Enhancer RNAs

Gene distal regulatory elements, primarily enhancers, are responsible for cell type- and state-specific gene expression, but the full molecular mechanisms underlying enhancer functions remain elusive. Recent evidences from genome-wide studies demonstrated that enhancers are pervasively transcribed to produce enhancer-derived long non-coding RNAs (lncRNAs), or eRNAs.¹ Several features of eRNAs make them similar yet distinguishable from other classes of lncRNAs, e.g. they are generated from genomic regions with high H3K4me1-to-me3 ratio of histone methylation, in majority display no polyadenylation or splicing, exhibit similar transcription rates compared to lncRNAs or coding mRNAs, but produce less stable transcripts, at least partially due to exosome activity.^1-5 Studies have suggested a high correlation between eRNA transcription and enhancer activation,^1-5 but the functional significance of eRNAs has been largely elusive.

Several recent studies provided evidence supporting functional roles of eRNAs. First, knock-down of eRNAs with RNAi or antisense-oligonucleotides resulted in decreased expression of the interrogated eRNAs themselves and their neighboring cognate coding genes.^2,3,6-8 Importantly, down-regulation of these eRNA transcripts appeared to cause minimal alternation of the local enhancer chromatin.^3,8 Furthermore, independent designs of reporter experiments, by overexpressing or tethering an exogenous eRNA to a promoter-containing reporter construct,^3,6,8 or by altering eRNA sequences² to manipulate the enhancer activity, supported functional importance of eRNA transcripts. One potential mechanism of eRNA-dependent enhancer activity in gene regulation is that eRNAs can interact with cohesin³ and Mediator⁸ complex to stabilize enhancer:promoter (E:P) looping (Figure 1). Considering the lack of trans localization of one interrogated eRNA,³ and the low abundance/stability of most eRNAs found in different systems,^1,3,4 most eRNAs likely function in cis. Yet the possibility of trans action of certain eRNAs cannot be excluded. Indeed, a recent finding found that the knockdown of an androgen-activated eRNA from KLK3 locus in prostate cancer cells not only inhibited its target gene KLK3, but also selectively impacted on a subset of androgen-regulated genes, some of which even reside on different chromosomes.⁸ In this regard, the role of eRNAs in stabilizing chromatin looping^3,8 provides an alternative interpretation of the classic definition of ‘in trans’ - eRNA may stay in the locus of its production but be capable of modulating cognate gene through long-distance genomic interactions. However, in another report, transcription elongation inhibitor flavopiridol diminished eRNA and coding gene expression but appeared to not affect E:P looping of the interrogated loci.⁵ Further experimentation is needed to clarify the relationship between eRNA transcription, its elongation and E:P looping .

Figure 1.

Functional roles of eRNA transcripts and enhancer transcription. Enhancer RNAs could functionally contribute to gene activation at least partially by modulating the stability of enhancer:promoter (E:P) looping via interacting with looping factors, including cohesin³ and Mediator,⁸ or could regulate the chromatin accessibility of its target promoter region.⁷ For some enhancer cohorts, eRNA transcription elongation (purple arrows) could have independent roles in deposition of histone methylation .

Another mechanism contributing to eRNA function appears to reflect modulation of chromatin accessibility of its target gene promoter (Fig. 1).⁷ During myogenesis, an eRNA from the core enhancer is needed for the chromatin accessibility and RNA polymerase II (Pol II) recruitment to the promoter and gene body of the cognate MYOD1 gene.⁷ Intriguingly, in the case of Arc gene in mouse neurons, on the contrary, gene promoter is required for its eRNA transcription, but not Pol II binding at the enhancer.¹ It seems consistent that transcription on enhancers and promoters is well correlated, but their inter-regulation is not fully understood. Furthermore, alteration of eRNAs can impact on biological processes including cell cycle⁶ and cell growth⁸ of cancer cells, or change gene expression in animal models in vivo,² suggesting a therapeutic potential.

With the rapid advance of epigenomics, it is now estimated that >400k potential enhancers exist in the human genome. Undoubtedly, the functions of eRNAs provide new mechanistic insights into the process of temporal and spatial gene regulation, enhancer activity and chromosomal interactions. Remaining questions about eRNAs include, but are not limited to: (1) further experimentation to fully understand the functions and underlying mechanisms of eRNAs in gene regulation and chromosomal interactions; (2) in cis versus in trans activities of eRNAs; (3) the distinctive and overlapping mechanisms underlying eRNA biogenesis and transcription regulation in comparison to other lncRNAs and protein coding genes; (4) the structural or sequence codes that enable certain eRNAs to play similar or distinctive roles, by themselves or together with protein partners; (5) additional eRNA functions besides those in transcription regulation; (6) the relevance of eRNAs to physiology and diseases. The answers to these questions will shed important new lights on the full understanding of the fundamental roles of enhancer elements in gene regulation, homeostasis and diseases.

Funding

The authors are supported by Department of Defense breast cancer postdoctoral fellowships (BC110381 to WL and BC103858 to DN) and National Institute of General Medical Science T32 (GM007198-37 and GM008666 to MTYL), respectively.