Identification of the endonuclease domain encoded by R2 and other site-specific, non-long terminal repeat retrotransposable elements

The non-long terminal repeat (LTR) retrotransposon, R2, encodes a sequence-specific endonuclease responsible for its insertion at a unique site in the 28S rRNA genes of arthropods. Although most non-LTR retrotransposons encode an apurinic-like endonuclease upstream of a common reverse transcriptase domain, R2 and many other site-specific non-LTR elements do not (CRE1 and 2, SLACS, CZAR, Dong, R4). Sequence comparison of these site-specific elements has revealed that the region downstream of their reverse transcriptase domain is conserved and shares sequence features with various prokaryotic restriction endonucleases. In particular, these non-LTR elements have a Lys/Arg-Pro-Asp-X12-14aa-Asp/Glu motif known to lie near the scissile phosphodiester bonds in the protein-DNA complexes of restriction enzymes. Site-directed mutagenesis of the R2 protein was used to provide evidence that this motif is also part of the active site of the endonuclease encoded by this element. Mutations of this motif eliminate both DNA-cleavage activities of the R2 protein: first-strand cleavage in which the exposed 3' end is used to prime reverse transcription of the RNA template and second-strand cleavage, which occurs after reverse transcription. The general organization of the R2 protein appears similar to the type IIS restriction enzyme, FokI, in which specific DNA binding is controlled by a separate domain located amino terminal to the cleavage domain. Previous phylogenetic analysis of their reverse transcriptase domains has indicated that the non-LTR elements identified here as containing restriction-like endonucleases are the oldest lineages of non-LTR elements, suggesting a scenario for the evolution of non-LTR elements.