Résumé
The story of biotechnology of wine yeasts started a long time ago. The selection of yeasts from wineries for commercial purposes was developed from the 1970s. The aim was to display satisfactory fermentation profi les and to ensure product quality. From the 1980s, classical breeding, interspecifi c breeding and mutagenesis were applied for the improvement of existing strains. This resulted in many developments: for example, better nitrogen assimilation and fermentation kinetics were obtained by random mutagenesis (Salmon and Barre 1998). These classical approaches are still used; EMS (ethyl methyl sulfonate) mutagenesis resulted in the selection of commercial wine strain variants which produce less reduced hydrogen sulfi de due to mutations into genes encoding sulfi de reductase subunits (Cordente et al. 2009). But genomic features of wine yeasts were rapidly identifi ed as strong limitations for these approaches. Indeed, wine strains are mainly diploid, polyploid or aneuploid. They exhibit a chromosomal polymorphism (Bidenne et al. 1992). Chromosomal trisomies or tetrasomies result in impaired sporulation ability and in highly variable spore viability (Johnston et al. 2000). In addition, wine strains are generally homothallic, i.e., able to switch mating type when haploid, and highly heterozygous. All these features confer to wine yeast, high genome plasticity and limit the stability of lineage of variants.