Insights into the Dekkera bruxellensis Genomic Landscape: Comparative Genomics Reveals Variations in Ploidy and Nutrient Utilisation Potential amongst Wine Isolates

Abstract

The yeast Dekkera bruxellensis is a major contaminant of industrial fermentations, such as those used for the production of biofuel and wine, where it outlasts and, under some conditions, outcompetes the major industrial yeast Saccharomyces cerevisiae. In order to investigate the level of inter-strain variation that is present within this economically important species, the genomes of four diverse D. bruxellensis isolates were compared. While each of the four strains was shown to contain a core diploid genome, which is clearly sufficient for survival, two of the four isolates have a third haploid complement of chromosomes. The sequences of these additional haploid genomes were both highly divergent from those comprising the diploid core and divergent between the two triploid strains. Similar to examples in the Saccharomyces spp. clade, where some allotriploids have arisen on the basis of enhanced ability to survive a range of environmental conditions, it is likely these strains are products of two independent hybridisation events that may have involved multiple species or distinct sub-species of Dekkera. Interestingly these triploid strains represent the vast majority (92%) of isolates from across the Australian wine industry, suggesting that the additional set of chromosomes may confer a selective advantage in winery environments that has resulted in these hybrid strains all-but replacing their diploid counterparts in Australian winery settings. In addition to the apparent inter-specific hybridisation events, chromosomal aberrations such as strain-specific insertions and deletions and loss-of-heterozygosity by gene conversion were also commonplace. While these events are likely to have affected many phenotypes across these strains, we have been able to link a specific deletion to the inability to utilise nitrate by some strains of D. bruxellensis, a phenotype that may have direct impacts in the ability for these strains to compete with S. cerevisiae.

Author Summary

The yeast D. bruxellensis is of great importance in biofuel and fermented beverage industries, largely as a contaminant and/or spoilage organism. Its lifestyle is not unlike that of the wine/brewing/baking yeast S. cerevisiae, with independent evolutionary pathways having led to this convergence; these species are phylogenetically very distant. Unlike S. cerevisiae, D. bruxellensis is highly intractable in the laboratory; it is difficult to mate and to transform, making even the most basic genetic analysis very difficult. Thus we still have a great deal to learn about this economically important yeast. The latest gene sequencing technologies are, however, providing a means of addressing these limitations. The current manuscript describes a comparative genomics approach to providing insights into inter-strain variations that shape the genomic landscape of D. bruxellensis. Like other industrial yeasts, it has a diploid core genome, but there are also triploid isolates which possess the core diploid complement with an additional, more distantly related, full set of chromosomes. Evidence presented in this paper suggests that this form of triploidy has arisen more than once in the evolutionary history of D. bruxellensis, and it confers a selective advantage for strains of this yeast isolated from wineries.