The fate of linear DNA in Saccharomyces cerevisiae and Candida glabrata: the role of homologous and non-homologous end joining.

In vivo assembly of plasmids has become an increasingly used process, as high throughput studies in molecular biology seek to examine gene function. In this study, we investigated the plasmid construction technique called gap repair cloning (GRC) in two closely related species of yeast - Saccharomyces cerevisiae and Candida glabrata. GRC utilizes homologous recombination (HR) activity to join a linear vector and a linear piece of DNA that contains base pair homology. We demonstrate that a minimum of 20 bp of homology on each side of the linear DNA is required for GRC to occur with at least 10% efficiency. Between the two species, we determine that S. cerevisiae is slightly more efficient at performing GRC. GRC is less efficient in rad52 deletion mutants, which are defective in HR in both species. In dnl4 deletion mutants, which perform less non-homologous end joining (NHEJ), the frequency of GRC increases in C. glabrata, whereas GRC frequency only minimally increases in S. cerevisiae, suggesting that NHEJ is more prevalent in C. glabrata. Our studies allow for a model of the fate of linear DNA when transformed into yeast cells. This model is not the same for both species. Most significantly, during GRC, C. glabrata performs NHEJ activity at a detectable rate (>5%), while S. cerevisiae does not. Our model suggests that S. cerevisiae is more efficient at HR because NHEJ is less prevalent than in C. glabrata. This work demonstrates the determinants for GRC and that while C. glabrata has a lower efficiency of GRC, this species still provides a viable option for GRC.