Ends-in vs. ends-out targeted insertion mutagenesis in Saccharomyces castellii

Gene replacement (knock-out) is a major tool for the analysis of gene function. However, the efficiency of correct targeting varies between species, and is dependent on the structure of the DNA construct. We analyzed the targeted insertion mutagenesis method in the budding yeast Saccharomyces castellii, phylogenetically positioned after the whole genome duplication event in the Saccharomyces lineage. We compared the targeting efficiency for target DNA constructs in the respective ends-in and ends-out form. For some of the constructs S. castellii showed a similar high degree of homologous recombination as S. cerevisiae. In agreement with S. cerevisiae, a higher targeting efficiency was seen for the diploid strain than for the haploid. Surprisingly, a higher degree of targeting efficiency was seen for ends-out constructs compared to ends-in constructs. This result may have been influenced by the difference in the length of the homologous target sequences used, although long homology regions of 300 bp-1 kb were used in all constructs. Remarkably, very short regions of cohesive heterologous sequences at the ends of the constructs highly stimulated random illegitimate integration, suggesting that the pathway of non-homologous end joining is highly active in S. castellii.