Tapping into yeast diversity


  • Justin C. Fay

    Corresponding author
    • Department of Genetics and Center for Genome Sciences and Systems Biology, Washington University, St. Louis, MO, USA
    Search for more papers by this author

Correspondence: Justin C. Fay, Fax: 314-362-2156; E-mail: jfay@genetics.wustl.edu


Domesticated organisms demonstrate our capacity to influence wild species but also provide us with the opportunity to understand rapid evolution in the context of substantially altered environments and novel selective pressures. Recent advances in genetics and genomics have brought unprecedented insights into the domestication of many organisms and have opened new avenues for further improvements to be made. Yet, our ability to engineer biological systems is not without limits; genetic manipulation is often quite difficult. The budding yeast, Saccharomyces cerevisiae, is not only one of the most powerful model organisms, but is also the premier producer of fermented foods and beverages around the globe. As a model system, it entertains a hefty workforce dedicated to deciphering its genome and the function it encodes at a rich mechanistic level. As a producer, it is used to make leavened bread, and dozens of different alcoholic beverages, such as beer and wine. Yet, applying the awesome power of yeast genetics to understanding its origins and evolution requires some knowledge of its wild ancestors and the environments from which they were derived. A number of surprisingly diverse lineages of S. cerevisiae from both primeval and secondary forests in China have been discovered by Wang and his colleagues. These lineages substantially expand our knowledge of wild yeast diversity and will be a boon to elucidating the ecology, evolution and domestication of this academic and industrial workhorse.

Saccharomyces species diversity

Yeast was discovered as the cause of fermentation by Louis Pasteur over 100 years ago (Pasteur 1866). Yet, it was not until recently that S. cerevisiae was clearly delineated from other Saccharomyces species through studies of reproductive isolation and application of the biological species concept (Naumov 1996). The Saccharomyces sensu structu complex is made up of eight known species, all capable of fermenting glucose to ethanol (Fig. 1). Saccharomyces species have been isolated from a diverse array of substrates, but most often from trees, bark, decaying leaves and soil. The most recently discovered species, S. arboricolus (Wang & Bai 2008), was isolated from oak and evergreen trees in China by the same authors as the study reported in this issue. S. eubayanus, the wild stock from which S. bayanus brewing strains were derived, was also recently found in forests in Patagonia (Libkind et al. 2011). However, S. cerevisiae is prevalent species in human-associated fermentations. How S. cerevisiae became so widely used in food and beverage production requires some knowledge of the wild populations from which it was derived.

Figure 1.

Saccharomyces species phylogeny. Species phylogeny is shown along with sources from which they have been isolated. Saccharomyces bayanus is listed in parenthesis to indicate it was derived from multiple hybridization events. S. pastorianus, an S. cerevisiae-S. eubayanus hybrid, is not shown.

Wild Saccharomyces cerevisiae diversity

Most of our knowledge of S. cerevisiae comes from strains present in the laboratory, vineyard or brewing environment. The identification of wild S. cerevisiae from oak trees in Siberia and North America (Naumov & Naumova 1991; Naumov et al. 1998) suggested that S. cerevisiae is not entirely a human commensal species. Subsequent population genetic studies showed that wild oak tree populations are differentiated from those associated with humans (Fay & Benavides 2005; Liti et al. 2009). However, with the exception of clinical isolates of S. cerevisiae from immuno-compromised patients (Muller et al. 2011), much of our understanding of population structure has come from strains associated with human fermentations (Fay & Benavides 2005; Legras et al. 2007; Liti et al. 2009).

To understand the structure of wild populations of S. cerevisiae, Wang et al. (2012) collected thousands of samples from diverse arboreal habitats across China, two of which are shown in Fig. 2. Tree samples were from fruit, bark, soil and rotten wood of primeval forests undisturbed by humans, secondary forests, planted orchards and urban trees in both tropical and temperate regions. Genetic analysis of 99 S. cerevisiae isolates revealed eight genetically distinct groups, five of which are basal to all previously defined groups, including that from North American oak trees. Interestingly, the three groups that fell within previously described populations were all isolated from secondary forests and orchards. While some amount of recombination can be inferred from genealogical incongruences among the 13 loci used, the three basal groups of strains obtained from primeval forests harbour the most diversity and have remained distinct. Consistent with this genetic differentiation, representatives of these groups exhibit partial reproductive isolation as measured by reduced spore viability between but not within groups.

Figure 2.

Arboreal Saccharomyces cerevisiae habitats. Primeval rainforest on Bawangling Mountain (left) and Wuzhi Mountain (right) in Hainan, China (Photo courtesy of Dr Niankai Zeng).

The extensive sampling combined with strong population genetic structure makes it possible to infer migration events and likely origins of some strains. Strains related to the Wine/European group were isolated from grapes and planted soil, and strains related to the Sake group were isolated from fruit and oak trees. There is also evidence for migration among arboreal habitats; one group contains strains predominantly isolated from primeval forests but also includes strains isolated from planted Chinese Chestnut and Chinese Plum trees. While it remains difficult to know where most domesticated strains originated from, it is noteworthy that the earliest evidence for wine production is from ceramic containers found in the Henan province in China dating to 9000 BP (McGovern et al. 2004). Although the modes of migration are not entirely clear, there is strong evidence for both human- (Goddard et al. 2010) and insect (Stefanini et al. 2012)-mediated migration.

At present, studies of S. cerevisiae phenotypic diversity have been dominated by genetically similar laboratory or domesticated yeast strains (but see Warringer et al. 2011). Domesticated strains are already known to have evolved a number of phenotypes related to non-arboreal environments, such as resistance to copper, sulphites and osmotic stress and their enological properties (Kvitek et al. 2008; Liti et al. 2009; Will et al. 2010; Hyma et al. 2011). Although the diversity of phenotypes present in arboreal population remains to be uncovered, their discovery will greatly expand our ability to apply the power of yeast genetics to ecological and evolutionary studies of wild S. cerevisiae.

J.C.F. is interested in domestication and diversification of yeasts.