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.

Despite the low average pairwise sequence similarity across this superfamily, all representatives are known or predicted to possess a similar core fold comprising of four conserved helices and six strands ( Figure 1 , Supplementary Data). Strands 1 and 2 form a b-hairpin and the strands 3-6 form a 4-stranded b-meander; however, the b-hairpin and the b-meander show only limited or no hydrogen-bonding along their length, despite being spatially beside each other. Thus, the structural core of the SUKH domain can be described as a split b-sheet with only weak interaction between its two parts. This structural peculiarity could potentially be critical for the functional interactions of the domain (see below). Based on sequence-similarity-based clustering and phylogenetic analysis five major groups can be recognized within the SUKH domain superfamily (Figure 1 , Supplementary Data). The first of these, and the most widespread, is the one typified by Smi1/Knr4, FBXO3, SKIP16, PGs2 and YobK (that entirely includes the PFAM model PF09346, 'SMI1/KNR4 family', and additional proteins not detected by that model within it) and is seen in both bacteria and eukaryotes. This ensemble, which we term Smi1-like or SUKH-1 group includes the majority of the SUKH domains. We term the second group, prototyped by Syd, the Syd-like or SUKH-2 group. This group is largely restricted to the gammaproteobacteria and firmicutes. The SUKH-3 group prototyped by CA_C3700 (gi: 15896931) is widely distributed across most bacterial lineages. The group prototyped by SGR_4389 (gi: 182438182), the SUKH-4 group, is again seen in several bacteria and sporadically in fungi. The SUKH-5 or US22-like group is present in fowl adenoviruses, various vertebrate iridoviruses, archosaur poxviruses (Crocodilepox virus and Fowlpox virus), and in multiple copies in several herpesviruses (representatives of the alphaherpesvirus, betaherpesvirus and alloherpesvirus clades). Members of this group are also encoded by genomes of the early-branching chordate Branchiostoma, the salmon, the frog Rana catesbeiana and the duckbilled platypus, where they appear to have been acquired from the genomes of integrated herpesviruses (46) . Phylogenetic analysis of each group, along with the phyletic patterns, strongly suggests that SUKH domain proteins have been widely disseminated both within and across the superkingdoms via extensive lateral transfer (Supplementary Data). In light of this pattern, the near complete absence of this superfamily in archaea suggests that there could be certain specific functional barriers that prevent acquisition of the SUKH domain by that superkingdom. Phylogenetic analysis strongly suggests that the groups SUKH-2-5 are monophyletic clades. The largest group, SUKH-1 is likely to represent the ancestral group from within which the above clades have diversified through rapid sequence divergence.