Site-specific recombination in human cells catalyzed by phage lambda integrase mutants.

Phage lambda Integrase (Int) is the prototype of the so-called integrase family of conservative site-specific recombinases, which includes Cre and FLP. The natural function of Int is to execute integration and excision of the phage into and out of the Escherichia coli genome, respectively. In contrast to Cre and FLP, however, wild-type Int requires accessory proteins and DNA supercoiling of target sites to catalyze recombination. Here, we show that two mutant Int proteins, Int-h (E174 K) and its derivative Int-h/218 (E174 K/E218 K), which do not require accessory factors, are proficient to perform intramolecular integrative and excisive recombination in co-transfection assays inside human cells. Intramolecular integrative recombination is also detectable by Southern analysis in human reporter cell lines harboring target sites attB and attP as stable genomic sequences. Recombination by wild-type Int, however, is not detectable by this method. The latter result implies that eukaryotic co-factors, which could functionally replace the prokaryotic ones normally required for wild-type Int, are most likely not present in human cells.

RNA splicing of bacterial genes in eukaryotes.

The presence of intervening sequences or introns in eukaryotic genes has been known for more than 20 years, and the mechanisms underlying RNA splicing have been studied in depth both genetically and biochemically. In recent years, however, an increasing number of bacterial genes have been introduced into higher eukaryotes as important tools for genetic studies. Their gene products are frequently used as an indirect measure for cell type-specific promoter activity, as, for example, in the case of chloramphenicol acetyl transferase (CAT assay) or beta-galactosidase. Here we show that RNA splicing of two prokaryotic genes encoding site-specific DNA recombinases occurs in eukaryotic cells. In one case, splicing is only observed after treatment of cells with the cytokine alpha interferon. We further demonstrate that mutating an intragenic donor splice site in a bacterial gene apparently activates a second, alternative splicing pathway. In conjunction with previous reports, our findings should also be regarded as a warning that splicing of bacterial genes in higher eukaryotes is a more common phenomenon than presently recognized, which may be difficult to overcome and may cause problems in the interpretation of experimental results.

The resolvase encoded by Xanthomonas campestris transposable element ISXc5 constitutes a new subfamily closely related to DNA invertases.

BACKGROUND: Conservative site-specific recombination is responsible for the resolution of cointegrates which result during the transposition of class II transposable elements. Resolution is catalysed by a transposon-encoded recombinase, resolvase, that belongs to a large family of recombinases, including DNA invertases. Resolvases and the related invertases are likely to employ similar reaction mechanisms during recombination. There are important differences, however. Resolvases require two accessory DNA binding sites within each of the two directly repeated recombination sites. Invertases instead need a host factor, Fis, and an enhancer type DNA sequence, in addition to two inversely orientated recombination sites. RESULTS: The resolvase encoded by transposable element ISXc5 from the gram-negative phytopathogen Xanthomonas campestris shows two features which distinguish it from other known resolvases. First, it is more closely phylogenetically related to invertases than other resolvases. In particular, two functionally important regions seem highly conserved between this resolvase and members of the invertase subfamily. Second, the enzyme exhibits a large extension of its carboxy-terminal domain with unknown function. We purified ISXc5 resolvase and analysed its resolution reaction in vitro. Our biochemical and DNA topological analysis reveals that critical features of resolution are similar, if not identical, to that carried out by gammadelta resolvase. However, despite its apparent similarity to invertases, we were unable to detect recombination on standard substrates for DNA inversion, in either the presence or absence of Fis. CONCLUSIONS: ISXc5 resolvase employs a reaction mechanism which is common to members of the resolvase family. Its position near the evolutionary borderline to invertases and its high degree of identity within two functionally important regions with members of the DNA invertase subfamily suggest that only a few replacements of critical residues may suffice to convert this resolvase into a functional, possibly Fis-dependent invertase.