The road less traveled: A new phosphorothioate antiviral defense mechanism discovered in Archaea

The story of phosphorothioate modifications of DNA begins over 50 years ago and has unfolded like Frost's walk in the woods [1]. In 1970, Fritz Eckstein synthesized oligonucleotides in which a non-bridging oxygen in the phosphodiester backbone was replaced by sulfur – the first phosphorothioate-modified nucleic acid [2]. Twenty years passed until Zixin Deng and colleagues discovered a DNA degradation phenotype in bacteria possessing a 5-gene dnd cluster, with electrophoretic degradation caused by incorporation of sulfur into bacterial genomes [3]. These roads of chemistry and biology met again another 20 years later with the discovery by Wang et al. that Dnd proteins insert sulfur as a phosphorothioate in the genomes of diverse bacteria [4]. This walk in the woods now picked up speed.

As an epigenetic mark, the earliest functional role for phosphorothioates was identified in restriction-modification “immune surveillance”, with modification regulated by Dnd proteins ABCDE and DndFGH performing the restriction cleavage of unmarked foreign DNA [5]. But the dnd cluster did not behave like a typical methylation-based restriction-modification system, with modification of all consensus sites in a genome. Advancements in next generation sequencing facilitated genomic mapping of phosphorothioates, which revealed that only ∼10–15% of short consensus sequences (e.g., C_PSCA, G_PSTTC/G_PSAAC) were modified, with frequent partial modification at bistranded sites on the genome [6]. Beyond restriction-modification, new functions have emerged for phosphorothioates in regulating gene expression, cellular metabolism, and oxidative stress response [7,8].

With the walk moving at a rapid pace, the story of phosphorothioate DNA modifications recently took a sharp turn in the discovery by Xiong et al. of a completely new phosphorothioate-based anti-viral defense system in Archaea [9]. Archaeal microbiology is not as well studied as that of bacteria, which led Xiong et al. to explore phosphorothioate modifications on a road less traveled. And this road made all the difference. Using a combination of genomics and mass spectrometry, Xiong et al. initially found phosphorothioate modifications and functional dndCDEA gene homologs in four archaea. It was not surprising that they also found homologs of the archaeal dndC and dndD genes in 2600 bacteria and 42 other archaeal genomes, and that short consensus sequences were only partially modified with phosphorothioates. What was surprising, however, was the discovery of a restriction system completely different than the previously recognized dndF-I genes. The new “phosphorothioate-blocked DNA exclusion” gene cluster, pbeABCD, did indeed restrict infection by bacteriophage lacking phosphorothioate modifications, but, remarkably, the pbeABCD system did not result in the expected DNA cleavage associated with typical restriction-modification systems and caused by dndF-I (Fig. 1) [10]. Instead, PbeABCD functioned to prevent viral DNA replication. While the biochemical mechanism of PbeABCD is a mystery, the fact that pbeABCD genes were found to be widespread among both bacteria and archaea and sometimes combined with dndF-I restriction genes points to an unappreciated complexity of bacteriophage defense mechanisms. By taking the road less traveled, Xiong et al. [9] have changed how we think about bacterial epigenetics. It is exciting to wonder where this new road will lead.

Fig. 1.

PT modification and related defense systems in Bacteria and Archaea.