We’ve got our mind on methylation and methylation in our mind, but we are often left wondering how DNA methylation got there in the first place. Well, that cognitive tongue twister has now been cracked by a class of RNA that likes to interact with DNA methyltransferases.
Extra coding RNAs (ecRNAs) originate from the sense strand of protein coding genes; however, they are distinct from mRNA as they go beyond the gene boundaries and are non-polyadenylated. Previous work by the lab of Daniel Tenen at Harvard has shown that ecRNAs can bind the maintenance methyltransferase DNMT1 and prevent methylation of their locus of origin. Now, the Lab of Jeremy Day from the University of Alabama at Birmingham brings forth their neuroepigenetic expertise to reveal that ecRNAs target DNA methylation in the brain as well.
Genome-Wide ecRNA Expression is Ubiquitous
First, the team characterized genome-wide patterns in embryonic rat cortical neuronal cultures:
- A directional RNA-seq workflow was used to analyze both polyadenylated and non-polyadenylated transcripts.
- This revealed ubiquitous ecRNA expression from protein coding genes, which is highly correlated with gene-specific mRNA expression.
- Methyl-Binding Domain (MBD)-seq uncovered that high ecRNA expression from a locus is not only correlated with increased mRNA expression, but also decreased DNA methylation in that gene’s promoter.
- Incubation of the cells with chemicals that alter neuronal activity uncovered that a small set of mRNA expression is altered by neuronal activity.
- Genes with mRNA expression regulated by neuronal activity also produce ecRNAs regulated by the same neuronal activity.
- They confirmed these findings by RT-qPCR.
- In terms of functionally, the activity-regulated ecRNAs belong to genes involved in neuronal processes (including related disorders) and the immune system.
ecRNA Biogenesis Lets Fos Remember by Binding DNMT3a and Regulating DNA Methylation
With the activity-dependent regulation of ecRNAs established, the team then focused in on the activity-regulated Fos gene to dig deeper into ecRNA biogenesis:
- Time course analysis of mRNA and ecRNA induction by RT-qPCR revealed that the ecRNA is synthesized before the mRNA.
- This finding and other experimentation with receptor agonists and different RNA polymerase inhibitors suggest that ecRNA is not a simple by-product of regular gene expression: it undergoes a unique biogenesis.
- RNA immunoprecipitation (RIP) with DNA methyltransferase antibodies (DNMT1 and DNMT3a) revealed that Fos ecRNA was enriched, while Fos mRNA was not.
- The Fos ecRNA and DNMT3a were investigated in further detail using in vitro assays and synthetic fluorescently labeled RNA and DNA probes based on the ecRNA to show that it directly binds to DNMT3a.
- Next, they designed anti-sense oligonucleotides (ASOs) to specifically target the Fos ecRNA.
- Both MeDIP-qPCR and bisulfite-seq revealed increased methylation at regulatory elements.
- There was also a progressive decline in Fos mRNA levels.
- Finally, since Fos is upregulated during memory formation, they went in vivo. By surgically infusing ASOs into the hippocampus of the rat brain and using behavioral testing (contextual fear conditioning) they showed that Fos ecRNA is required for long-term memory formation.
Activity Induced Priming of the Neuroepigenetic Landscape
Day shares, “In this paper, we show that neuronal activation can induce transcription of ecRNAs from a number of genes. For example, ecRNA from the Fos gene is induced by several different forms of neuronal stimulation, and this induction obeys different temporal dynamics and utilizes distinct transcriptional machinery as compared to the mRNA product from the same gene. We demonstrate that these ecRNAs can bind with high affinity to DNMT3a and block methylation of the corresponding genomic DNA.”
“We think that activity-dependent changes in ecRNA may be a way for neurons to alter DNA methylation at certain genes based on activity patterns, possibly to prime the response of that gene to future stimulation.”
ecRNA and DNA Interactions: RNA-DNA Hybrids?
Day also discussed the relationship of the ecRNAs to recent results from the lab of Frédéric Chédin at the University of California Davis. Day elaborates, “The idea that ecRNAs might utilize RNA-DNA hybrid structures such as R-loops to achieve genetic specificity is a very intriguing idea. As revealed in Sanz et al. (2016 Mol. Cell), R-loops are enriched at gene promoters and end sites, which by definition overlaps with the extra-coding sites that we described.”
“Critically, the ability of an ecRNA to be specific to the “parent” genomic DNA would seem to require some active anchoring mechanism and/or a very short transcript half-life. R-loop formation would be consistent with both of these features.”
“Further, the finding that R-loop formation at promoters is inversely correlated with DNA methylation would be consistent with our observations that ecRNA presence is associated with lower levels of promoter DNA methylation, and our findings that ecRNA knockdown results in gene silencing via methylation.”
The Future of ecRNA
Day concludes with his outlook, “I think a key step will be to characterize the mechanism that confers ecRNA specificity – this is still a big unknown. However, it will also be interesting to further understand the processes that regulate ecRNA synthesis and degradation, and continue to explore the functional role of ecRNA at different sites.”
“It has not escaped our attention that CRISPR–guided targeting of ecRNAs (as employed by John Rinn and others) will allow further insight into the functional contribution of ecRNAs. Finally, it is possible that ecRNAs may also bind other proteins that have conventionally been thought to only bind DNA. Future work by our lab and other labs will continue to explore how these interactions contribute to normal and disease function in neuronal systems.”
Go induce the rest of your neuroepigenome over at Nature Communications, July 2016