This paper, expanding the field of toxin-antitoxin (TA) systems, identified the proteins encoded by a mystery plasmid in E coli O157:H7 named pOSAK1, as a TA system. This system consists of as the name specifies, a 'toxic' gene product and another gene encoding the 'antidote'. Usually the antidote is less stable and if lost the stable toxic gene acts quickly to kill or severely inhibit the growth of its host. Thus, the bacterial host is now addicted. Some functions of such systems may include growth polymorphisms (persistors, static growth) or anti-phage mechanisms.
One TA system RnlA/B, acts as an endoribonuclease actively degrading T4 phage mRNA. However, T4 bacteriophage has evolved to inhibit this process by encoding a protein Dmd, which negatively interacts with RnlA. The authors found that the mysterious plasmid pOSAK1 encodes genes with homology to rnlA/B, and they subsequently demonstrated that indeed, 2 of the 3 genes encoded by pOSAK1, LsoA/B make up a toxin-antitoxin system. Despite this genetic homology between these two systems, they do not interact with one another. Interestingly, the bacteriophage protein Dmd acts as an antitoxin which inhibits both toxins of each system, demonstrating a remarkable evolutionary feat (the first of its kind). Given that T4 replicates effectively in E coli K-12, which contains 36 TA systems, the authors speculate that this broad specificity of phage-encoded antitoxins may apply to many other systems, and possibly outside this T4-E coli model. Thus, it seems that in the arms race of bacteria and phage, the former is once again trying to follow the ladder.