Putting microbial diversity to work in shaping wine style
Abstract
There is substantial genetic diversity both within and between different yeast species, with significant potential to provide benefits to wine producers. This project has put that genetic diversity to work by characterising existing strains to help build the foundations of future strain development, introducing new traits through targeted selection, and expanding the range of characteristics available through the introgression of genetic elements not previously accessible in a winemaking context. In achieving these tasks the project team has also developed new methods which will help to build future generations of wine yeast.
Summary
A general constraint in microbial strain development programs can be access to germplasm free of third-party IP constraints. To this end, the project team sought to identify and gain access to Saccharomyces cerevisiae isolates that could be used for our ongoing strain development program. Three robust S. cerevisiae strains were identified (AWRI 1742, 1939 and 1796), which exhibited strong fermentation kinetics, an absence of negative quality attributes and produced a variety of compounds associated with positive aromas in wine.
Physiological characterisation of yeasts isolated by Project 4.4.3 was undertaken to identify novel characteristics and opportunities for strain refinement. In total, 250 genetically distinct yeast isolates were screened on solid media for winemaking-relevant traits. Just over 1% were classified as low-H2S producers. Seven of the 250 wild isolates exhibited potential for varietal thiol release during fermentation. One isolate exhibited robust fermentation kinetics in trial ferments and had a high-thiol-releasing phenotype.
Recent changes to the definitions of genetically modified organisms by the Office of the Gene Technology Regulator prompted an exploration of the use of CRISPR methodology to modify diploid yeast genotype. The approach typically makes use of a molecular repair feature of many organisms called non-homologous end joining (NHEJ). NHEJ is suppressed in diploid yeast, leading to low frequency of multi-allelic modification. The project team explored several approaches to overcome this issue and identified an approach that stimulates NHEJ in diploid wine yeast which substantially improved the efficiency of the CRISPR method. This development has laid the foundation for the use of this approach in the future development of robust yeast for industrial fermentations.
The non-Saccharomyces yeast genera Hanseniaspora can have a very active metabolism of the amino acid phenylalanine, which enhances the formation of benzaldehyde and benzyl alcohol, precursors to the ‘flinty’ compound phenylmethanethiol. The potential for strain-specific production of different benzenoid-derived compounds by a subset of H. vineae and H. osmophila strains was investigated. Even though some of the isolates from this genera produced higher concentrations of both benzaldehyde and benzyl alcohol than selected S. cerevisiae yeasts, this was not translated into increased formation of phenylmethanethiol.
The aromatic higher alcohols 2-phenylethanol (2-PE), tryptophol, and tyrosol are yeast-derived compounds that modulate the aroma and palate of fermented beverages such as beer, wine, and sake. A group of five variants of the widely used industrial wine strain AWRI 796, previously isolated due to their elevated production of the desirable ‘rose-like aroma’ compound 2-PE, were characterised in pilot-scale fermentation of a Chardonnay must. Results showed that these variants increased both the concentration of 2-PE and modulated the formation of the higher alcohols tryptophol, tyrosol, and methionol, in addition to volatile sulfur compounds derived from the amino acid methionine. This work highlighted the connections between nitrogen and sulfur metabolism in yeast during fermentation.
Such increased production of specific compounds may not suit all wine styles. Extensive trialling of three of the ‘rose’ yeast strains (AWRI 2940, AWRI 2965 and AWRI 4390) at pilot scale over several vintages has allowed an assessment of which wine styles may be more compatible with enhanced ‘rose’ aromas. It was demonstrated that the ‘rose’ yeast benefitted varieties with neutral profiles such as Chardonnay and Pinot Gris, sparkling Chardonnay and Grenache rosé.
The foundation yeast used for the production of the original ‘rose’ yeast is not an ideal commercial foundation for the production of sparkling wines. Therefore, a second generation of 2-PE overproducing yeasts was generated to improve compatibility with this winemaking style. A pilot-scale winemaking trial assessed the suitability of second generation 2-PE overproducing yeast for primary and secondary fermentation during sparkling wine production using the traditional method. The most prominent effect was achieved when the ‘rose’ yeasts were used during primary fermentation, with 2-PE being produced at between five to seven times higher concentration than the control. While the yeast were capable of conducting secondary fermentation, there was no compositional or sensorial benefit to their use at this winemaking stage. Both ‘rose’ yeasts had distinct effects on sparkling wine style. AWRI 4390 resulted in wines with more intense ‘rose’ and ‘fruity’ aromas than the control, while AWRI 2965 was characterised by an intense ‘brioche’ aroma.
Interspecific hybridisation is the term used when two different species are mated together. The project team has used this process to introduce genetic diversity into traditional wine yeasts and to harness traits from more distantly related yeasts that can then be used in winemaking. While the blending of genomes often results in additive combinations of characteristics, occasionally, a hybrid may exhibit characteristics different from either parent.
Two new hybrids containing partial genomes of S. eubayanus and S. arboricola were created during the project. This achievement completed the collection of possible hybrids between S. cerevisiae and other Saccharomyces yeasts with complementary mating type features from the sensu stricto group.
The seven hybrids in this completed set were characterised in pilot-scale fermentations of red and white grapes. The red wines differed across nine descriptive attributes. Key among the more positive attributes were fruity aromas of cherry, red berry, baking spice and vanilla. However, some strains produced wines with more prominent 'boiled egg' and 'canned vegetable' aromas. This work identified hybrid strains that appear well suited to the production of red wine.
White wines made using hybrids differed in yellow colour intensity, tropical fruit aroma, passionfruit aroma, banana aroma, honey aroma, stone fruit flavour and banana flavour. Overall, the S. mikatae hybrid strain AWRI 1504 appeared most favourable to white wine production, imparting stone fruit aroma and flavour.
One of the hybrids in the set, a cross between S. cerevisiae and S. uvarum, displays robust fermentation properties in high-sugar juices while producing wines with lower concentrations of acetic acid than either parent. The team set about discovering the genetic basis of the 'low acetic acid' trait in the uvarum hybrid. The genome of the hybrid was picked apart by sporulating it to generate hundreds of progeny. This work pointed toward the loss, or retention, of a specific uvarum chromosome being a defining feature in the hybrid, responsible for the strain's ability to produce a very low concentration of acetic acid.