Using glycosides and other flavour precursors for improved wine flavour
Abstract
Much of the flavour in wine comes from non-flavour-active precursors that have to be broken down to provide known varietal attributes. Through chemical, sensory and winemaking experiments, the potential of different unrealised flavour precursors to improve wine flavour was determined. Additives prepared from grape marc were effective at increasing flavour of wine and no- and low-alcohol products. Flavour formation was not reliant on fermentation but largely dependent on wine pH. Additionally, flavour evolves from precursors in the mouths of tasters, although their response to precursors depends largely on their individual sensitivity to the odorant released.
Summary
The route from grape to wine flavour can occur via many different pathways. Many classes of compounds are present in the grape in a non-volatile form, bound to sugars or amino acids. For them to be perceived as flavour or aroma, they must firstly be broken down and the volatile portion released. While this release of flavour into wine can happen either during winemaking or wine storage, there is a portion of these precursors that is not extracted into wine and can be found in grape marc. Additionally, monoterpene glycosides present a novel opportunity to add flavour via in-mouth breakdown during wine consumption.
Using previously prepared Chardonnay and Riesling wines with added Gewürztraminer marc extract, analysis after six months of storage showed marked changes. The extract that was rich in monoterpene glycosides provided no added bitterness in the wines, confirming adequate removal of phenolics from the isolation procedure. All wines showed increased ‘floral’ attributes which resulted from the hydrolysis of monoterpene glycosides to yield the aroma-active free monoterpenes. Adding the Gewürztraminer marc extract before or after fermentation made little difference to the outcome, showing that chemical hydrolysis was the driver for volatile release and not enzymatic hydrolysis during fermentation. As such, marc extracts can be added post-fermentation once the wine can be assessed for flavour intensity.
Importantly, when tasted, there were observed increases in ‘floral aftertaste’ for the wines with added Gewürztraminer marc extract. However, this observation was largely from those sensory panellists who had previously been shown to perceive flavour from glycosides via in-mouth release of the volatiles (deemed ‘tasters’). This suggests that a major factor in ‘floral aftertaste’ attribute rating was the in-mouth hydrolysis of monoterpene glycosides present in the wines.
The underlying reasons for panellists’ different responses to glycosides were investigated. While the microbial populations in-mouth were linked to the breakdown of glycosides when in contact with saliva samples, these did not relate to the ability of the panellists to detect aroma from the glycosides. Instead, the main factor in determining the response to glycosides was the sensitivity to the released volatile. It was also shown that individual thresholds for glycosides and the released volatiles were wide-ranging. The classification of the panellists into either ‘tasters’ or ‘non-tasters’ was a function of their threshold being above or below the concentration used in the test. Regardless, it was shown that glycoside additions to wines produce flavour via two mechanisms:
- in-mouth hydrolysis of the glycosides themselves during tasting and retronasal aroma detection of the released volatile
- slow chemical hydrolysis due to the acidic nature of wine to yield the volatile compound (and further rearrangement products) in wine.
The Chardonnay and Riesling wines with added Gewürztraminer marc extract were monitored chemically during three years of storage. The low pH Riesling (pH 3.1) showed a more rapid hydrolysis of geraniol glucoside, with key monoterpenes also reducing in concentration from the six-month analysis timepoint. For the Chardonnay wines (pH 3.4), the concentration of linalool was still increasing at the three-year analysis point. The 0.3 difference in pH created significant differences in the evolution of flavour as the wines aged. Adjusting for the speed of breakdown at different pH and comparing the Riesling wines after 11 months with the Chardonnay wines at 31 months showed a better alignment of the volatile profiles. While there were a number of chemical reactions taking place in these wines, the hydrolysis step of geraniol glucoside alone appears useful in understanding shelf-life characteristics of the wines.
The winery-scale production of glycoside-rich extracts was further optimised using cross-flow filtration from a 300 kg parcel of Muscat Gordo Blanco marc. This produced a similar extract to the Gewürztraminer extraction but with much faster throughput in the winery. Additionally, an in-depth review of the FSANZ regulations against the protocol for extraction indicated that these marc extracts were permissible additives for winemaking.
The Gordo extract was used to understand the first six months of storage in wine. The initial winemaking trials showed an increase in geraniol glucoside from what was in the extract to what was observed at the first timepoint after six months. When the Gordo extract was added to a commercial Chardonnay and monitored monthly, there was a continual increase in the concentration of geraniol glucoside. However, the quantification of disaccharide analogues did not account for the increases seen. As such, there must be another contributor to geraniol glucoside that is yet to be identified. Regardless, the concentration of geraniol glucoside in the extracts provided a good predictor for the evolution of volatile monoterpenes in wine.
As the bound monoterpenes are not odour-active and can largely be disposed of with marc, there was potential that traditionally ‘non-floral’ varieties might be hiding a pool of flavour precursors that were never realised. However, when marc from a number of floral and non-floral varieties was used to create extracts, this was shown to be incorrect. When stored in commercial wine for six months, the floral varieties produced much higher concentrations of volatiles, especially the Muscat varieties. Therefore, to produce potent glycoside-rich extracts from grape marc it would be beneficial to keep Muscat varieties separated from others, although this would present logistical hurdles.
From industry consultation on the use of glycoside-rich extracts in winemaking, it was highlighted that there was strong interest in using them to improve the flavour outcomes of no- and low-alcohol (NOLO) products. When the Gordo extract was tested in NOLO matrices, it was observed that the aggregated knowledge on the use of glycosides in winemaking is transferable. Initial tastings of Gordo marc extract in NOLO products showed potential to increase ‘wine-likeness’.
When a similar isolation protocol was performed on Sauvignon Blanc juice to obtain extracts rich in tropical thiol precursors, there was less success. Firstly, the requirement to use juice created extraction and filterability issues due to the high viscosity. Secondly, when the extracts were used in informal tastings, extremely high concentrations were required for only a portion of the panel to perceive flavour. It was determined that this process is unlikely to be applicable, for a number of reasons.
This project has delivered an extraction method for the isolation of purified glycosidic material that can be added to wine or wine-like products. The method has been developed specifically to be used in a winery setting and involves materials that are accessible to a winery. The development of the glycosidic extract has undergone significant industry consultation, specifically looking at segments of the wine industry that would be amenable to integrating it into their usual practices, with a number of large wine companies indicating they would use such an extract if it were available. Furthermore, there appears to be a natural fit between glycoside-rich extracts and a desire to achieve greater flavour outcomes in NOLO products.