Yeast flocculation is a community-building cell aggregation trait that is a key contributor to brewing but a nuisance in other industrial processes, the laboratory, and clinical settings. This dichotomy demonstrates the importance of understanding and manipulating the genetic determinants of cell aggregation to achieve phenotypically desirable outcomes on both ends of the spectrum. Experimental evolution in the laboratory provides a unique testing ground for understanding the breadth of genetic mechanisms Saccharomyces cerevisiae uses to achieve cell aggregation. The popular lab strain S288C does not aggregate due to a nonsense mutation in the gene encoding transcription factor Flo8, which regulates flocculation. Despite this mutation, these strains evolve the ability to aggregate during 35% of our continuous culture evolution experiments. These evolved strains present an opportunity to examine the adaptive routes that lead to cell aggregation and whether one or many routes are favored.
We have analyzed 23 evolved clones using a combination of whole genome sequencing, quantitative phenotyping, and Bulk Segregant Analysis (BSA), a pooling and sequencing approach. All of the clones exhibit a significantly stronger aggregation phenotype than the ancestor, and none of the strains reverted the point mutation in FLO8. We recovered two loss of function mutations in ACE2, which has been previously shown in the literature to cause mother-daughter separation defects. BSA revealed causal mutations in TUP1, a known regulator of flocculation gene FLO1, and ROX3, which has not been previously connected to flocculation in the literature. Most interestingly, we discovered a common mutation among 13 of the evolved clones: a Ty1 transposable element insertion in the promoter region of the flocculation gene FLO1. Discovering this common adaptive mechanism indicated that rather than many equally favored adaptive routes, changing the regulation of FLO1 is the preferred route for evolving flocculation. We engineered a flo1 knockout strain and evolved it comparatively against a wild-type lab strain, with 32 replicates of each. We determined that this single gene deletion reduced flocculation occurrences from 25% to 3% over 250 generations. Deleting FLO1 had no effect on the evolution of unrelated phenotypes (separation defects and wall sticking), demonstrating the efficacy of this evolve-sequence-design approach for engineering targeted phenotypic outcomes.