Molecular drivers of wine texture and taste

Abstract

This project focused on the impact of non-volatile compounds on wine texture and taste. The evolution of phenolics and polysaccharides during white wine production was studied. Another primary aim was to identify new texture and taste target compounds in wine. Work on the bitterant tryptophol sulfonate showed that it was more important in white than red wines, with concentrations increasing in response to post-bottling SO2 additions. A coumaric acid glycoside was initially identified as a potential bitterant, but later shown through sensory analysis not to impart bitterness. A survey of glutamic acid revealed that it was present above detection threshold in most wines, and could impart ‘savoury’ flavour.

Summary

This project aimed to create new knowledge on the molecular drivers of taste and texture in wine and how they are affected by various winemaking inputs. Areas explored included compounds responsible for bitterness and savoury character and the impacts of dissolved oxygen management and winemaking steps on phenolics and polysaccharides (and the resultant wine taste and mouthfeel).

Formation of the ‘bitter’/‘hard’ tasting compound tryptophol sulfonate (TSO3) in wines when using a high-tryptophol-producing yeast was investigated post-bottling in Chardonnay, Riesling and Gewurztraminer wines, and it was found that TSO3 concentration stabilised 3-6 months post-bottling. The results further confirmed that wine SO2 concentration during bottle storage was a significant driver of TSO3 formation. Threshold testing of individuals to TSO3 revealed a heterogeneous distribution across tasters resulting in only around 1 in 10 tasters being able to taste TSO3 at the concentrations found in the white wines at 18 months post-bottling. Subsequent formal sensory analysis of the white wines found no significant differences in bitterness between the white wines with different TSO3 concentrations. Shiraz fruit was fermented using the same high tryptophol-producing yeast and adjusted to two different SO2 and pH values prior to bottling. Consistent with findings of the white wine study, little TSO3 was formed during winemaking. However, in contrast to the white wines, TSO3 in the red wines did not increase post-bottling, regardless of SO2 concentration or pH. The white wine study implicated SO2 availability post-bottling as a critical factor in TSO3 production, which may explain the differences between TSO3 production in the white and red wine. TSO3 could not be mitigated through intervention with either protein or polysaccharide agents. The difficulty in removing TSO3 from wine once formed in bottle suggests that mitigation strategies (particularly the judicious application of SO2 pre bottling) should be employed to avoid TSO3 formation.

Phenolic fractions were isolated from a bitter-tasting white wine prepared from hard pressings. A non-targeted metabolomics approach was used to analyse the composition of these fractions and identify compounds that correlated with their perceived bitterness. A glycoside of trans-p-coumaric acid (4-β-d-glucopyranosyl coumarate) was identified as a potential bitterant. The compound was synthesised and purified on a preparative scale with its identity and purity validated by NMR. It was then subjected to formal taste and texture descriptive analysis, presented in model white wine. Contrary to expectations, the sensory assessment indicated that at low concentration the compound suppressed the bitterness of the model wine, which is presumably elicited by ethanol. The hydroxycinnamic acids (of which coumaric acid is a member) are the predominant class of phenolic compounds found in white wine. This result suggests that the presence of glycosylated forms of hydroxycinnamates may positively contribute to the taste of white wine by reducing bitterness.

The term ‘savoury’ is synonymous with complex, high-quality wines. While the molecular drivers of ‘savoury’ character in wine are uncertain, the amino acid glutamic acid, succinic acid, and sodium and potassium salts are likely to contribute to savouriness in wine either directly or via their interactions. A survey of the concentrations of these compounds was conducted on 425 Australian commercial wines. The survey was also extended to include concentrations of sodium chloride in wine, as this salt is known to contribute to mouthfeel characteristics such as viscosity and soapiness. Results indicated that glutamic acid was above sensory detection threshold in most wines, and may impart ‘savoury’ character. The data generated will be used to design sensory experiments that further explore the interactions between compounds in wine known to produce savouriness, in an effort to gain knowledge that could benefit winemakers seeking to enhance the savoury character in their wines. 

In a scoping study to assess possible winemaking influences on glutamic acid and succinic acid concentrations in red wine, the impacts of must sugar concentration, YAN, fermentation temperature, yeast strain and press fraction were determined in two red varieties. Succinic acid concentration was influenced by yeast strain, YAN and fermentation temperature. Glutamic acid production was enhanced with higher fermentation temperatures, and significantly, the glutamic acid concentrations in the pressing wines were three to four times higher than in the free run wines.

In the late 1990s winemakers began using flotation to clarify white juices, as grape solids could be removed more quickly and at a lower cost than cold settling due to its lower energy requirements and continuous nature. Reports on the efficiency of must clarification using flotation are commonplace, but data on the influence of flotation on wine composition and wine sensory properties have not been widely available. A study involving two commercial wineries from the Murray Valley was conducted to compare the composition, taste and mouthfeel properties of Chardonnay and Frontignac wines made from juices clarified using different methods: flotation using nitrogen, cold settling/racking and unclarified. The non-volatile composition including total phenolics, total polysaccharides, phenolic profile and polysaccharide molecular weight profile of the wines produced by flotation were found to be much like those made using cold-settled juices, which was reflected in the wines having similar mouthfeel and taste properties. Both the Chardonnay and Frontignac wines were perceived to be slightly more viscous than the wines made from either the settled or unclarified juice. Perceived viscosity was best correlated with higher pH and higher total phenolic content and specifically with caftaric acid concentration. Surprisingly, higher total polysaccharide concentration was associated with lower perceived viscosity. The cold-settled Frontignac wines were slightly less bitter than the high-solids and low-solids floated wines, but the differences in bitterness could not be attributed to total wine phenolics nor to any of the quantified phenolic compounds including the known bitterants, catechin and epicatechin. This suggested that the compounds responsible for the greater bitterness in the settled wines made from this variety were not captured in this study. Perceived hotness, astringency and acidity were not consistently affected by juice settling methods. Overall, while a few differences in mouthfeel attributes between wines made from floated and cold settled juices were observed, the magnitude of the differences was small. Winemakers can benefit from the known efficiency and cost advantages of switching from cold settling to flotation, while being confident that the sensory and compositional quality of their wines is unlikely to be diminished by changing to the more efficient system of clarification.

The evolution of polysaccharides and total phenolics during the entire process of white winemaking including during short and extended periods of juice settling was determined for the first time in Sauvignon Blanc and Chardonnay. Total phenolics increased in the first 12 hours of settling, suggesting that they were being rapidly extracted from the grape solids. However, at the completion of fermentation and pre-bottling, the total phenolic content of wines was independent of the time taken to settle the juice off solids. The concentration of high molecular weight mannoprotein polysaccharides that assist in protein and cold stabilisation increased significantly following the start of fermentation. The concentration of medium molecular weight arabinogalactan protein polysaccharides (previously shown to contribute positively to white wine texture by reducing alcohol hotness and increasing perceived viscosity) reduced both during settling off grape solids and particularly during fermentation. Longer juice settling time, contrary to expectation, did not have a significant effect on the polysaccharide composition of the settled juices or of the finished wines. This was reflected in the sensory outcomes where no consistent differences were found between the wines.

Dissolved carbon dioxide (DCO2) was previously found to be a factor influencing the textural properties of white wines. This work was expanded to assess the influence of DCO2 on red wine taste and mouthfeel. The impact of DCO2 concentration on red wine taste and texture in the presence of increased red wine tannin and ethanol was determined in a Shiraz and Cabernet Sauvignon wine. Higher DCO2 concentration resulted in reduced bitterness and astringency and increased sweetness, while not affecting viscosity, hotness or overall flavour. However, the dominant effect of increased DCO2 on red wine texture was the addition of a ‘spritz’ character. A further study investigated the loss of DCO2 from the act of pouring wine into a glass and leaving it standing. The study found that significant losses in DCO2 resulted from the act of pouring, and that higher alcohol wines experienced greater DCO2 losses during standing. Based on the published difference threshold for DCO2, the loss would have resulted in a perceptible reduction in spritz character, but only after leaving wine in the glass for 15 minutes. This work indicates that winemakers should consider their protocols regarding pouring of samples and the time between pouring and assessment when making decisions at the tasting bench. 

Based on the results of the project, winemakers now have a clearer understanding of a range of influences on wine taste and texture that they can incorporate into their practices.