Skip to main content
NCBI home page
Transcription logoLink to Transcription
. 2017 Nov 3;9(2):123–130. doi: 10.1080/21541264.2017.1372044

Super enhancers – new analyses and perspectives on the low hanging fruit

Feda H Hamdan 1, Steven A Johnsen 1,✉
PMCID: PMC5834217  PMID: 28980882

ABSTRACT

Significant attention has recently been given to a class of enhancers termed “super enhancers”, while implying that “typical enhancers” are less important. In this report, we examine criteria for identification of super enhancers and address the need to evaluate the differences between BRD4-occupied “typical” and “super” enhancers.

KEYWORDS: BRD4, enhancer RNA, ERα activation, ROSE algorithm, super enhancers, transcriptional regulation, typical enhancers

Introduction

Bromo- and extra-terminal (BET) domain proteins, most notably BRD4, represent an important class of epigenetic regulators with a particularly strong relevance for various human diseases including cancer, fibrosis, heart failure, etc.1 BET proteins function as “epigenetic readers” which recognize acetylated lysine residues on both histone and non-histone proteins and serve to promote target gene transcription.2,3 A number of recent studies from our group and others have revealed a particular importance of BET proteins in the control of gene expression via distal enhancer regions. Notably, the discovery of small molecule inhibitors (BETi) which block the binding of BET proteins to chromatin4,5 have led to an explosion of research into the biology of these proteins and ultimately to numerous early phase clinical trials to test their efficacy in the treatment of various malignancies.6 However, a major hurdle that remains is the ability to accurately predict the biological and transcriptional effects of BETi based on transcriptional and/or genome-wide occupancy profiles. We have recently shown that BRD4 plays a significant role in the transcription of lineage-specific genes in human fetal osteoblasts, mainly through localization with different transcription factors at enhancer regions.7 Similarly, we showed that the majority of BRD4-enriched regions in mammary epithelial cells were associated with putative enhancers.8 Remarkably, more than one-third of epithelial-to-mesenchymal transition-related genes showed an enrichment of BRD4 at an adjacent distal region. Overall, enhancers emerge as a common and decisive mechanism in BRD4-mediated regulation of gene transcription.

Since the term “super enhancer” (SE) was coined,9,10 a significant emphasis has been placed on this class of enhancers while implying that other “typical enhancers” (TEs) are of less importance. However, the exact characteristics that render an enhancer “super” and not “typical” are still poorly defined and fairly arbitrary. In our previous work,7,8 we identified potential enhancers important for mediating tissue-specific BRD4 activity using a differential occupancy approach, independent of “super” classification. This work resulted in important insights into BRD4 function at putative distal enhancer regions, which could be functionally verified in both systems. In this report, we sought to examine the potential relevance of “super enhancers” by following different approaches to identify BRD4-dependent “super enhancers” with the goal of testing if they indeed represent a special class of regulatory elements with a particularly strong influence on gene regulation. Furthermore, we examined if we can more meticulously identify these presumed highly efficient regulatory units that can be optimally harvested to effectively predict the effects of their pharmacological perturbation on gene transcription.

Super vs. typical enhancers: What is so bad about being typical?

In general, SEs are considered to be large clusters of regulatory elements that are highly occupied by transcription factors and have high potential to activate transcription of their target genes.11,12 It was recently found that single nucleotide polymorphisms (SNPs) within BRD4-enriched SEs increase the chances of the development of breast and prostate cancers.13 BRD4 was also reported to occupy the Colon Cancer Associated Transcript 1 (CCAT1) super enhancer, thereby enhancing the expression of the nearby MYC oncogene and accounting for the anti-proliferative effects of BET inhibition in colorectal cancer.14 Cell lines that do not include a super enhancer driving MYC expression have been reported to be largely unresponsive to BET inhibition. Deleting the upstream super enhancer region of Myc in mice specifically affected its expression in tissues like the colon and prostate and led to a partial loss of the mammary tumorigenic phenotype in these mice.15 While these and other studies validate the increased interest in SEs, this might inadvertently lead to the disregard of important “typical” enhancers that can significantly mediate the effects of BRD4 and may efficiently be exploited in its perturbation.

In order to evaluate the individual contributions and importance of SEs and TEs, we used our previously published dataset (GSE55921/2) in which we examined the importance of BRD4 in controlling ERα-activated gene transcription in ER-positive MCF7 breast cancer cells.16 We specifically chose this system as it ideally suits our analyses due to the rapid and direct effect of ligand (estrogen) binding to the estrogen receptor-alpha (ERα) and the robust effects on transcriptional activation. Additionally, ERα activity in this system is highly dependent on BRD4 demonstrated by the observation that 83% (126 out of 152) of the expressed genes that are significantly upregulated by estradiol (E2) treatment (>2 folds, q-value<0.05) are significantly downregulated upon knockdown of BRD4 (<1.4 folds, q-value<0.05). We hypothesized that if SEs play a more prominent role in mediating BRD4 effects than TEs, the 126 genes downregulated upon knockdown of BRD4 will show a higher correlation with BRD4-driven “super” than “typical” enhancers.

In order to identify putative enhancers, we used genome-wide occupancy data for histone 3 acetylated on lysine 27 (H3K27ac) in E2-treated MCF7 (GSE40129)17 and identified SEs based on their BRD4 signal intensity using the Ranking of Super Enhancers (ROSE) algorithm,9,10 the most commonly utilized approach to identify SEs. In this way, we were able to identify 324 super enhancer regions (Fig. 1A) and 24,427 TEs. To selectively narrow enhancers to those related to ERα function, we intersected them with ERα peaks followed by intersection with BRD4 peaks to select only for TEs that are occupied by both ERα and BRD4 and eliminate any regions where a direct effect of estrogen and BRD4 could not be established. Subsequently, we extracted the nearest genes in a window of 500 kb using the BETA-minus version 1.0.0 on the Galaxy platform and utilized CTCF boundaries to filter the genes.

Figure 1.

Figure 1.

Open in a new tab

Different approaches followed to identify super enhancers using the ROSE algorithm. (A-D) Feeding ROSE with H3K27ac peaks and ranking regions based on BRD4 signal (A), only H3K27ac peaks that co-localize with or are adjacent to ERα (B), H3K27ac peaks that only co-localize with ERα (C) or H3K27ac peaks and ranking regions based on GRO-seq signal (D). In each panel, the specific workflow of the SE and TE identification is shown, followed by the ROSE output showing approximate ranks of enhancers associated with GREB1 and ESR1 as examples. Venn diagrams show the overlap of SE- and TE-associated genes with BRD4-dependent genes (126 genes that are upregulated by E2 treatment and significantly downregulated by BRD4 treatment). At the bottom of each panel, a box plot shows the general tendency of regulation of SE- or TE-associated genes upon knockdown of BRD4.

Interestingly, there were twice as many genes associated with TEs compared to SEs which were downregulated in response to BRD4-depletion (Fig. 1A, Venn diagram). This indicates that TEs likely play a highly significant role in BRD4-mediated gene transcription regulation and should not be overlooked. On the other hand, SE-associated genes included major players in estrogen response, including GREB1, TFF1, XPB1, in addition to the widely known BRD4 target, MYC. This may imply that master regulatory genes show a tendency to be more related to SE-mediated regulation, possibly due to more efficient control by a cluster of enhancers rather than by individual ones. In general, we failed to find significant differences between SE- and TE-associated genes in their tendencies for downregulation upon BRD4 knockdown (Fig. 1A, box plot).

Picking the petals of ROSE: An enhancer is super, an enhancer is not super…

The ROSE algorithm is currently the gold standard for the identification of super enhancers in addition to other scripts which follow the same rationale. Certain regions are used as an input for ROSE and stitched together if they are less than the default 12.5 kb apart. Subsequently, stitched regions are ranked according to the intensity of the signal of a chosen transcription factor or cofactor (e.g., Mediator, BRD4, etc.). Super enhancers are identified as those for which the signal is particularly high and surpasses a specific cut-off point. The regions that are close to the transcriptional start site can also be disregarded with a default threshold of 2500 bp. Thus, the process of the identification of super enhancers can be affected by different variables.

A deciding factor for the identification of super enhancers is the input regions that are inserted into ROSE. They form the general population among which SEs are selected. Naturally, the higher the number of the regions, the more “difficult” it is for a certain region to rise above the specific cut-off point and be classified as a SE. Super enhancers comprise a luring target not only due to their high activation potential but more importantly due to their specific and dependent mediation of gene regulation. As we were specifically interested in the potential role of BRD4-occupied super enhancers in the context of ERα function, we intersected our enhancer regions with ERα peaks after the identification of SEs by ROSE. To check the effect of limiting the regions before running ROSE, we used H3K27ac peaks that are co-localized or adjacent to ERα peaks as input for the algorithm. As expected, this approach reduced the number of SEs to 95 (Fig. 1B). Interestingly, as the processed region numbers were less than 20% of those used in the standard way, the cut-off point increased and ranks of estrogen-regulated genes improved, which could be expected since only estrogen-related regions were taken into account. Conversely, the number of SE-associated genes that are dependent on BRD4 decreased, while still including the same key ERα target genes defined previously (Fig. 1B, Venn diagram). Remarkably, this approach started to reveal a more significant dependence of SE-associated genes on BRD4 compared to TE-associated genes.

Overall, by inserting a simple focused “tuning” to ROSE, we were able to more precisely select for a predictive subgroup of enhancers. Ultimately, our goal is to identify subgroups of enhancers within a larger population of regions, which may more precisely predict the effects of BETi treatment. Altogether, the ranking of enhancers by ROSE is highly dependent on the number and nature of the regions which are inserted as input into the algorithm in addition to the stitching threshold. As such, the ranking should be used as a tool rather than a goal in identifying highly functional transcriptional regulatory subunits. This can support the rationale to limit the tested regions in the hope of identifying meaningful mediators regardless of the fact if their “super” rank is ultimate (from all possible regions) or regional (from specifically picked regions).

Super enhancer subcomponents: One for all or all for one?

While super enhancers are generally considered to comprise very broad H3K27ac peaks, chromatin accessibility can frequently provide an opportunity to narrow down the regions in order to decipher individual components of SEs. Usually, these enhancers include multiple peaks of chromatin accessibility (e.g. from DNase or ATAC-seq), where chromatin is accessible to transcription factors. This raised the question if these clusters work in an additive or synergistic manner.11 Recent developments in ground-breaking genome editing techniques have enabled scientists to investigate the contribution of individual subcomponents of certain SEs. Mosaic-seq uses CRISPR-mediated deletion of certain clusters of a super enhancer in order to evaluate their role in gene activation. This revolutionary method has revealed that in the case of the β-globin locus control region (LCR), one cluster within this super enhancer region is largely responsible for the activation of its target gene, HBG2.18 Conversely, deletion of individual regulatory elements of the α-globin SE in mice showed no significant preference of any of the SE subcomponents.19 Concordantly, the high activation potential of the SE associated with the mammary-specific Wap locus was partially dependent on each of the singular components.20

Accordingly, an important question arises of whether SEs comprising multiple components solely related to ERα can have greater effects on gene regulation than other SEs which include ERα in only one or a few of its components. For this purpose, we performed a highly biased analysis by using only H3K27ac peaks that intersect with ERα peaks as input for ROSE (Fig. 1C). This approach significantly decreased the cut-off point of SE identification to half of that compared to our previous approach when we allowed for all components independent of co-localization with ERα. Surprisingly, this increased the numbers of SEs to 117, but did not increase the number of BRD4-dependent SE-associated genes. SE-associated genes in this case also showed more significant downregulation by BRD4 knockdown (Fig. 1C, box plot).

To identify whether SEs affect more prominent regulators than TEs, we used the GREAT analysis tool and performed gene ontology analyses on the two nearest genes in a 1000 kb window to SEs and TEs.21 To correct for the large difference in numbers between the SE- and TE-associated genes, which may lead to a profound bias when calculating significance, we intersected the TE- and SE-associated peaks with DNase-seq from E2-treated MCF7 (GSE33216).22 Only SE regions defined in the most biased approach (Fig. 1C), where we intersected the H3K27ac peaks with ERα peaks, showed significant association with ontology terms related to estradiol, hormone, and estrogen response, while TE regions did not comprise any significant ontology terms related to estrogen or hormones. Both SE and TE regions failed to show meaningful gene associations using the other previous approaches (data not shown).

Overall, our observations imply that SEs can include subcategories depending on the transcription factors and niche of each subcomponent. Overall, not all SEs will necessarily follow the same rule. As we learn more about SEs, we will be able to better understand the impact of their various individual components and identify which are additive and which are synergistic. Currently, it is of high interest to test if the activation of any element of a specific SE will be sufficient to activate a gene that is otherwise not active. To do so, a driving transcriptional cofactor such as BRD4 or p300 can be tethered to specific regions using nuclease-deficient Cas9 (dCas9) in a system where the SE is not active [model in Fig. 2]. This approach has been successfully used to ectopically induce DNA methylation23 and induce acetylation of enhancer regions.24 Thus, using this approach will enable us to test the effects of targeting different transcriptional regulatory proteins to specific loci (e.g. in TE or SE) in order to derive conclusive data that will help in identifying the dependencies of enhancers and further validate their effects.

Figure 2.

Figure 2.

Open in a new tab

Tethering of BRD4 to enhancer regions via nuclease-deficient Cas9 (dCas9) can provide causative information about the contribution of individual components of (super) enhancers. (A) An inactive enhancer region is shown which displays no significant transcriptional activity. (B) In the case of an active super enhancer, different or identical transcription factors can lead to the recruitment of other activators, in this model p300-mediated acetylation of histones serves to promote recruitment of BRD4. This then leads to increased transcription of the target gene. (C) Tethering of BRD4 in an acetylation-independent manner to subcomponents of a super enhancer can validate the contribution of each element to the enhancer. By tethering BRD4 to one region, we can verify if one of the elements are sufficient to account for the high activation potential of the super enhancer (1+0+0 = 10). We can also evaluate the loss of the one component (0+1+1 = 5) and thus verify if this certain super enhancer works synergistically (1+1+1 = 10) or in an additive manner (1+1+1 = 3). Finally, tethering can also provide causative data to identify target genes of specific enhancers and determine the complex interactions between these enhancers and their target gene(s).

Enhancer RNAs: Whether bystanders or effectors, they are anyways a great help

Enhancers were recently shown to frequently produce bi-directional non-coding, generally short-lived transcripts referred to as enhancer RNAs (eRNA), which appear to be involved in inducing nearby target genes.25 It was reported that eRNAs near E2 activated genes are upregulated upon E2 treatment and function as stabilizers for the looping of enhancers with the promoters of ERα target genes.26 Conversely, a recent study has found that eRNAs rarely co-localize with an enhancer-promoter-loop associated with active transcription.27 Hah et al. showed that the majority of macrophage SEs in mice produced eRNAs in contrast to only one third of TEs.28 To date, it still remains unclear whether eRNAs are a mere by-product given their presence in the midst of transcriptional factories or actual effectors that are important central regulators of gene transcription.

In order to examine this further, we used publically available Global run-on (GRO-seq) (GSE43836)29 data for E2-treated MCF7 cells to test whether using this information to predict super enhancers results in similar findings to BRD4 ChIP-seq data. Remarkably, we were able to identify a similar number of SEs by ranking them based on GRO-seq, rather than BRD4 (Fig. 1D). Out of the 177 SEs that co-localize with ERα and BRD4, 106 enhancers were contained within the 218 genes identified using the BRD4 signal, suggesting there is a substantial, but only partial overlap in the results obtained using the two approaches. Notably, using GRO-seq to classify SEs increased the number of BRD4-dependent genes that are related to SEs and included other estrogen responsive genes like HSP80 and RAB31. However, it did not affect the tendency of downregulation following BRD4 depletion between SE- and TE-associated genes, which stayed insignificant when using all the H3K27ac regions (Fig. 1D, box plot). Whether eRNAs are bystanders or effectors, in any case, they can be highly effective predictors of enhancers and may possibly be used in place of the most commonly-used transcription factors like MED1 or BRD4 when identifying and analyzing enhancer regions.

Conclusion

In this report, we sought to challenge the current approach for identifying functionally important distal enhancer regions and make a simple comparison between SEs and TEs. We aimed to shed light on the distinctive roles of TEs and SEs since many researchers remain skeptical about this classification, at least in part due to the lack of a clear definition. Irrespective of categories, increasing our understanding of enhancer mechanisms of action and dependencies will not only provide us with novel approaches to predict the effects of epigenetic manipulation of enhancer function (e.g., through BETi), but will also endow us with unprecedented potential to identify and develop new strategies to specifically manipulate gene expression in different contexts and diseases. We observed that, irrespective of the approach for SE identification, a high number of genes showed regulation by TEs, which underscores the importance of studying the role of all enhancers rather than focusing only on the SE subcategory. While the frequent association of SE with master regulators may justify the recent interest in their characterization, SEs may simply be the “low hanging fruit” of enhancers that can prejudice our investigations and prevent us from identifying other important regulatory elements that can have significant impacts and effects. Instead, a differential occupancy approach as we recently described,7,8 may provide a more effective approach to identify functionally important enhancers and their underlying transcription factor networks.

Analysis

Fastq files were mapped to the hg19 genome using BOWTIE2/2.2.6 with very sensitive end-to-end options. Peaks were called using MACS2/2.1.0 without building the shifting model and with input peaks as background. Narrow peaks were called for ERα and broad peaks with a cut-off of 0.05 were called for BRD4 and H3K27ac. RNA-seq data were mapped by TOPHAT/2.1.0 to the hg19 genome and differential analysis was performed using CUFFLINKS/2.2.1. GRO-seq (SRR653425/6) was mapped using BOWTIE2/2.2.6 and bigwig files were generated using DEEPTOOLS/2.4.0 with ignoring the duplicates and extending for 200bp. Box plots were generated using GraphPad Prism 5 with significance calculated using the Mann-Whitney test.

Funding Statement

Deutsche Krebshilfe [111600]; German Ministry for Science and Education (BMBF) [01KU1401A]; Deutsche Krebshilfe (PiPAC Consortium) [70112505]; Deutsche Forschungsgemeinschaft (DFG) [JO 815/3-1].

Disclosure of potential conflicts of interest

No potential conflicts of interest were disclosed.

Acknowledgments

The authors would like to thank Xin Wang, Madhobi Sen, and Matthias Dobbelstein for valuable discussions.

References

  • 1.Ferri E, Petosa C, McKenna CE. Bromodomains: Structure, function and pharmacology of inhibition. Biochem Pharmacol. 2016;106:1–18. doi: 10.1016/j.bcp.2015.12.005. PMID:26707800 [DOI] [PubMed] [Google Scholar]
  • 2.Dhalluin C, Carlson JE, Zeng L, He C, Aggarwal AK, Zhou MM. Structure and ligand of a histone acetyltransferase bromodomain. Nature. 1999;399:491–6. doi: 10.1038/20974. PMID:10365964 [DOI] [PubMed] [Google Scholar]
  • 3.Xu Y, Vakoc CR. Targeting cancer cells with BET bromodomain inhibitors. Cold Spring Harb Perspect Med. 2017;7(7):a026674. doi: 10.1101/cshperspect.a026674. PMID:28213432 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 4.Nicodeme E, Jeffrey KL, Schaefer U, Beinke S, Dewell S, Chung CW, Chandwani R, Marazzi I, Wilson P, Coste H, et al. Suppression of inflammation by a synthetic histone mimic. Nature. 2010;468:1119–23. doi: 10.1038/nature09589. PMID:21068722 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5.Filippakopoulos P, Qi J, Picaud S, Shen Y, Smith WB, Fedorov O, Morse EM, Keates T, Hickman TT, Felletar I, et al. Selective inhibition of BET bromodomains. Nature. 2010;468:1067–73. doi: 10.1038/nature09504. PMID:20871596 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 6.Hessmann E, Johnsen SA, Siveke JT, Ellenrieder V. Epigenetic treatment of pancreatic cancer: Is there a therapeutic perspective on the horizon? Gut. 2017;66:168–79. doi: 10.1136/gutjnl-2016-312539. PMID:27811314 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Najafova Z, Tirado-Magallanes R, Subramaniam M, Hossan T, Schmidt G, Nagarajan S, Baumgart SJ, Mishra VK, Bedi U, Hesse E, et al. BRD4 localization to lineage-specific enhancers is associated with a distinct transcription factor repertoire. Nucleic Acids Res. 2017;45:127–41. doi: 10.1093/nar/gkw826. PMID:27651452 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Nagarajan S, Bedi U, Budida A, Hamdan FH, Mishra VK, Najafova Z, Xie W, Alawi M, Indenbirken D, Knapp S, et al. BRD4 promotes p63 and GRHL3 expression downstream of FOXO in mammary epithelial cells. Nucleic Acids Res. 2017;45:3130–45. doi: 10.1093/nar/gkw1276. PMID:27980063 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9.Loven J, Hoke HA, Lin CY, Lau A, Orlando DA, Vakoc CR, Bradner JE, Lee TI, Young RA. Selective inhibition of tumor oncogenes by disruption of super-enhancers. Cell. 2013;153:320–34. doi: 10.1016/j.cell.2013.03.036. PMID:23582323 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Whyte WA, Orlando DA, Hnisz D, Abraham BJ, Lin CY, Kagey MH, Rahl PB, Lee TI, Young RA. Master transcription factors and mediator establish super-enhancers at key cell identity genes. Cell. 2013;153:307–19. doi: 10.1016/j.cell.2013.03.035. PMID:23582322 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Pott S, Lieb JD. What are super-enhancers? Nat Genet. 2015;47:8–12. doi: 10.1038/ng.3167. PMID:25547603 [DOI] [PubMed] [Google Scholar]
  • 12.Evan GI, Hah N, Littlewood TD, Sodir NM, Campos T, Downes M, Evans RM. Re-engineering the pancreas tumor microenvironment: A “regenerative program” hacked. Clin Cancer Res. 2017;23(7):1647–55. doi: 10.1158/1078-0432.CCR-16-3275. PMID:28373363 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 13.Zuber V, Bettella F, Witoelar A, Andreassen OA, Mills IG, Urbanucci A. Bromodomain protein 4 discriminates tissue-specific super-enhancers containing disease-specific susceptibility loci in prostate and breast cancer. BMC Genomics. 2017;18:017–3620. doi: 10.1186/s12864-017-3620-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14.McCleland ML, Mesh K, Lorenzana E, Chopra VS, Segal E, Watanabe C, Haley B, Mayba O, Yaylaoglu M, Gnad F, et al. CCAT1 is an enhancer-templated RNA that predicts BET sensitivity in colorectal cancer. J Clin Invest. 2016;126:639–52. doi: 10.1172/JCI83265. PMID:26752646 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Dave K, Sur I, Yan J, Zhang J, Kaasinen E, Zhong F, Blaas L, Li X, Kharazi S, Gustafsson C, et al. Mice deficient of Myc super-enhancer region reveal differential control mechanism between normal and pathological growth. Elife. 2017;6:23382. doi: 10.7554/eLife.23382. PMID:28583252 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Nagarajan S, Hossan T, Alawi M, Najafova Z, Indenbirken D, Bedi U, Taipaleenmäki H, Ben-Batalla I, Scheller M, Loges S, et al. Bromodomain protein BRD4 is required for estrogen receptor-dependent enhancer activation and gene transcription. Cell Rep. 2014;8:460–9. doi: 10.1016/j.celrep.2014.06.016. PMID:25017071 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Theodorou V, Stark R, Menon S, Carroll JS. GATA3 acts upstream of FOXA1 in mediating ESR1 binding by shaping enhancer accessibility. Genome Res. 2013;23:12–22. doi: 10.1101/gr.139469.112. PMID:23172872 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Xie S, Duan J, Li B, Zhou P, Hon GC. Multiplexed engineering and analysis of combinatorial enhancer activity in single cells. Mol Cell. 2017;66:285–99. doi: 10.1016/j.molcel.2017.03.007. PMID:28416141 [DOI] [PubMed] [Google Scholar]
  • 19.Hay D, Hughes JR, Babbs C, Davies JO, Graham BJ, Hanssen LL, Kassouf MT, Marieke Oudelaar AM, Sharpe JA, Suciu MC, et al. Genetic dissection of the alpha-globin super-enhancer in vivo. Nat Genet. 2016;48:895–903. doi: 10.1038/ng.3605. PMID:27376235 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Shin HY, Willi M, Yoo KH, Zeng X, Wang C, Metser G, Hennighausen L. Hierarchy within the mammary STAT5-driven Wap super-enhancer. Nat Genet. 2016;48:904–11. doi: 10.1038/ng.3606. PMID:27376239 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.McLean CY, Bristor D, Hiller M, Clarke SL, Schaar BT, Lowe CB, Wenger AM, Bejerano G. GREAT improves functional interpretation of cis-regulatory regions. Nat Biotechnol. 2010;28:495–501. doi: 10.1038/nbt.1630. PMID:20436461 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 22.He HH, Meyer CA, Chen MW, Jordan VC, Brown M, Liu XS. Differential DNase I hypersensitivity reveals factor-dependent chromatin dynamics. Genome Res. 2012;22:1015–25. doi: 10.1101/gr.133280.111. PMID:22508765 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Xu X, Tao Y, Gao X, Zhang L, Li X, Zou W, Ruan K, Wang F, Xu GL, Hu R. A CRISPR-based approach for targeted DNA demethylation. Cell Discov. 2016;2:16009. doi: 10.1038/celldisc.2016.9. PMID:27462456 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Hilton IB, D'Ippolito AM, Vockley CM, Thakore PI, Crawford GE, Reddy TE, Gersbach CA. Epigenome editing by a CRISPR-Cas9-based acetyltransferase activates genes from promoters and enhancers. Nat Biotechnol. 2015;33:510–7. doi: 10.1038/nbt.3199. PMID:25849900 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Kim TK, Hemberg M, Gray JM, Costa AM, Bear DM, Wu J, Harmin DA, Laptewicz M, Barbara-Haley K, Kuersten S, et al. Widespread transcription at neuronal activity-regulated enhancers. Nature. 2010;465:182–7. doi: 10.1038/nature09033. PMID:20393465 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Li W, Notani D, Ma Q, Tanasa B, Nunez E, Chen AY, Merkurjev D, Zhang J, Ohgi K, Song X, et al. Functional roles of enhancer RNAs for oestrogen-dependent transcriptional activation. Nature. 2013;498:516–20. doi: 10.1038/nature12210. PMID:23728302 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 27.Rahman S, Zorca CE, Traboulsi T, Noutahi E, Krause MR, Mader S, Zenklusen D. Single-cell profiling reveals that eRNA accumulation at enhancer-promoter loops is not required to sustain transcription. Nucleic Acids Res. 2017;45:3017–30. doi: 10.1093/nar/gkw1220. PMID:27932455 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 28.Hah N, Benner C, Chong LW, Yu RT, Downes M, Evans RM. Inflammation-sensitive super enhancers form domains of coordinately regulated enhancer RNAs. Proc Natl Acad Sci U S A. 2015;112:6. doi: 10.1073/pnas.1424028112. PMID:25538304 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Hah N, Murakami S, Nagari A, Danko CG, Kraus WL. Enhancer transcripts mark active estrogen receptor binding sites. Genome Res. 2013;23:1210–23. doi: 10.1101/gr.152306.112. PMID:23636943 [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from Transcription are provided here courtesy of Taylor & Francis

RESOURCES