Towards construction and validation of an ends-in recombination system in Escherichia coli

Homologous recombination is the primary DNA repair pathway in bacteria and it is immensely important in repairing DNA double strand breaks. Components of the homologous recombination pathway have been well conserved throughout evolution as an essential part of cell survival. Homologous recombination plays an important role in cellular processes like DNA repair as well as exchange of genetic information through chromosomal crossover. During homologous recombination, DNA strand exchange leads to formation of a heteroduplex joint between the invading and displaced DNA strands. This hetereoduplex joint is called a Holliday Junction. Resolution of the Holliday Junction proceeds via one of two pathways. In the presence of RuvC and/or RecG, Holliday Junction resolution proceeds via a “cut and paste” pathway where the invading DNA strand replaces a region of homologous DNA on the target DNA. In the absence of RuvC and RecG, Holliday Junction resolution takes place via a “copy and paste” pathway during which DNA synthesis needs to be primed at Holliday Junction intermediates formed during strand invasion. In an effort to separate this myriad of different requirements, I have attempted to develop a novel “ends-in” recombination assay system using E. coli as a model organism. This ends-in system would allow recombinant molecule formation by DNA synthesis of approximately 200 to 2000 bp size interval between the two converging ends of an invading linear dsDNA substrate oriented just like the greek letter Ù, but with the arms pointing inwards. In this study, a number of linear dsDNA assay templates were constructed and analyzed. All the constructs had two “arms” of homology to the chromosome pointing inwards i.e. in the ends-in orientation. Using this ends-in system, it was demonstrated that the presence of chi (Crossover Hotspot Initiator) sites was an important requirement for ends-in recombination in wild type E. coli cells. Our studies also showed that ends-in homologous recombination did not occur if chi sites were placed at or very near to the ends of the incoming linear dsDNA molecule, suggesting that the chi site recognition is efficient only if the incoming dsDNA has chi sites internal to the ends. Moreover, it was shown that neither RuvC nor RecG were required for successful recombinant product formation using the ends-in assay. This finding reinforces previous observations that suggest the idea that Holliday Junctions can be resolved independent of both RuvC and RecG