Abstract
Wine fermentation has long been conducted using Saccharomyces cerevisiae. This species is the workhorse of wine, beer, cider, sake and bread production all around the globe, as a pure species or, to a lesser extent, as a part of interspecific hybrids with other Saccharomyces species. However, other Saccharomyces species have shown a high potential to diversify the organoleptic properties of wine and tackle the environmental challenges that the wine industry has been facing in recent years. In this work, we sought to shed light on the phenotypic diversity in the genus Saccharomyces for winemaking. We phenotyped 92 yeast strains of all the current Saccharomyces species in synthetic grape must fermentation. Contrary to expectations, all Saccharomyces species fermented efficiently under the conditions used. Remarkably, strains of S. kudriavzevii and S. arboricola and interspecific Sc x Sk hybrids fermented more efficiently than wine S. cerevisiae strains. Regarding metabolite production, we observed a high strain variability for some species. Even more interestingly, we observed specificities at the species level: some non-cerevisiae Saccharomyces produced high amounts of industrially relevant compounds such as glycerol, succinate and fermentative aromas, or extremely low amounts of acetic acid, compared to S. cerevisiae. Overall, the potential of alternative Saccharomyces was higher than expected. They constitute a promising alternative to diversify the current set of commercially available yeast strains, either as pure species or following interspecific hybridisation. Given the interest of these phenotypes, we next aimed to determine their genetic basis through QTL mapping in Saccharomyces interspecific hybrids. Hybrids between different Saccharomyces species can be easily obtained in the lab thanks to the weak pre-zygotic barriers in this genus. Although viable, those hybrids are sterile, as the high genetic divergence between homologous chromosomes prevents their correct segregation during meiosis. However, allotetraploid hybrids are fertile, as recombination in these organisms occurs between chromosomes of the same species, minimising sequence divergence and facilitating correct segregation during sporulation. We used this phenomenon to construct fertile S. uvarum x S. mikatae hybrids. After genotyping and phenotyping its F12 progeny in oenological conditions, we performed the first QTL mapping study ever in non-cerevisiae Saccharomyces species. We found several genomic regions in the S. uvarum subgenome affecting the production of central carbon metabolites and, to a lesser degree, fermentative aromas and kinetic parameters. Verifying some candidate genes will shortly provide new tools for improving S. uvarum and its hybrids in a winemaking context. Overall, this study uncovered the potential of non-cerevisiae Saccharomyces species to tackle the current challenges of the wine industry, establishing innovative lines of action for future research.