Thread by @NilsBirkholz

Very pleased to share our ✨New Paper✨ about an epigenetic conflict between a restriction–modification system and a methylation-dependent nuclease.

Want to know how that pans out? Continue reading here 🧵👇🏻 or head over to @NAR_Open for the full story: academic.oup.com/nar/advance-ar…

It all began when we started investigating the anti-#CRISPR repressor Aca2 from Pectobacterium #phage ZF40. Back then, we stumbled upon an RM system in Pecto that inhibited transformation and conjugation and completely blocked phage infection. 🧬🚫 [2/n]

Because this strain could only take up DNA with RM-compatible methylation, we knocked out the restriction endonuclease of the RM system, enabling us to pursue our anti-CRISPR regulation project:
academic.oup.com/nar/article/47…

And that could have been the end of it... [3/n]

...but when we removed the entire RM system, something curious happened: this ΔRM strain displayed a different defence phenotype, restricting DNA with the RM methylation pattern; in other words, the new defence acted like the polar opposite of the RM system! 🔄🤔 [4/n]

After a long search 🔎🧬, #RNAseq finally brought the revelation: a methylation-dependent nuclease was strongly upregulated in the ΔRM strain, and a knockout abolished the mysterious defence phenotype. But why had this gene been inactive in the presence of the RM system? [5/n]

Using reporter assays, we could show that the RM methyltransferase methylates and thus epigenetically shuts off 🚧 the promoter of the antagonistic endonuclease. Hence, the RM system not only provides strong protection but also silences a competing defence. [6/n]

Importantly, due to the antagonistic target specificities of the two nucleases (unmethylated DNA 🆚 DNA methylated by the RM system) we believe silencing is essential for their co-existence. But why and how would such a paradoxical situation arise in the first place? [7/n]

From genome comparisons we concluded that the RM system was mobile and acquired after the independent nuclease. The cell then had to maintain the RM system, or else the conflict would flare up and cause mayhem. A striking example of selfish behaviour by a defence system! [8/n]

Maybe the methylation-dependent nuclease was once meant to protect from parasitic RM systems ⚔️ by cleaving freshly methylated DNA and clearing any invaded cells from the population 🚮. But this particular RM system found a sneaky strategy to avoid that! [9/n]

And there might be other intriguing implications of this rivalry, including horizontal gene flow, pan immunity and the emergence of new evolutionary lineages. 🤯 [10/n]

Recently, a few cool papers have revealed defence system synergies 🛡️🤝🏻🛡️ – for example by the @MarraffiniLab (cell.com/molecular-cell…) and the @BlowerLab (academic.oup.com/nar/article/49…).

However, our study illustrates that defence systems can also be mortal enemies! 🛡️☠️🛡️ [11/n]

This has been a super fun project throughout my PhD and beyond, with a great team including @SimonJacksonNZ, @RobFagerlund and @PeterFineran 👨‍👨‍👦‍👦. Time to let go! 👋🏻🥲

Thanks to the four (oof) reviewers for their useful input and to @BioprotectionNZ and @otago for funding. [12/12]