A lot of things just don’t mix well, dogs and cats, oil and water, and as most geneticists would agree, transposons and DNA methylation. DNA methylation is the mortal enemy of transposon activity; across the tree of life it has a conserved role silencing transposable elements from plants, to algae, to vertebrates. This suggests DNA methylation may have originally evolved in eukaryotes to silence transposable elements. But a recent study suggests that DNA methylation was sought after by some retrotransposons in distantly related eukaryotes. DNA methyltransferases (DNMTs) are ancient, some going back to the last common ancestor. In most species, DNMTs and transposons were engaged in an evolutionary arms race, with transposons finding ways to escape methylation. However, some species have a very odd relationship between the two which challenges our assumptions about this arms race.
Ryan Lister’s lab from The University of Western Australia followed up on a peculiar finding regarding transposable elements and DNMTs. Previous studies showed that DNMTs were counterintuitively linked to reverse transcriptase domains in certain eukaryotes. This may represent a novel mechanism by which transposons escape DNA methylation, or actually use it to their benefit. To interrogate this possible mechanism, the talented team sought out to characterise how cytosine-specific DNMTs have been incorporated into distinct classes of retrotransposons independently across evolution. They used eukaryotic species encoding a high number of DNMTs: the dinoflagellates (phylum including marine plankton species) and charophytes (division including green algae species). The talented team used RNA-seq, whole genome sequencing, and MethylC-seq to annotate and characterize these species genomes. Here’s what they found:
- RNA-seq uncovered that a large number of less characterized Dnmts (Dnmt5, Dnmt6, and Dntm2) are the only expressed DNMT transcripts across dinoflagellates
- These strange transcripts also happen to only contain a DNMT domain
- RNA-seq guided genome annotation revealed that the dinoflagellate species Symbiodinium kawagutii has hundreds of evolutionarily old, and active Dnmts in retrotransposons.
- Using MethylC-seq, they found that Symbiodinium also has a distinct epigenome among algae and plants, with widespread non-specific hemi-methylated CpGs
- This may suggest a complex, co-adaptive evolutionary relationship
- CpG methylation in Symbiodinium is not correlated with gene expression, and is present at active promoters
- Charophytes independently evolved the incorporation of DNMTs into the coding regions of very different retrotransposon classes
- Using methylation induction experiments, the authors cloned and expressed the transposon DNMTs to reveal that the retrotransposon DNMTs are active and can de novo methylate CpGs, suggesting that retrotransposons could self-methylate retrotranscribed DNA
This study shows that retrotransposons recruited host DNMT genes in two distantly related eukaryotes. The nearly ubiquitous CpG methylation in Symbiodinium could explain why the functions of DNA methylation is mostly lost in this lineage. The function of the retrotransposon DNMTs is still unclear, but it may be involved in methylation of the retrotranscribed cDNA. This methylated cDNA would then resemble the highly methylated host genome which might allow it to escape being recognised as exogenous. Overall, these findings challenge our long-held view of transposons and DNA methylation as opponents. Their relationship may be much more complex and could change our view of the evolution and function of the epigenome as a whole. Even cats and dogs can get along … if it’s evolutionarily advantageous.
Read more about this evolutionary cease-fire at Nature Communications, April 2018.