Understanding SCRaMbLEd GenomesView all posters
Johns Hopkins University, United States
As part of the Synthetic Yeast Genome Project (Sc2.0), an inducible evolutionary system (SCRaMbLE) was developed using a synthetic yeast chromosome arm, synIXR, by inserting recombinase target sites in specific regions, according to predefined design principles. By inducing a chemically regulated form of Cre recombinase, cells undergo structural variations (inversions, deletions, duplications) generating potentially complex populations with many distinct derivatives (SCRaMbLEotypes) through in vivo scrambling. By sequencing SCRaMbLEd genomes, it will be possible to elucidate evolutionary trajectories compatible with viability. Identifying and quantifying recombination events and determining the novel genomic sequence generated by scrambling poses challenging computational problems. We performed in-depth sequence evaluation of >70 genomes subjected to SCRaMbLE using a custom analysis pipeline to detect structural variations, quantify copy number variations, and perform sequence reconstruction. The strains had for the most part been selected for loss of function in one of two genes. The pipeline is efficient and scalable for massive sequencing analysis of SCRaMbLEd strains. We identified 491 novel junctions that resulted from scrambling; under the conditions used, scrambled derivatives of synIXR had on average approximately 2 deletions and 1 inversion per synIXR chromosome, along with smaller numbers of duplications and complex rearrangements that are difficult to characterize. Because all strains were selected for loss of one gene, which only occurred via deletion, the data are consistent with equal frequencies of deletion and inversion, as predicted by the theory and confirmed by our statistical model. Our analysis shows that rearrangements occurred only at loxP sites, with no preference. Also, no ectopic rearrangements were observed in synIXR or in the remaining 99% non-synthetic genome. SCRaMbLE is an effective mechanism for generating genome diversity in a predefined genomic region, and is expected to be scalable to a fully synthetic genome designed by the same principles as synIXR.