Defining the nutritional drivers of yeast performance and matching yeast to must
From the AWRI’s extensive experience in the rescue of suboptimal fermentations, it has become increasingly clear that the ability to predict problematic fermentations, beyond an understanding of the impacts of sugar and nitrogen, is extremely poor. Work is therefore in progress to develop a greater understanding of the interactions of yeast strains with their environment. While currently over 200 yeast strains are available to the wine industry, there is limited detail available on how they perform in different contexts. This project aims to expand knowledge of yeast strain performance under a range of environmental conditions (i.e. in grape juices of widely varying composition as used for a range of wine styles) and inform winemakers about how they can reduce the risks of suboptimal fermentations though yeast strain choice.
Defining yeast strain relationships through whole genome sequencing
With so many yeast strains available, both commercially and within collections (for example, the AWRI wine microorganism culture collection, AWMCC), a major goal was to evaluate wine yeast genetic diversity. This was achieved through the use of whole genome sequencing and subsequent comparative genomic analysis. Draft genome sequences are now available for 200 wine yeast, comprising S. cerevisiae, S. uvarum and hybrids of S. cerevisiae with non-cerevisiae Saccharomyces yeast including S. eubayanus, S. paradoxus and S. kudriazevii. The strains sequenced included commercial and non-commercial yeast. Non-commercial strains were accessed from the AWMCC and represent isolates collected from numerous wineries, both in Australia and overseas, since the 1940s. Many of these have not been previously characterised. Older wine yeast isolates provide some historical context to the relationships between the various wine yeasts in this study and offer a potential untapped source of diversity for strain phenotypic assessment.
The shared genomic differences of the wine yeast in this work, compared to the lab yeast S288c, form the primary comparative dataset used to inform relationships between strains. Overall, the comparison showed a high level of genetic similarity between wine strains, consistent with a historically recent point of divergence. Of the 200 strains sequenced, 54 (27%) form a group with a large degree of similarity to strain EC1118. S. cerevisiae EC1118 is a robust workhorse yeast, isolated many years ago in France from secondary champagne fermentation. Within the larger comparison, small and tightly defined subgroups are evident, speaking to the influence of domestication through strain selection and breeding on the generation of increased levels of genetic divergence. Subgroups within the larger EC1118 group show particularly high levels of genetic divergence and evidence of breeding and selection. For example, seven strains are interspecific hybrids arising from crosses of non-cerevisiae strains with an EC1118-like parent. Surprisingly, there are fewer identical strains in circulation than were expected prior to commencing this work.
Parallel phenotypic assessment using barcoded strains
Information about relationships between strains obtained through comparative genomic analysis was used to select strains that are representative of the broader diversity of all wine yeast. This subset of wine yeast has been individually barcoded using a short molecular element that can be used to identify strains in mixed culture environments. The barcoded wine yeast collection allows parallel evaluation of multiple yeast strains in a single ferment and is now being used to assess strain fitness in a variety of environments that reflect the natural compositional variation in winery-produced grape juice.