Opening the Chromatin by eRNAs

Abstract

Enhancer RNAs (eRNAs) have emerged as an important component of transcriptional activation. In this issue, Mousavi et al. (2013) have uncovered a critical role for eRNAs in regulation of myogenic differentiation program through increasing chromatin accessibility at MyoD and MyoG loci.

Temporal and tissue-specific expression of key protein-coding genes is orchestrated by distal regulatory elements known as transcriptional enhancers. These DNA elements are often decorated with unique chromatin signatures, including monomethyl histone H3 lysine 4 (H3K4me1) and acetylated H3K27, and display stimulus-dependent recruitment of transcriptional coactivators such as p300/CBP. However, the revelations that enhancers are occupied by RNA polymerase II (RNAPII) and are the sites of synthesis of long noncoding RNAs came as a complete surprise (De Santa et al., 2010; Kim et al., 2010). Furthermore, it was shown that the production of such enhancer-derived transcripts (termed eRNAs) correlated with the activation of the neighboring protein-coding genes (De Santa et al., 2010; Kim et al., 2010). Concomitantly, studies aimed at assessing the function of long noncoding RNAs in mammalian cells uncovered a role for a class of extragenic long noncoding RNAs in activation of their neighboring protein-coding genes reminiscent of the functional enhancers (Ørom et al., 2010; Wang et al., 2011).

A number of recent studies have started to shine light on the generality of eRNA’s function from a variety of stimulusregulated enhancers and the possible molecular basis of their action by promoting enhancer-promoter interactions (Hah et al., 2013; Lam et al., 2013; Li et al., 2013; Melo et al., 2013). The study by Mousavi et al. (2013) has extended the role of eRNAs to the differentiation of the skeletal muscles by myogenic regulatory factors. The authors determined the genome-wide chromatin residence of MyoD and MyoG using mouse C2C12 skeletal muscle cells. While MyoD is expressed in both proliferative myoblasts (MB) and terminally differentiated myotubes (MT), MyoG expression is only observed at the onset of differentiation in MT. Importantly, there was a significant (~77%) overlap between MyoD and MyoG occupancy in MT, and a large portion (~50%) of these MyoD/MyoG sites mapped to extragenic regions. The authors determined that a fraction of these sites (~20%) contain RNAPII, the chromatin signatures resembling transcriptional enhancers and express eRNAs as determined by RNA-sequencing of steady-state levels of cellular RNA. These eRNAs were manifested as sense and/or antisense eRNAs extending ~1–2 kb from MyoD/MyoG sites. Interestingly, the expression of these eRNAs was preferentially regulated by MyoD, as depletion of MyoG did not significantly alter their expression. While the eRNAs described by Mousavi et al. (2013) resembled the previously described eRNAs (Kim et al., 2010), it is becoming evident that the best methodology for assessing eRNAs requires the capture of pioneering rounds of transcription using techniques such as global run-on sequencing (Gro-seq), since eRNAs may not be as stable as messenger RNAs (Hah et al., 2013). Moreover, the status of eRNA 3′ end processing and whether they are predominantly monoexonic or undergo splicing is not clear. Unfortunately, Gro-seq does not allow for analysis of polyadenylation or splicing, and therefore one needs to use other RNA sequencing techniques in conjunction with Gro-seq to examine the mature forms of eRNAs in detail.

Next the authors chose to focus on a ~24 kb genomic fragment upstream of MYOD1, which was shown to contain a regulatory landscape controlling spatiotemporal expression of MyoD during embryogenesis. Analysis of RNA-sequencing data revealed a prevalence of eRNAs throughout this region in C2C12 MT as well as muscle progenitors. Using multiple small interfering RNAs (siRNAs) against noncoding RNAs expressed from this region, the authors uncovered an enhancer function for eRNAs expressed from core enhancer (CE) in regulation of MyoD gene expression, located approximately 20 kb from the MyoD promoter. In contrast, depletion of the eRNAs from distal regulatory region (DRR) did not significantly impact MyoD expression. However, surprisingly, depletion of eRNAs corresponding to the DRR region diminished MyoG expression located in a different locus in late-differentiating cells and hindered the myogenic differentiation program. This effect was deemed to be mediated in trans since overexpression of a fragment (1.2 kb and 2.0 kb) of eRNA corresponding to DRR activated MyoG expression as well as the cascade of myogenic gene regulatory network in the absence of increased MyoD expression.

To uncover the underlying molecular basis of eRNAs expressed from CE or DRR sites, the authors initially examined the possible involvement of the Cohesin complex in eRNA-mediated activation of MyoD. Depletion of eRNAs corresponding to the CE region did not affect the chromatin residence of the Cohesin subunits at either CE or MyoD promoter. However, reduction of this eRNA significantly diminished RNAPII occupancy at MyoD locus, consistent with decreased levels of transcription. Similarly, depletion of eRNA corresponding to the DRR, which regulated MyoG in trans, decreased RNAPII chromatin residence at MyoG. Surprisingly, diminished levels of eRNA corresponding to the DRR region also decreased MyoD occupancy at MyoG locus without any effect on the MyoD gene itself. Finally, the authors assessed whether eRNAs corresponding to CE of DRR regulate the chromatin accessibility at the MyoD or MyoG loci. Indeed, depletion of eRNA corresponding to DRR resulted in decreased DNaseI sensitivity at MyoG promoter, while depletion of CE eRNA decreased chromatin accessibility at both MyoD and MyoG promoters.

Taken together, the authors have uncovered a number of eRNAs expressed during the myogenic differentiation program mapping to MyoD and MyoG extragenic binding sites in mammalian cells. Their detailed characterization of two of these eRNAs corresponding to the proximal and distal sites of MyoD regulatory regions revealed differences in their targets and their possible mode of action. While the distal eRNA (CE) regulated a neighboring MyoD gene presumably in a cis-mediated manner (Figure 1), the proximal eRNA (DRR) unexpectedly acted in trans to regulate MyoG gene at a different locus. However, the authors found that both eRNAs were not only critical for increasing chromatin accessibility but also for the recruitment of RNAPII to their respective targeted promoters (Figure 1). These findings suggest an interesting mechanism by which eRNAs, through recruitment of critical transcription factors and/or chromatin remodeling complexes, allow for a permissive chromatin environment in which preinitiation complexes containing RNAPII and basal transcription factors could operate.

Figure 1. eRNAs Expressed from Core Enhancer Element Lead to Chromatin Opening and Transcription at MyoD Locus.

eRNAs are expressed from the core enhancer element of the MyoD regulatory region following a differentiating stimuli. CE-expressed eRNAs function by recruiting chromatin remodeling factors to provide chromatin accessibility leading to increased RNA polymerase II occupancy.

While we have begun to scratch the surface in understanding eRNAs and their mechanism of action, many questions remain unanswered. Some of these questions include: What are the factors required for eRNA synthesis and their 3^′ end processing? Why are some eRNAs bidirectional while others unidirectional? Are there common sequences or structural features among eRNAs that could be used to discern their function? Does misregulation of eRNAs result in development of human diseases? Without a doubt, these are exciting times to unravel new and unanticipated layers of regulation for tissue- and temporal-specific gene expression.