In addition to protein-coding messenger RNAs, our cells produce a plethora of diverse non-coding RNA molecules. Many of these are generated from sequences that are distant from genes, and include regulatory DNA sequences called enhancers. Transcription factors bound at enhancers are thought to regulate gene expression by looping towards genes in 3D space. The potential functions of non-coding enhancer RNAs (eRNAs) in this process have been avidly debated, but there has been a tendency to write them off as accidentally transcribed by-products of enhancer–gene interactions. After all, how could short, unstable, heterogeneous RNAs have a role in gene regulation? Writing in Cell, Bose et al. reveal that these eRNAs can indeed be functional, when produced in proximity to the enzyme CBP.
Many proteins that modulate gene expression have RNA-binding surfaces— this includes DNA-binding transcriptional regulators, as well as epigenetic factors, which add or remove molecular modifications to DNA and other regulatory proteins. However, there are few examples of these RNA–protein interactions modulating an enzyme's catalytic activity. One exception is the interaction between RNA and the enzyme PRC2, which deposits repressive methyl groups on histone proteins, around which DNA is packaged in a structure called chromatin. Nascent transcription prevents PRC2 from depositing methyl groups onto histones in the vicinity, helping to maintain gene activity. Bose and colleagues now show that this phenomenon of local regulation by non-coding RNA might be general, affecting transcriptional co-activator proteins as well as repressors such as PRC2.
One transcriptional co-activator is the acetyltransferase enzyme CBP, which, along with its close relative p300, associates with DNA in enhancer regions, where it adds acetyl groups to histones and transcription factors. This acetylation promotes the recruitment of numerous transcriptional co-activators and chromatin-remodelling proteins that have acetyl-binding regions, along with the RNA-synthesizing enzyme polymerase II (Pol II). The authors began by isolating and sequencing thousands of RNA molecules that interact with CBP in vivo. Most CBP-bound RNAs were derived from intergenic regions and non-coding intron sequences that lie within genes, rather than protein-coding exon sequences. Moreover, the RNAs typically originated from DNA regions to which CBP binds, including enhancers.
Indicating that many of the CBP-bound RNAs are eRNAs, these transcripts seemed to be generally unstable. The researchers detected low or undetectable levels of these molecules on examination of all the RNA molecules present in the cell, but a prominent proportion in an assay of all nascent RNAs. Moreover, CBP-bound RNA levels correlated with both nascent RNA levels and CBP-binding intensity in a given region, suggesting that interactions occur on chromatin between CBP and newly transcribed RNAs.
Bose et al. went on to define the primary region in CBP that interacts with RNA as a disordered, basic loop within the protein's acetyltransferase domain. This region had previously been defined as an auto-inhibitory domain whose action could be relieved by regulated auto-acetylation. This suggested that neutralization of positive charges in the basic loop — by either acetyl groups or nucleic acids, both of which are negatively charged — could stimulate the acetyltransferase activity of CBP. The authors found that various RNAs bound CBP with low affinity and little sequence specificity, consistent with a promiscuous RNA-binding site, as observed for PRC2. Strikingly, RNA binding to this domain stimulated acetyltransferase activity in vitro, supporting the idea that RNA binding relieves auto-inhibition by the basic loop and regulates acetylation levels.
Of note, there was previous evidence that the catalytic activity of CBP and p300 is regulated at enhancers. Histone acetylation by these proteins is a hallmark of active enhancers that are communicating with their target genes to stimulate transcription. But CBP and p300 are also found associated with inactive enhancers, which lack histone acetylation. Accordingly, although CBP and p300 are bound at most enhancers, their levels on DNA do not effectively predict the expression of target genes. By contrast, when focusing only on enhancers that generated CBP-bound eRNAs, Bose et al. found that the depletion of CBP did reduce the expression of selected target genes. It will be fascinating to probe this effect further, and to discover whether the presence of CBP-bound RNAs is an improved determinant of where CBP is actively modulating gene transcription.
The authors' work suggests a mechanism for region-specific control of CBP, resolving how CBP can be bound throughout the genome, but active only at a subset of sites. A similar finding has been reported for another defining characteristic of enhancers, methylation of the amino-acid residue lysine 4 on histone H3, which also depends on the onset of eRNA synthesis. Interestingly, Bose et al. found that depletion of CBP did not have consistent effects on eRNA production, implying that eRNA transcription occurs early in enhancer activation, often independently of CBP activity and histone acetylation levels. Thus, acetylation by CBP may be more important for maintaining enhancer activity than for initiating the process.
A model is emerging in which transcription is itself an early step in enhancer activation. Pol II is recruited by transcription factors and maintains opens chromatin. Once the enzyme begins to transcribe, the nascent eRNA it produces stimulates co-activator proteins such as CBP in the region in a sequence- and stability-independent manner. The activities of these proteins promote the recruitment of more transcription factors, Pol II and chromatin-remodelling proteins, enabling full enhancer activation. In addition, Pol II itself can serve as a vehicle for attracting chromatin-modifying enzymes that spread more molecular marks associated with chromatin activation across the transcribed region. In this manner, transcription of enhancers can generate a positive-feedback loop that stabilizes both enhancer activity and gene-expression profiles.
Overall, the current study fundamentally changes the discourse around eRNA functions, by demonstrating that these RNAs can have major, locus-specific roles in enhancer activity that do not require a particular RNA-sequence context or abundance. Furthermore, by providing strong evidence that CBP interacts with eRNAs as they are being transcribed, this study highlights the value of investigating nascent RNAs for understanding enhancer activity.