Bioinformatic approaches for the genetic and phenotypic characterization of a Saccharomyces cerevisiae wine yeast collection

Abstract

In the years following the publication of the Saccharomyces cerevisiae genome, enough evidence was provided showing substantial genetic differences among wine yeast strains. At present, about 200 S. cerevisiae strains are commercially available, whereas each strain harbours a combination of specific traits that are of interest for winemakers. Exploring the biodiversity and genetic constitution of indigenous fermentative strains is an important contribution towards the selection of strains with specific phenotypes. From 54 spontaneous fermentations of grapes collected in three vineyards of the Vinho Verde Region during three consecutive harvests, a total of 1620 yeast isolates were obtained and categorized in 300 strains based on mitochondrial DNA RFLP patterns. For each pattern a unique allelic combination for six polymorphic microsatellite loci (ScAAT1 – ScAAT6) was found. No correlation was apparent between geographic distance and genetic similarity among individual isolates. However, genetic differences among S. cerevisiae populations were apparent from gradations in allele frequencies rather than from distinctive “diagnostic” genotypes, and the accumulation of small allele-frequency differences across six loci allowed the identification of population structures. Additional bioinformatics approaches (Bionumerics software) were performed with a subset of 47 strains in order to search for associations between microsatellite data and phenotypic traits, using semiquantitative results from an enzymatic activities test panel (api®ZYM, bioMérieux® SA). Most strains showed similar phenotypic profiles, whereas five strains were clearly distinctive based on the activity for acid phosphatase, naphthol-phosphohydrolase and β-galactosidase. Three of them showed rare alleles (allelic frequency 0.05 - 0.2) for the more polymorphic microsatellites ScAAT1 and ScAAT4.