Transposons to toxins: the provenance, architecture and diversification of a widespread class of eukaryotic effectors

Abstract

Enzymatic effectors targeting nucleic acids, proteins and other cellular components are the mainstay of conflicts across life forms. Using comparative genomics we identify a large class of eukaryotic proteins, which include effectors from oomycetes, fungi and other parasites. The majority of these proteins have a characteristic domain architecture with one of several N-terminal 'Header' domains, which are predicted to play a role in trafficking of these effectors, including a novel version of the Ubiquitin fold. The Headers are followed by one or more diverse C-terminal domains, such as restriction endonuclease (REase), protein kinase, HNH endonuclease, LK-nuclease (a RNase) and multiple distinct peptidase domains, which are predicted to carry their toxicity determinants. The most common types of these proteins appear to have originated from prokaryotic transposases (e.g. TN7 and Mu) and combine a CDC6/ORC1-STAND clade NTPase domain with a C-terminal REase domain. Other than the so-called Crinkler effectors of oomycetes and fungi, these effectors are encoded by other eukaryotic parasites such as trypanosomatids (the RHS proteins) and the rhizarian Plasmodiophora, and symbionts like Capsaspora Remarkably, we also find these proteins in free-living eukaryotes, including several viridiplantae, fungi, amoebozoans and animals. These versions might either still be transposons or function in other poorly understood eukaryote-specific inter-organismal and inter-genomic conflicts. These include the Medea1 selfish element of Tribolium that spreads via post-zygotic killing. We present a unified mechanism for the recombination-dependent diversification and action of this widespread class of molecular weaponry deployed across diverse conflicts ranging from parasitic to free-living forms.

Figure 1.

(A) Domain architecture network of eukaryotic CR proteins. Domains linked in the same polypeptide are connected by arrows, with the arrow head pointing to the C-terminal domain. The arrow thickness reflects the number of associations found in the total dataset. (B–M) Representative domain architectures of CR proteins in different eukaryotic species. Proteins are labelled by the species abbreviation of organisms in which they are found followed by the NCBI Genbank id (gi). Species abbreviations are provided in the figure and also available in the ‘Materials and Methods’.

Figure 2.

(A) Multiple sequence alignment of various eukaryotic CR-NTPase families and related AAA+ ATPases. Protein sequences are labelled by their species abbreviation followed by the NCBI gis. For species abbreviations refer to the Supplementary data. (B) Topology diagram depicting the conserved core of the STAND-CDC6/ORC1-like AAA+ ATPases highlighting family-specific and overall features. (C) 3-D cartoon of representative CDC6 structure (pdb: 2QBY) illustrating its DNA-contacting interface. (D) Representative structure of HEH domains depicting their DNA-binding modes. (E) Phylogenetic tree depicting the inter-relationships between various members of the STAND-CDC6/ORC1-like AAA+ATPases.

Figure 3.

Multiple sequence alignment and predicted topologies for widely present CR-REase families. The conserved catalytic site residues are highlighted in both the alignment and the topology figure. The equivalence of core secondary elements (α1–β1–β2–β3–α2) between these families is also illustrated. Protein sequences are labelled by their species abbreviation followed by the NCBI gis. For species abbreviations refer to the ‘Materials and Methods’.

Figure 4.

(A–B) Multiple sequence alignment of CR-HNH nuclease families. (C) Multiple sequence alignment and topology of CR-ubiquitin-like (CR-Ubl) Header domains. (D–F) Multiple sequence alignment of other lineage-specific Header domains including VP-NTD, Mnag-NTD and Caps-NTD domains. Protein sequences are labelled by their species abbreviation followed by the NCBI gis. For species abbreviations refer to the Supplementary data.

Figure 5.

(A–F) Phylogenetic trees illustrating LSEs of eukaryotic CR protein domains and specific gene transfers from bacterial homologs (indicated by the curved arrows). LSEs are shown as coloured triangles/sectors in the tree. Bootstrap values are shown for the major branches only. The bacterial branches are coloured black. The complete trees from which these were derived can be retrieved from the Supplementary data. For species abbreviations refer to ‘Materials and Methods’. (G) Positional entropy comparison between CR-NTD and CR toxin domains. (H) Entropy plot for CR-Ubl 1+ CR-REase 2 type proteins in P. infestans.