Identification of yeast genes involved in sauvignon blanc aroma development

Abstract

The grape variety Sauvignon Blanc (SB) is the flagship of New Zealand’s wine industry and accounted for over 75 % of the value of total wine exports in 2008. Two volatile thiols, 3-mercaptohexan-1-ol (3MH) and 3-mercaptohexyl-acetate (3MHA), reminiscent of grapefruit and passion fruit respectively, are critical for the main varietal characters in New Zealand SB. These aromatic thiols are not present in the grape juice, but are synthesized and released by the yeast during alcoholic fermentation from non-aromatic precursors. The aim of this work was to elucidate the underlying genetics of volatile thiol synthesis in yeast (Saccharomyces cerevisiae) during alcoholic fermentation of grape juice. A gene-deletion strategy was chosen for the investigation of putative genes influencing 3MH and 3MHA release. The first part of this thesis optimized fermentation conditions in grape-juice-based media, which enabled auxotrophic laboratory strains, derived from S288C, to ferment grape juice to completion with high efficiency. Key steps to achieving this goal were the supplementation of the grape juice with higher than recommended amounts of amino acids, which increased the fermentation rate of auxotrophic yeast strains. Lysine auxotrophic strains especially benefited from this measure. In combination with the dilution of SB grape juice by 25 % with synthetic grape juice without sugars, the auxotrophic laboratory yeast BY4743 was able to metabolize all sugars in the grape juice-based media in a time frame similar to that of a commercial wine yeast. The key properties of the resulting wine were comparable to wine made with a commercial wine yeast under the same conditions. In the second part, these newly developed fermentation conditions were employed to screen 69 single-gene deletion strains in the laboratory yeast background BY4743. The list of the 69 candidate genes was compiled by combining existing knowledge about thiol production in yeast with the mining of several biological databases. Screening of the single-gene deletions revealed 17 genes which caused biologically relevant increases or decreases in volatile thiol production, but none abolished it. The majority of the 17 genes were related to the sulfur and nitrogen metabolism in yeast. A subset of these thiol-influencing genes were also deleted in a wine yeast, and were overexpressed in both wine yeast and laboratory yeast, to gain more insight in their regulatory effects. The findings confirmed that sulfur and nitrogen metabolism in yeast were important in regulating 3MH and 3MHA synthesis. Different sulfur and nitrogen sources were added to the grape must prior to fermentation and their effect on thiol release was studied. It was found that nitrogen sources urea and DAP, as well as, the sulfur compound S-ethyl-L-cysteine (SEC) increased 3MH and 3MHA concentrations in the resulting wines. The addition of cysteine to grape juice fermented with wine yeast deleted in genes CYS3 and CYS4 more than doubled total thiol production. Mapping approaches to investigate thiol production in yeast were employed in the final part of this thesis. Genetically mapped F2 progeny of a cross between a low thiol-producing yeast strain and a high-thiol producer were screened for their thiol phenotype. The 3MH and 3MHA phenotypes across 48 screened F2 progeny resembled normal distributions, indicating a quantitative trait. Subsequent mapping identified a locus on chromosome 14 with a small effect on the 3MHA phenotype, but no obvious candidate genes were evident in the region. Another approach to investigate the evolution of volatile thiols in yeast included the use of SEC, a thiol compound resembling the cysteinylated precursor of 3MH, as a sole nitrogen source in a yeast growth assay. It was found that most wine yeast, European yeast isolates and laboratory yeasts could utilize SEC as a nitrogen source, whereas various other S. cerevisiae isolates could not. Crosses between three pairs of Sec- and Sec+ yeast strains strongly indicated that this trait was monogenically inherited. However, no direct correlation between the SEC phenotype and volatile release could be observed. Genetic mapping experiments in one SEC-segregating yeast population linked this SEC phenotype to the leu2-D0 deletion in a cross between a Leu+ and Leu- yeast strain. It was shown that leucine auxotrophy most likely caused the Sec- phenotype. In a second F2 population of a cross between prototrophic Sec+ and Sec- strains, strong linkage was established to a region on chromosome 6 containing two candidate genes, DUG1 and IRC7. DUG1 was proved not to be the cause of the SEC phenotype, whereas IRC7 remains a strong candidate gene.