A novel immunity system for bacterial nucleic acid degrading toxins and its recruitment in various eukaryotic and DNA viral systems

Zhang, Dapeng; Iyer, Lakshminarayan M.; Aravind, L.

In this article we present the results of such a strategy that helped us uncover and characterize a remarkable, diverse class of nuclease toxins, whose immunity appears to depend primarily on a protein superfamily prototyped by the Saccharomyces cerevisiae protein Smi1/Knr4. The Smi1/Knr4 protein was first recovered in a screen for S. cerevisiae mutants that confer resistance to the killer toxin produced by the competing yeast species Hansenula mrakii (16, 17) . Smi1/Knr4 was shown to physically interact with the tyrosyl tRNA synthetase and it appears to functionally interact with the non-ribosomal peptide ligase Dit1, with a tRNA-synthetase-like catalytic domain, in the efficient synthesis of dityrosine a peptide metabolite that is typical of fungal spore-walls (18) . Interestingly, it also shows synthetic lethal and physical interactions with a great number of proteins (19) . Nevertheless, its exact significance and biochemical action has remained poorly understood (20) . Parallel studies recovered other Smi1/Knr4 eukaryotic homologs namely FBXO3, a subunit of a SCF-type E3 ubiquitin ligase in vertebrates (21) , and PGs2, a subunit of the tubulin polyglutamylase, which is a non-ribosomal peptide-ligase that links multiple glutamates to the g-carboxyl group of target proteins (22, 23) . Exploratory sequence surveys suggested that Smi1/Knr4 homologs are also abundantly represented in bacteria (Smi1/Knr4 domain, Pfam: PF09346). Furthermore, our preliminary contextual analysis of conserved gene neighborhoods of these representatives suggested that they might be functionally linked to potential nucleases. Very recently, a novel contact-dependent inhibitory (CDI) toxin system has been reported in proteobacteria that delivers multiple nuclease toxins into target cells (24, 25) . Our observations indicated that Smi1/Knr4 homologs are potential immunity proteins in a subset of these CDI systems. Together, these observations prompted us to systematically investigate both the bacterial and eukaryotic Smi1/Knr4 homologs and explore their potential connection to nuclease toxins, their delivery and immunity against them. As a result we were able to identify a diverse group of previously unknown nuclease toxins and immunity proteins that are present across all the major bacterial lineages with considerable significance for intra-specific and host interactions. This investigation also allowed us to uncover diverse, previously unknown nuclease and deaminase domains in bacterial toxins and predict their folds and biochemical mechanisms. We also show that the Smi1/Knr4 homologs, which were ultimately derived from bacterial toxin-immunity systems, have been recruited by eukaryotic double-stranded DNA viruses to perform multiple roles in intracellular survival and morphogenesis of these viruses. Finally, we present evidence that the ability of the conserved domain in the Smi1/Knr4 superfamily of proteins to bind structurally diverse protein partners has been re-used in eukaryotes as a means to recruit targets to peptide-modifying systems such as the ubiquitin and the polyglutamylase systems.