normalized to one for each WT ESC experimental replicate. IGV images showing HaeIII cut sites, 3C primers, and ChIP-seq binding profiles in WT ESCs are shown below each graph. The anchor primer is not shown in (C). Error bars indicate the SEM between experimental replicates. *p < 0.05 and #p < 0.05, ## p < 0.01 using a one-sample and two-sample Student’s t test, respectively.

are putative Nanog enhancers, we asked whether each CRE is required for robust Nanog expression. To determine in situ enhancer function, we systematically deleted each CRE using CRISPR/Cas9 genomic editing (Figure 3A). To delete the
5 and +60 enhancer, we designed sequence-specific guide RNAs (gRNAs) flanking each enhancer (Figures S2A and S2B). We obtained multiple independent clones containing large biallelic deletions (10.5 kb) of the +60 enhancer. Using the same approach, we efficiently recovered clones containing only monoallelic deletions (2.5 kb) of the proximal
5 enhancer (Figure S2C). Finally, using an alternative CRISPR-mediated homology-directed repair (HDR) approach, we obtained several clones containing biallelic deletions (2.85 kb) of the
45 enhancer (Figure S2D). All super-enhancer-deleted clones recovered retain colony morphology and stain positive for alkaline phosphatase, a surface marker of pluripotency (Figure S2E).

Surprisingly, knockout of individual super-enhancers resulted in different effects on gene expression at the extended Nanog locus. Specifically, biallelic deletion of the
45 enhancer results in decreased expression of both nearest neighbor genes Nanog and Dppa3, while biallelic deletion of the +60 enhancer results in minimal changes in expression of any gene at the Nanog locus (Figures 3B and 3C). Presence of a neomycin resistance (Neo) cassette in place of the
45 enhancer had modest effects on Dppa3 expression but did not alter Nanog expres-sion. Differences in expression of Apobec1 and Oct4 between
45-enhancer-deleted clones with and without Neo is likely clone-to-clone variation (Figure S3C). Interestingly, monoallelic
5-enhancer-deleted ESCs show an 50% decrease in Nanog expression. This suggests that the
5 enhancer is required for robust Nanog expression and biallelic deleted clones were not recovered likely due to intolerance for enhancer loss secondary to loss of Nanog expression, in line with work demonstrating

Nanog+/
, but not Nanog
/
, mice are viable (Mitsui et al., 2003). No significant change in expression was observed at other nearby NOS-regulated genes (Gdf3 and Apobec1) following deletion of any enhancer (Figure 3B). While
5- and
45-enhancer-deleted clones showed a 40%–50% decrease in Nanog expression, ESCs retain wild-type (WT) expression of other pluripotency-associated TFs, indicating the pluripo-tency network is intact (Figure S3A). Accordingly, there was no biologically significant increase (10- to 100-fold) in a range of lineage-specific markers as we have observed previously when ESCs lose pluripotency and differentiate (Stelloh et al., 2016;Figure S3B). These findings are in contrast to genome editing studies at the Sox2 locus where biallelic deletion of a distal super-enhancer in ESCs resulted in an 80%–90% decrease in Sox2 expression and elevated expression of germ layer lineage markers (Zhou et al., 2014; Li et al., 2014). Taken together, despite identical epigenetic signatures, re-porter activity, and chromatin interaction, genomic editing reveals important functional variation of super-enhancers in

Nanog regulation, where only the most proximal CRE (
5

enhancer) appears critical to Nanog expression. These results provide insights into the complexity of enhancer function during development and demonstrate that isolated super-enhancers have differential function on cell-critical gene expression that can only be uncovered by deleting the CREs (Figure 3C).

More-over, this parallels work showing differential function of constit-uent enhancers within one large super-enhancer (Huang et al., 2016).

The
45 Enhancer Regulates Nanog and Dppa3

Surprisingly, deletion of the
45 enhancer resulted in an 60% decrease in Dppa3 expression (Figure 3B). Dppa3 lies 35 kb upstream of the
45 enhancer and has roles in DNA imprinting in very early embryogenesis but is dispensable for pluripotency (Nakamura et al., 2007;Figures 1A and 3B). To confirm that reduction in Dppa3 expression is not a result of decreased Nanog (or vice versa), we depleted either Dppa3 or Nanog mRNA by RNAi. Nanog depletion resulted in a >10-fold increase in Dppa3 expression, consistent with a previous report that Nanog represses Dppa3 (Sharov et al., 2008; Figure S4A). Dppa3 depletion led to minimal changes in Nanog expression and other pluripotency genes such as Sall4 (Figure S4B), demon-strating that Dppa3 does not regulate pluripotency. We identified an interaction of the
45 enhancer with the Dppa3 promoter in ESCs using 3C, which was not present in fibroblasts (Figure 4A). Finally, we show that the
45 enhancer activates Dppa3 in reporter assays in ESCs, but not in fibroblasts (Figure 4B). Thus, we determined that the
45 enhancer regulates both nearest neighbor genes within a highly active region of co-regu-lated genes in ESCs. Further, the modest increase in Dppa3 expression in
5 enhancer monoallelic ESCs demonstrates that the
5 enhancer regulates Nanog, but not Dppa3 (Figures 3B and 3C). Interestingly, despite studies showing clustering of Apobec1 and Gdf3 with Nanog (Levasseur et al., 2008; Apostolou et al., 2013), our genome editing results reveal that none of the super-enhancers regulate these genes (Figure 3B). This suggests linear DNA proximity of the super-enhancer is of functional importance within this compact cluster of co-regu-lated CREs.

eRNAs Produced at the
45 Enhancer Specifically

Regulate Dppa3

In just a few years since their discovery, eRNAs have been shown to have diverse functions in transcriptional regulation, including roles in enhancer-promoter looping (Lai et al., 2013; Li et al., 2013; Hsieh et al., 2014). We previously identified ESC-specific eRNAs produced at the
45 enhancer (Pulakanti et al., 2013). The
45 enhancer serves as an interesting opportunity to study eRNA function, as this CRE regulates two neighboring genes, Nanog and Dppa3. To identify activating properties of
45 eRNAs at these two genes, we tar-geted eRNAs for depletion with modified antisense oligonucleotides (ASOs). eRNAs are restricted to the nucleus, and therefore ASOs, which act by a nuclear-specific RNase H mechanism, have been used to efficiently deplete them. ASOs were designed to target peak levels of eRNAs as determined by global run-on sequencing (GRO-seq;Kaikkonen et al., 2013; Li et al., 2013;Figure S2D), which detects nascent RNA transcription and thus is a sensitive technique to detect eRNAs. Our level of eRNA depletion is comparable to other studies, as ASOs reproducibly depleted eRNAs by >60% rela-tive to a non-targeting ASO (Kaikkonen et al., 2013; Li et al., 2013; Figure 5A). Maintaining >50% depletion is challenging,

0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

-5 Enhancer +/- -45 Enhancer +60 Enhancer

-/-Fold C h an ge R elati vet o W T E S C s 0 0.2 0.4 0.6 0.8 1 1.2

-5 Enhancer +/- -45 Enhancer +60 Enhancer

-/-Fold C hange R elat ive to W T ESC s

**

_**

*

*

*

**

-5 Enhancer +/- -45 Enhancer +60 Enhancer

-/-Fold C h ange Rel ati ve toW T E S C s 0 0.2 0.4 0.6 0.8 1 1.2 1.4

-5 Enhancer +/- -45 Enhancer +60 Enhancer

-/-Fold C h ange R el at iv e to W T E SC s 2.85 kb -/-2.5 kb

+/-*

**

Gdf3 mRNA Apobec1 mRNA

C A Nanog Dppa3 -45 -5 +60 10.5 kb -/-Cas9 gRNA Nanog Dppa3 -45 Enhancer Biallelic Deletion -5 Enhancer Monoallelic Deletion +60 Enhancer Biallelic Deletion

Gene Expression Changes Following Enhancer Deletion

**

##

Figure 3. Genomic Editing Reveals Functional Differences in Super-Enhancer Regulation of Nanog

(A) Schematic of CRISPR-mediated super-enhancer deletion at the Nanog locus.

(B) Expression of genes within the extended Nanog locus in WT ESCs. Two representative clones are shown for
45- (Neo present),
5-, and +60-enhancer-deleted clones. Error bars indicate the SEM between experimental replicates. n = 3. All samples had a fold change (y axis) and statistical analysis (one-sample Student’s t test) calculated relative to WT ESCs that were set to one (dashed line). A two-sample Student’s t test was used to compare clones. *p < 0.05, **p < 0.01 and ## p < 0.01 using a one-sample and two-sample Student’s t test, respectively.

(C) Chart depicting changes in gene expression of Nanog and Dppa3 following deletion of each enhancer.

as eRNAs are a short-lived molecule and restricted to focal region(s) of the nucleus. Interestingly, 24-hr depletion of eRNAs at the
45 enhancer showed a >40% decrease in Dppa3 expression, with no change in Nanog expression (Figure 5A). This is in contrast to changes following
45 enhancer deletion, which result in decreased expression of both genes ( Fig-ure 3B). This demonstrates specificity of eRNAs to regulate only one of the nearest neighbor genes and that eRNAs are partially required for the cis-activating properties of this su-per-enhancer.

Previous studies have shown that lncRNAs interact with epigenetic regulators involved in DNA methylation (Berghoff et al., 2013), catalyzation of enhancer histone marks H3K27Ac (Wang et al., 2008), and histone 3 Lysine 4 methylation (H3K4me; Yang et al., 2014). To examine the mechanism by which eRNAs regulate Dppa3, we first tested for changes in active epigenetics marks of the enhancer. Following eRNA depletion, we observed no change in H3K4me1 or H3K27Ac levels at the
45 enhancer and a putative transcribed enhancer upstream of Oct4 (control genomic region on a different chromo-some) by chromatin immunoprecipitation (Figures S5A and S5B). Highly active enhancers also exhibit DNA hypomethylation compared to typical enhancers. Following eRNA depletion, we observed no change in DNA methylation at two CpG clusters within the
45 enhancer (Pulakanti et al., 2013;Figure S5C). In sum, transient depletion of eRNAs does not result in alteration of epigenetics marks of an active enhancer, implying eRNAs are not required for their maintenance.

Recent work highlights a role of eRNAs in stabilizing enhancer-promoter looping through interactions with the cohesin and Mediator complexes (Li et al., 2013; Lai et al., 2013). Cohesin is enriched at super-enhancers, which are RNAPII bound and transcribed (Hnisz et al., 2013). Using published enhancer data-sets, we show that Smc1a (a component of the cohesin complex) is enriched at all ESC-transcribed enhancers relative to eRNA-negative enhancers (Pulakanti et al., 2013;Figure S5D). This sug-gests that eRNAs may have a generalized role to tether looping factors such as cohesin to stabilize chromatin interactions in ESCs. To determine the mechanism underlying altered

Dppa3 expression, we tested whether eRNA depletion disrupts

enhancer-promoter looping by 3C. Using eRNA ASO 1, which results in the most efficient depletion of eRNAs, we observed a 50% decrease in looping of the
45 enhancer at the Dppa3 promoter (Figure 5B). There was no change in interaction of the
45 enhancer with Nanog or the
5 enhancer (Figure 5B). Other interactions at the Nanog locus were also unchanged following eRNA depletion (Figure S5E). Thus,
45 eRNAs stabi-lize specific enhancer-promoter interactions within a cluster of pluripotency-associated genes to drive robust Dppa3 expression. Our results are in agreement with studies showing no change in active enhancer marks following disruption of enhancer-promoter looping by depleting cohesin (Seitan et al., 2013). Finally, we confirmed that deletion of the
45 enhancer does not alter other super-enhancer interactions with Nanog (Figure 5C), further substantiating that
45 eRNAs specifically regulate a single gene within the cluster of CREs and that the
5 enhancer is sufficient to maintain pluripotency (Figures 3B and5C). 0 10 20 30 40 50 60

-45 Plus Enhancer -45 Minus Enhancer

Fold Change Relativ e to Dppa3 P ro moter O nl y WT ESCs Fibroblasts B A

Dppa3 Promoter Luciferase Enhancer

*

*

Distance (kb) From Dppa3 Promoter

WT ESCs Fibroblasts * # Primers HaeIII H3K27Ac Smc1a Nanog Dppa3 -45 Enhancer (0-195) (0-20) (0-17) # #

Figure 4. The
45 Enhancer Regulates Dppa3

(A) A looping event is detected between the
45 enhancer and Dppa3 by 3C in WT ESCs. The x axis indicates genomic distance (kilobases) from the Dppa3 promoter (anchor), and the y axis indicates the normalized interaction ratio. The peak interaction was normalized to one for each WT ESC experimental replicate. The IGV image showing HaeIII cut sites, 3C primers, and ChIP-seq binding profiles in WT ESCs is shown below the graph.

(B) A luciferase reporter plasmid containing the Dppa3 promoter and a frag-ment of the
45 enhancer in both orientations was transfected into WT ESCs or a fibroblast cell line (NIH/3T3). All samples had a fold change (y axis) and statistical analysis calculated relative to a plasmid containing Dppa3 promoter only, which was set to one (dashed line). n = 3.

Error bars indicate the SEM between experimental replicates. *p < 0.05 and # p < 0.05 using a one-sample and two-sample Student’s t test, respectively.

DISCUSSION

In summary, our data demonstrate that in situ genomic editing is required to unravel the functional role of enhancers, particu-larly when numerous co-regulated CREs cluster within a com-pacted chromatin space. Expression changes following CRE

ablation suggests there is a super-enhancer functional hierar-chy at the Nanog locus, where the
5 enhancer is the only CRE required for robust Nanog expression and pluripotency. Further, by combining genome editing with eRNA depletion, we demonstrate that eRNAs have a functional role by specifically regulating one of the enhancer associated genes

0 0.2 0.4 0.6 0.8 1 1.2 -45:Dppa3 -45:-5 -45:Nanog Normal ized Interacti on Rati o -5 0 0.2 0.4 0.6 0.8 1 1.2 1.4

Nanog mRNA Dppa3 mRNA -45 eRNAs

F o ld C h an g e R el ati ve to N o n-Targeting ASO eRNA ASO 1 eRNA ASO 2 A

* *

*

**

** **

Bidirectional RNAPII (Ser5P) Chromatin Looping Machinery Gene Super-Enhancer Nanog +60 -45 Dppa3 eRNA Depletion 5’ _{5’ 3’}

Elongating RNAPII (Ser2P) -5 Nanog Dppa3 -45 -5 +60

*

Figure 5.
45 eRNAs Regulate Dppa3 Expression by Stabilizing Enhancer-Promoter Interactions

(A) 24-hr depletion of
45 eRNAs using antisense oligonucleotides (ASOs) results in a reduction in Dppa3 expression. All samples had a fold change (y axis) and statistical analysis calculated relative to a non-targeting ASO, which was set to one (dashed line). n = 4.

(B) Chromatin interactions (3C) between the
45 enhancer and neighboring cis-regulatory elements following eRNA depletion using eRNA ASO 1. Statistical analysis was calculated relative to the non-targeting ASO, which was set to one (dashed line). n = 3.

(C) Chromatin interactions (3C) between Nanog and the
5 and +60 enhancers in a
45-enhancer-deleted ESC line (Neo removed). Statistical analysis was calculated relative to WT ESCs, which was set to one (dashed line). n = 3.

For (A)–(C) error bars indicate SEM between experimental replicates. *p < 0.05, **p < 0.01 using a one-sample Student’s t test.

(D) Two-dimensional schematic of the extended Nanog locus before and after
45 eRNA depletion.
45 eRNAs maintain interaction of Dppa3 and the
45 enhancer and high levels of Dppa3 expression.

(Figure 5D). This is in contrast to recent work byParalkar et al., (2016), where depletion of an enhancer-transcribed long inter-genic non-coding RNA (Lockd), which belongs to a class of lncRNAs distinct from eRNAs and is capable of acting in

trans, did not phenocopy gene expression changes following

enhancer deletion. Rather, our work supports a model whereby eRNAs have critical roles in organizing three-dimensional nu-clear interactions, thereby activating a select subset of interact-ing genes regulated by a super-enhancer (Li et al., 2013; Lai et al., 2013; Hsieh et al., 2014).

Under commonly used in vitro serum and leukemia inhibitory factor (LIF) culture conditions, Nanog and Dppa3 are two of the most heterogeneously expressed genes on a cell-to-cell basis (Galonska et al., 2015). ESCs maintained in serum-free defined conditions containing two kinase inhibitors to suppress differ-entiation cues (2i; Mek and Gsk3 inhibitors) have higher and more homogenous levels of Nanog as well as Dppa3 (Silva et al., 2009; Galonska et al., 2015). These ESCs are in an epigenetically and transcriptionally distinct ‘‘naive’’ pluripotent state, likely as a result of dynamic NOS binding at enhancer elements (Galonska et al., 2015). Interestingly, gene expres-sion differences in Nanog and Dppa3 in these two in vitro conditions parallel expression changes observed in WT and
45-enhancer-deleted ESCs. This suggests that dynamic epigenetic changes in the
45 enhancer landscape in the two ESC states result in differential regulation of Nanog and

Dppa3.

In addition to driving expression of lineage-critical factors during development, super-enhancers play a critical role in pathologic processes, including malignancies (Hnisz et al., 2013). Inhibition of transcriptional co-activators and chromatin regulators that densely occupy super-enhancers has been shown to decrease aberrant expression of oncogenes in tumor cells, including Myc (Love´n et al., 2013). However, current approaches using small molecules to target co-activators such as Brd4 lack specificity and disrupt most genes regulated by super-enhancers. Our work points to a method to selectively target a subset of genes regulated by super-enhancers in pathologic states. ASOs, which are used as RNA-targeting therapies for a variety of diseases (Arun et al., 2016), offer an intriguing alternative to small molecules, as they allow direct pharmacological targeting of activating eRNAs at oncogenic loci.

EXPERIMENTAL PROCEDURES

Cell Culture, RNA Isolation, and Real-Time qRT-PCR

ESCs were cultured under previously described conditions (Rao et al., 2010). Total RNA isolation, cDNA conversion, and qPCR was performed as previously described (Pulakanti et al., 2013; Stelloh et al., 2016). Primers used for qRT-PCR, ChIP-qPCR, and 3C are listed inTable S1(all primers in this study were designed using mm9).

Reporter Plasmid Generation

Luciferase reporter plasmids were generated and analyzed similar toPulakanti et al. (2013). Each enhancer was cloned into the SalI site (in both orientations) and each promoter into the KpnI/XhoI site of the Firefly luciferase plasmid (pGL2, Promega). Reporter constructs were transiently transfected into cells using Lipofectamine 2000 (Invitrogen) along with a control plasmid (Renilla luciferase, pRL-EF) for 30 hours.

Chromosome Conformation Capture

3C libraries for ESC and NIH/3T3 crosslinked chromatin were generated as described inStelloh et al., (2016)using a modified protocol fromHage`ge et al., (2007). Interaction between the anchoring point and distal fragments was determined by SYBR Green qPCR and normalized to bacterial artificial chromosome/clone (BAC) templates. Fold changes were divided by relative interaction at the Ercc3 locus (a housekeeping gene known to loop in all cell types) to account for differences in library generation (Table S1). At least two independent libraries were used for each cell line, and statistical analysis was calculated relative to WT ESCs. For each genomic region tested, the peak normalized interaction for each ESC library was set to one.

Genome Editing

To generate
45 Nanog super-enhancer-deleted ESC clones, a single gRNA designed using the CRISPR design tool (http://crispr.mit.edu/) was cloned into the Cas9-expressing vector px459 (Cong et al., 2013). An HDR vector (pL452) containing a floxed Neo cassette flanked by homology regions was co-transfected along with the gRNA in WT ESCs. Individual clones that were resistant to both puromycin and G418 were isolated and expanded for geno-typing (Table S1). To generate
5 and +60 Nanog super-enhancer-deleted ESC clones, we used two gRNAs (designed same as above) as described in

Zhou et al., (2014)andHuang et al., (2016)(Table S1).

Antisense Oligonucleotides

ASOs targeting eRNAs produced at the
45 Nanog super-enhancer were de-signed by Integrated DNA Technologies (Table S1). ASOs were designed with phosphorothioate bonds and a 10-bp gapmer flanked by 5-bp blocks of 20 -O-methyl modified ribonucleotides that protect the ASO from nuclease degrada-tion. ASOs were transiently transfected at 100 nM with Lipofectamine 2000.

SUPPLEMENTAL INFORMATION

Supplemental Information includes Supplemental Experimental Procedures, five figures, and one table and can be found with this article online athttp:// dx.doi.org/10.1016/j.celrep.2016.09.002.

AUTHOR CONTRIBUTIONS

S.B. performed all experiments, while M.H.R. provided assistance with spe-cific experiments. K.P. assisted with computational analysis. S.B. and S.R. conceived the study and designed experiments.

ACKNOWLEDGMENTS

We thank Christopher Glass and Minna Kaikkonen for the suggestion of using ASOs for eRNA depletion, Kim Lennox (Integrated DNA Technologies) for designing ASOs, and Sergey Tarima for assistance with statistical analysis. This work was supported in part by funding from the NIDDK (grant DK103502) to S.B. and the MCW MSTP T32 (NIGMS, grant GM080202). The majority of support came from Midwest Athletes against Childhood Cancer, the American Society of Hematology, and Hyundai Hope on Wheels (all to S.R.). Received: February 5, 2016 Revised: June 13, 2016 Accepted: August 30, 2016 Published: September 27, 2016 REFERENCES

Apostolou, E., Ferrari, F., Walsh, R.M., Bar-Nur, O., Stadtfeld, M., Cheloufi, S., Stuart, H.T., Polo, J.M., Ohsumi, T.K., Borowsky, M.L., et al. (2013). Genome-wide chromatin interactions of the Nanog locus in pluripotency, differentiation, and reprogramming. Cell Stem Cell 12, 699–712.

Arun, G., Diermeier, S., Akerman, M., Chang, K.C., Wilkinson, J.E., Hearn, S., Kim, Y., MacLeod, A.R., Krainer, A.R., Norton, L., et al. (2016). Differentiation of

mammary tumors and reduction in metastasis upon Malat1 lncRNA loss. Genes Dev. 30, 34–51.

Berghoff, E.G., Clark, M.F., Chen, S., Cajigas, I., Leib, D.E., and Kohtz, J.D. (2013). Evf2 (Dlx6as) lncRNA regulates ultraconserved enhancer methylation and the differential transcriptional control of adjacent genes. Development 140, 4407–4416.

Chambers, I., Colby, D., Robertson, M., Nichols, J., Lee, S., Tweedie, S., and Smith, A. (2003). Functional expression cloning of Nanog, a pluripotency sus-taining factor in embryonic stem cells. Cell 113, 643–655.

Chen, X., Xu, H., Yuan, P., Fang, F., Huss, M., Vega, V.B., Wong, E., Orlov, Y.L., Zhang, W., Jiang, J., et al. (2008). Integration of external signaling pathways with the core transcriptional network in embryonic stem cells. Cell 133, 1106–1117.

Cong, L., Ran, F.A., Cox, D., Lin, S., Barretto, R., Habib, N., Hsu, P.D., Wu, X., Jiang, W., Marraffini, L.A., and Zhang, F. (2013). Multiplex genome engineering using CRISPR/Cas systems. Science 339, 819–823.

Das, S., Jena, S., and Levasseur, D.N. (2011). Alternative splicing produces Nanog protein variants with different capacities for self-renewal and pluripo-tency in embryonic stem cells. J. Biol. Chem. 286, 42690–42703.

De Santa, F., Barozzi, I., Mietton, F., Ghisletti, S., Polletti, S., Tusi, B.K., Muller, H., Ragoussis, J., Wei, C.L., and Natoli, G. (2010). A large fraction of extragenic RNA pol II transcription sites overlap enhancers. PLoS Biol. 8, e1000384.

de Wit, E., Bouwman, B.A., Zhu, Y., Klous, P., Splinter, E., Verstegen, M.J., Krijger, P.H., Festuccia, N., Nora, E.P., Welling, M., et al. (2013). The pluripo-tent genome in three dimensions is shaped around pluripotency factors. Nature 501, 227–231.

Feng, J., Bi, C., Clark, B.S., Mady, R., Shah, P., and Kohtz, J.D. (2006). The Evf-2 noncoding RNA is transcribed from the Dlx-5/6 ultraconserved region and functions as a Dlx-2 transcriptional coactivator. Genes Dev. 20, 1470– 1484.

Galonska, C., Ziller, M.J., Karnik, R., and Meissner, A. (2015). Ground state conditions induce rapid reorganization of core pluripotency factor binding before global epigenetic reprogramming. Cell Stem Cell 17, 462–470.

Guo, C., Gerasimova, T., Hao, H., Ivanova, I., Chakraborty, T., Selimyan, R., Oltz, E.M., and Sen, R. (2011). Two forms of loops generate the chromatin conformation of the immunoglobulin heavy-chain gene locus. Cell 147, 332–343.

Hage`ge, H., Klous, P., Braem, C., Splinter, E., Dekker, J., Cathala, G., de Laat, W., and Forne´, T. (2007). Quantitative analysis of chromosome conformation capture assays (3C-qPCR). Nat. Protoc. 2, 1722–1733.

Hnisz, D., Abraham, B.J., Lee, T.I., Lau, A., Saint-Andre´, V., Sigova, A.A., Hoke, H.A., and Young, R.A. (2013). Super-enhancers in the control of cell identity and disease. Cell 155, 934–947.

Hsieh, C.L., Fei, T., Chen, Y., Li, T., Gao, Y., Wang, X., Sun, T., Sweeney, C.J., Lee, G.S., Chen, S., et al. (2014). Enhancer RNAs participate in androgen receptor-driven looping that selectively enhances gene activation. Proc. Natl. Acad. Sci. USA 111, 7319–7324.

Huang, J., Liu, X., Li, D., Shao, Z., Cao, H., Zhang, Y., Trompouki, E., Bowman, T.V., Zon, L.I., Yuan, G.C., et al. (2016). Dynamic control of enhancer reper-toires drives lineage and stage-specific transcription during hematopoiesis. Dev. Cell 36, 9–23.

Kaikkonen, M.U., Spann, N.J., Heinz, S., Romanoski, C.E., Allison, K.A., Stender, J.D., Chun, H.B., Tough, D.F., Prinjha, R.K., Benner, C., and Glass, C.K. (2013). Remodeling of the enhancer landscape during macrophage acti-vation is coupled to enhancer transcription. Mol. Cell 51, 310–325.

Kim, T.K., Hemberg, M., Gray, J.M., Costa, A.M., Bear, D.M., Wu, J., Harmin, D.A., Laptewicz, M., Barbara-Haley, K., Kuersten, S., et al. (2010). Widespread transcription at neuronal activity-regulated enhancers. Nature 465, 182–187.

Lai, F., Ørom, U.A., Cesaroni, M., Beringer, M., Taatjes, D.J., Blobel, G.A., and Shiekhattar, R. (2013). Activating RNAs associate with Mediator to enhance chromatin architecture and transcription. Nature 494, 497–501.

Levasseur, D.N., Wang, J., Dorschner, M.O., Stamatoyannopoulos, J.A., and Orkin, S.H. (2008). Oct4 dependence of chromatin structure within the extended Nanog locus in ES cells. Genes Dev. 22, 575–580.

Li, W., Notani, D., Ma, Q., Tanasa, B., Nunez, E., Chen, A.Y., Merkurjev, D., Zhang, J., Ohgi, K., Song, X., et al. (2013). Functional roles of enhancer RNAs for oestrogen-dependent transcriptional activation. Nature 498, 516–520.

Li, Y., Rivera, C.M., Ishii, H., Jin, F., Selvaraj, S., Lee, A.Y., Dixon, J.R., and Ren, B. (2014). CRISPR reveals a distal super-enhancer required for Sox2 expression in mouse embryonic stem cells. PLoS ONE 9, e114485.

Love´n, J., Hoke, H.A., Lin, C.Y., Lau, A., Orlando, D.A., Vakoc, C.R., Bradner, J.E., Lee, T.I., and Young, R.A. (2013). Selective inhibition of tumor oncogenes by disruption of super-enhancers. Cell 153, 320–334.

Mitsui, K., Tokuzawa, Y., Itoh, H., Segawa, K., Murakami, M., Takahashi, K., Maruyama, M., Maeda, M., and Yamanaka, S. (2003). The homeoprotein Nanog is required for maintenance of pluripotency in mouse epiblast and ES cells. Cell 113, 631–642.

Nakamura, T., Arai, Y., Umehara, H., Masuhara, M., Kimura, T., Taniguchi, H., Sekimoto, T., Ikawa, M., Yoneda, Y., Okabe, M., et al. (2007). PGC7/Stella protects against DNA demethylation in early embryogenesis. Nat. Cell Biol. 9, 64–71.

Nora, E.P., Lajoie, B.R., Schulz, E.G., Giorgetti, L., Okamoto, I., Servant, N., Piolot, T., van Berkum, N.L., Meisig, J., Sedat, J., et al. (2012). Spatial partition-ing of the regulatory landscape of the X-inactivation centre. Nature 485, 381–385.

Ørom, U.A., Derrien, T., Beringer, M., Gumireddy, K., Gardini, A., Bussotti, G., Lai, F., Zytnicki, M., Notredame, C., Huang, Q., et al. (2010). Long noncoding RNAs with enhancer-like function in human cells. Cell 143, 46–58.

Paralkar, V.R., Taborda, C.C., Huang, P., Yao, Y., Kossenkov, A.V., Prasad, R., Luan, J., Davies, J.O., Hughes, J.R., Hardison, R.C., et al. (2016). Unlinking an lncRNA from Its Associated cis Element. Mol. Cell 62, 104–110.

Pulakanti, K., Pienello, L., Stelloh, C., Blinka, S., Allred, J., Milanovich, S., Peterson, J., Wang, A., Yuan, G.C., and Rao, S. (2013). Enhancer transcribed RNAs arise from hypomethylated, Tet-occupied genomic regions. Epigenetics 8, 1303–1320.

Rao, S., Zhen, S., Roumiantsev, S., McDonald, L.T., Yuan, G.C., and Orkin, S.H. (2010). Differential roles of Sall4 isoforms in embryonic stem cell pluripo-tency. Mol. Cell. Biol. 30, 5364–5380.

Rao, S.S., Huntley, M.H., Durand, N.C., Stamenova, E.K., Bochkov, I.D., Rob-inson, J.T., Sanborn, A.L., Machol, I., Omer, A.D., Lander, E.S., and Aiden, E.L. (2014). A 3D map of the human genome at kilobase resolution reveals princi-ples of chromatin looping. Cell 159, 1665–1680.

Sanyal, A., Lajoie, B.R., Jain, G., and Dekker, J. (2012). The long-range inter-action landscape of gene promoters. Nature 489, 109–113.

Schoenfelder, S., Furlan-Magaril, M., Mifsud, B., Tavares-Cadete, F., Sugar, R., Javierre, B.M., Nagano, T., Katsman, Y., Sakthidevi, M., Wingett, S.W., et al. (2015). The pluripotent regulatory circuitry connecting promoters to their long-range interacting elements. Genome Res. 25, 582–597.

Seitan, V.C., Faure, A.J., Zhan, Y., McCord, R.P., Lajoie, B.R., Ing-Simmons, E., Lenhard, B., Giorgetti, L., Heard, E., Fisher, A.G., et al. (2013). Cohesin-based chromatin interactions enable regulated gene expression within preex-isting architectural compartments. Genome Res. 23, 2066–2077.

Sharov, A.A., Masui, S., Sharova, L.V., Piao, Y., Aiba, K., Matoba, R., Xin, L., Niwa, H., and Ko, M.S. (2008). Identification of Pou5f1, Sox2, and Nanog downstream target genes with statistical confidence by applying a novel algo-rithm to time course microarray and genome-wide chromatin immunoprecip-itation data. BMC Genomics 9, 269.

Silva, J., Nichols, J., Theunissen, T.W., Guo, G., van Oosten, A.L., Barrandon, O., Wray, J., Yamanaka, S., Chambers, I., and Smith, A. (2009). Nanog is the gateway to the pluripotent ground state. Cell 138, 722–737.

Stelloh, C., Reimer, M.H., Pulakanti, K., Blinka, S., Peterson, J., Pinello, L., Jia, S., Roumiantsev, S., Hessner, M.J., Milanovich, S., et al. (2016). The

cohesin-associated protein Wapal is required for proper Polycomb-mediated gene silencing. Epigenetics Chromatin 9, 14.

Wang, X., Arai, S., Song, X., Reichart, D., Du, K., Pascual, G., Tempst, P., Rosenfeld, M.G., Glass, C.K., and Kurokawa, R. (2008). Induced ncRNAs allo-sterically modify RNA-binding proteins in cis to inhibit transcription. Nature 454, 126–130.

Whyte, W.A., Orlando, D.A., Hnisz, D., Abraham, B.J., Lin, C.Y., Kagey, M.H., Rahl, P.B., Lee, T.I., and Young, R.A. (2013). Master transcription factors and

mediator establish super-enhancers at key cell identity genes. Cell 153, 307–319.

Yang, Y.W., Flynn, R.A., Chen, Y., Qu, K., Wan, B., Wang, K.C., Lei, M., and Chang, H.Y. (2014). Essential role of lncRNA binding for WDR5 maintenance of active chromatin and embryonic stem cell pluripotency. eLife 3, e02046.

Zhou, H.Y., Katsman, Y., Dhaliwal, N.K., Davidson, S., Macpherson, N.N., Sakthidevi, M., Collura, F., and Mitchell, J.A. (2014). A Sox2 distal enhancer cluster regulates embryonic stem cell differentiation potential. Genes Dev. 28, 2699–2711.