Non-mammalian c-integrases are encoded by giant transposable elements.

In a recent report in Trends in Genetics, Gao and Voytas [1] described a new family of integrase genes that were identified in diverse eukaryotic species, including slime mold, Caenorhabditis elegans, C. briggsae, zebrafish, Takifugu rubripes, Xiphophorus maculates, cow, dog and humans. These genes potentially encode proteins containing w400 amino acids with homology to retroviral integrases and transposases, including the integrase-like proteins encoded by Tlr, a family of atypical mobile elements with long terminal inverted repeats (TIRs) from the ciliate Tetrahymena thermophila [2,3]. Despite the similarity of c-integrases to the Tlr integrases and their presence in multiple copies in some genomes, Gao and Voytas found no evidence linking the c-integrase genes to retroelements [1]. The authors did not exclude the possibility that the c-integrases might be part of an unusual type of mobile element, but instead proposed that the c-integrases are ‘host’ genes. Two observations supported this hypothesis: (i) an excess of synonymous to nonsynonymous substitutions among c-integrase genes, indicative of purifying selection, and; (ii) the distant relationship of c-integrases to Fob1p, a protein from Saccharomyces cerevisiae involved in rDNA metabolism. Two observations prompted us to investigate the origin of c-integrases. First, we noticed that the ESTs encoding D. discoideum c-integrases reported by Gao and Voytas displayed O99% nucleotide identity with the currently active mobile element Tdd-4 (GenBank accession number U57081). Tdd-4 elements essentially consist of the integrase gene flanked by w125-bp TIRs, a structure reminiscent of DNA transposons [4]. Second, we were intrigued that they found two distinct pairs of c-integrases from C. elegans to be part of larger duplicated genomic regions flanked by large inverted-repeats (IRs; see Figure 3 in Ref. [1]). In light of the relationship betweenD. discoideum c-integrases and theTIR-containing Tdd-4 transposons, the presence of IRs flanking the C. elegans c-integrase genes might indicate that these genes are part of larger mobile elements. A hallmark of mobile-element transposition is the duplication of a short genomic sequence at the site of chromosomal integration [5]. These target-site duplications (TSDs) result from the staggered cleavage mediated by integrases and transposases. We examined each flank of the three IR pairs associated with the C. elegans c-integrase genes and found that in all three cases, the IRs were immediately flanked by a 6-bp direct