4. Genetic Diversity and Population Structure of S. cerevisiae
The approximately 9000 year domestication history of yeast
[33] is similar to that of key plants and animals, which usually have a domestication history of around 10,000 years
[34]. The domestication of plants and animals has been extensively studied since Darwin
[34][35], however, research centering on the domestication of yeast has rarely been performed until recently. The lag was partially due to the lack of reference wild populations of
S. cerevisiae and poor understanding about the natural history of the yeast. The phylogenetic distinction between wild and domesticated populations of
S. cerevisiae was shown for the first time in 2005
[36] based on sequence analysis of five genes (
CCA1,
CYT1,
MLS1,
PDR10, and
ZDS2) and their promoters in 81 strains.
The
S. cerevisiae strains employed in the early studies of population genetics and genomics were mainly from fermentation and human-associated environments, and wild strains were poorly represented. The wild strains designated in these studies were mainly from vineyards, oak tree bark and associated soil. Though the oak strains were considered “truly wild” in these studies, the association of the oak strains with human activities cannot be excluded, because the oak trees sampled were usually located in man-made environments or environments frequently visited by humans, such as parks or arboreta
[10][37].
5. Origin of the Domesticated Population of S. cerevisiae
Previous studies generally support the China/Far East Asia origin hypothesis of
S. cerevisiae. Ancient basal lineages of
S. cerevisiae have not been found outside China, despite extensive survey in Europe
[38][39], North America
[10][40], South America (including Amazonian rainforests)
[41], New Zealand
[42][43][44] and Africa
[13][45][46]. However, the origin of the domesticated population of
S. cerevisiae is still a debated issue
[47]. Basically, two hypotheses have been proposed: (1) Chinese or Asian wild
S. cerevisiae strains immigrated to other regions and were then domesticated independently in different areas
[48][49]; or (2) after a single ancestral domestication event occurring most likely in China or Asia, domesticated ancestors were later introduced to other regions
[12][13].
A population genomics study mainly on ale beer and wine yeasts showed that present industrial yeasts originated from only a limited number of ancestors
[50], but the ancestors were not specified. Other studies
[48][49] revealed close relationships of different domesticated lineages with different wild relatives of
S. cerevisiae, suggesting that multiple independent domestication events led to the origin of various domesticated lineages. This multiple domestication events scenario was also supported by additional studies based on different strain and data sets
[36][51][44][52][53]. However, the closest local wild relatives of individual domesticated lineages have not been specified, except for the wine lineage.
6. Intrinsically Different Life Strategies of the Wild and Domesticated Populations of S. cerevisiae
Previous studies have shown that
S. cerevisiae occurs in both natural and man-made environments with high genetic diversity and clear population structure. However, different studies resulted in different answers to a fundamental question of whether the diversity of
S. cerevisiae is primarily driven by niche adaptation and selection, or neutral genetic drift, echoing the long standing selectionist vs. neutralist debate in evolutionary biology. Some studies show that
S. cerevisiae strains are principally organized by geography, highlighting the role of genetic drift in shaping the population structure of
S. cerevisiae [48][40][44][53], while others recognize mainly ecologically defined populations, suggesting that natural selection may play a more important role than geographic factors in the diversification of
S. cerevisiae [36][37][54][55][56]. Recent studies suggest that the forces driving the evolution of
S. cerevisiae are more complicated, and neither geographic nor ecologic factors can fully explain the population structure of the species. Different levels of divergence and different lineages may have resulted from different driving forces.
In general, ecology seems to be the primary force driving the evolution of
S. cerevisiae, since the wild and domesticated populations are distinct phylogenetically and the domesticated population is apparently an outcome of natural and artificial selection for adaptation to nutrient- or sugar-rich environments
[12][13][36]. Extensive adaptive genome variations, including different patterns in heterozygosity, SNPs, gene contents and copy numbers, and allele distributions have been observed between wild and domesticated populations
[12][13][49], suggesting that wild and domesticated populations have evolved different life strategies for adaptation to generally different environments.
7. The Diversity of the Domesticated S. cerevisiae Is Primarily Driven by Ecology
Ecology apparently plays a main role in the divergence of the domesticated lineages of
S. cerevisiae. Osmolarity seems to be the primary selection pressure, since strains associated with liquid- and solid-state fermentation are clearly separated
[12][13]. The main difference between the two types of fermentation is the water content of the substrates. The water contents are usually 80–90% and 40–60% in the liquid- and solid-state fermentation, respectively
[13]. Within each of the LSF and SSF groups, strains associated with different food and beverage fermentation usually form distinct lineages. Remarkably, strains for the fermentation of grape juice, wort, milk, agave juice and honey, cluster in the Wine, Beer, Milk/Cheese, Mexican Agave and African Honey Wine lineages in the LSF group, respectively; while strains for the fermentation of dough, sorghum grain, barley grain, and cooked rice form the Mantou, Baijiu, Qingkejiu, and Huangjiu/Sake lineages in the SSF groups, respectively, regardless of their geographic origins
(Figure 4) [12][13][49][57].
Extensive genetic variations leading to consequent phenotypic trait changes for adaptation to specific niches have been identified in different domesticated lineages
[12][13][58][59][60]. Three unique HGT fragments (regions A–C) from
Zygosaccharomyces bailii were identified from wine yeast strains
[61]. These regions harbor key functional genes in wine fermentation and thus are believed to contribute to the adaptation of wine yeast strains to grape juice fermentation. Genes in these regions have also been found from other lineages, but are mostly limited in the LSF group
[12][13].
8. The Diversification of the Wild S. cerevisiae Is Largely Consistent with a Neutral Model
The genetic diversity of the whole species
S. cerevisiae is mainly contributed by its wild population, which is clearly structured with highly diverged lineages
[11][12][13][14]. Broadly, geography seems to play a main role in the diversification of the wild strains. Strains from forests in different countries or regions usually form different lineages, such as the North American Oak, Far East Russia, Ecuador, and Malaysian lineages
[13][49]. The
S. cerevisiae strains from Amazon forests in Brazil also formed different lineages
[41]. Within China, the primeval forest strains from south China are generally not mixed with those from north China
[11][12][13]. The forest strains from different regions (Shaanxi and Beijing) in north China also form different lineages (CHN-II and CHN-IV, respectively). However, the role of ecological factors cannot be excluded because different countries and regions may be ecologically different. The flora in tropical and subtropical forests in southern China are different from those in the temperate forests in northern China
[62].
Conversely, the high genetic diversities of wild strains from single locations have been well documented. Primeval forest strains from a single location may belong to highly diverged lineages, exhibiting a sympatric differentiation phenomenon
[11][12][13].
9. A Modified Genome Renewal Hypothesis for Explaining the Diversification of S. Cerevisiae in the Wild
The life cycle and mating behaviors of
S. cerevisiae (
Figure 1) probably contribute to the reproductive isolation and diversification of wild strains. Efficient sporulation might be a selected trait for
S. cerevisiae to survive in the wild
[21]. Repetitive starvation and aridity pressures in the wild would select for the capability to return efficiently to a diploid state, which is necessary for sporulation. Autodiploidization mediated by mating-type switch and intratetrad mating would apparently provide a selective advantage because these processes avoid the risk of the absence of adjacent mates with opposite mating types
[18][21]. Multiple reinventions of mating-type switching have occurred during the evolution of budding yeasts, suggesting strong natural selection in favor of this property
[19]. The seemly obligate homothallism of the wild
S. cerevisiae probably prevents outbreeding and genetic admixture. On the other hand, mutation or occasional outbreeding due to population admixture of the wild
S. cerevisiae caused by human or animal (insect) activities could create heterozygous strains. Reinstatement to a homozygous state of heterozygous strains due to haplo-selfing would produce new genotypes as predicted by Mortimer’s genome renewal hypothesis
[63][27][64]. In the case of wild strain diversification, the hypothesis needs to be modified, because purging of deleterious alleles is not a necessary function of this model. The neutral polymorphisms due to mutation or outbreeding in the occasionally formed heterozygous strains in nature can be fixed via subsequent haplo-selfing, as illustrated in
Figure 2. The modified genome renewal model can explain sympatric diversification observed in wild
S. cerevisiae, for neither geographic nor ecological isolation is required in this model.