This major goal of this project is to reduce bentonite use by better clarification of the conditions under which its use will be successful and to explore the feasibility of development of alternatives to bentonite. Bentonite functions in wine protein removal via simple ionic and absorptive interactions. Bentonite has a net negative charge at wine pH while most of the wine proteins are positively charged (have picked up protons) under the acidic conditions of wine. This simple charge interaction is not very specific and bentonite is known to strip wine of other components as well. An alternative method to achieve wine stability would be to utilize the specificity of an enzymatic system for protein destruction. In earlier grant years we identified proteases that are catalytically active in wine and that are capable of degrading wine proteins. However, while haze formation was reduced somewhat, haze was still formed in these wines. Our more recent goal was to determine why haze was still appearing in these wines to determine if a more effective protease could be identified. Using MALDI-TOF mass spectrometry we have determined that peptides and low molecular weight protein materials can also be main contributors to haze. Thus, protease treatment will likely not be fully effective in reducing the haze-forming potential of a wine. We have further characterized the proteins most associated with haze also using MALDI-TOF mass spectrometry in order to determine if an alternate strategy might be developed based upon the unique chemical properties of the unstable proteins. Our preliminary work with Sauvignon blanc indicates that the unstable proteins are proteins that have become “glucosylated”, that is, that have interacted and formed a covalent bond with juice glucose (or fructose). While this is an important finding towards understanding the kinetics of haze formation, we will not be able to take advantage of this property in the design of resins to remove wine protein as many key flavor components are also glucosylated and would likely be stripped by the new matrix. In addition, frequently the haze that forms in wine is not proteinaceous in nature so that bentonite is completely ineffective. Winemakers currently do not have any readily utilizable methodology to determine if their haze problem is caused by protein components that will respond to bentonite, is caused by proteins not interacting with bentonite or is not proteinaceous in nature at all. We have developed an amido black assay that can easily be used in a winery situation that will allow winemakers to determine the nature of the haze of their wines and to predict the effectiveness of bentonite.
MALDI-TOF mass spectrometry has been shown to be an excellent method for the direct analysis of the composition of wine haze. Previously developed methods allow a subtractive evaluation of wine components, assay of the wine pre and post haze treatment, but are not useful for the analysis of the haze itself. MALDI-TOF in contrast to SDS-PAGE is not specific for protein and can be used to simultaneously assess the presence of non-proteinaceous material in haze. It also allows accurate investigation of low molecular weight components, poorly resolved by other techniques. The MALDI-TOF analysis has demonstrated that the two commonly used methods to induce haze in wine as a means of assessing protein stability, the heat/chill and ethanol assays, destabilize specific components within the wine, but do not yield hazes of an identical composition to each other or to hazes forming naturally in wine under commercial conditions. Phenolic compounds were found to play a role in haze formation in wines made from Muscat of Alexandria but not in wines made from Sauvignon blanc. More wines from each varietal need to be evaluated before more conclusive statements can be made. It appears that the difference is due either to the greater level of phenolic compounds in Muscat or to differences in the peptide fractions between the two varietals.
In this grant year, we have nearly completed the analysis of genetically engineered wine strains of Saccharomyces over-expressing Protease A. The over-expression of Protease A does not lead to release of protease into the medium as is found in laboratory strains under laboratory conditions. We are in the process of determining if the vacuolar level of protease A activity is higher or not in these strains. MALDI and SELDI-TOF analyses are allowing characterization of wine proteins, and, more importantly, direct analysis of wine hazes, sediments and precipitates. The analysis of hazes has revealed a significant amount of low molecular weight material is present. This has not been previously observed, most likely because other techniques do not allow detection of low molecular weight substances. Protein profiles are similar for wines made form the same varietal, but there are marked seasonal differences in the level of protein present. Wines behave very differently upon aging on yeast lees. In the case sparkling wines from one winery, grape protein was completely degraded within 9 months, while in another winery protein persisted for years during aging. The proteins do not appear to be dramatically different among these wines, suggesting that other factors, such as the yeast used, are critical in proteolysis during aging.
Wine hazes are considered to be a visual sensory defect. Bentonite works extremely well at removing most species of proteins rendering wine protein stable. We do not have an equally reliable method for the removal of polysaccharide material which can also lead to a haze due to insolubility in ethanol. Bentonite is not specific and can strip wines of important flavor and aroma characters. In addition to this quality issue, bentonite use also generates solid waste and counties in California are seeking ways to minimize the volume of solid wastes produced. An alternative to bentonite that would result in protein stability without generating massive quantities of waste, adding any sensory attribute, or stripping wine characters is desirable. We are exploring the ability of the Saccharomyces protease, Protease A, to function in degrading wine proteins and elimination of protein hazes. We are also conducting a detailed compositional analysis of wine hazes to define the wine components that trigger haze formation. Once more is understood about the kinetics of haze formation, other alternatives to bentonite may become apparent. Our final goal is to characterize the haze material that is not protein and also to define the role of polysaccharide in the generation of haze. This year we will be including analysis of recurring haze problems being experienced by white Zinfandel producers.
Bentonite is the most effective, all-purpose reliable agent for the removal of wine proteins and haze-forming potential. However, bentonite is non-specific, having a tendency to over-fine wine removing complexity and can result in a significant loss of wine volume as bentonite lees. Bentonite lees also represent a significant amount of solid waste of high biological and chemical oxygen demand, the treatment and disposal of which may become increasingly costly. Alternative methods to achieving wine stability not suffering from these drawbacks is highly desirable. Therefore, we are exploring the feasibility of use of yeast acid proteases to degrade wine proteins preventing haze formation. We are also determining in finer detail the composition of hazes to determine the wine components involved in the formation of haze. Toward these ends, in this grant period a plasmid over-expressing Protease A of Saccharomyces was constructed which will be utilized in lees contact experiments to determine if the increase of protease activity is correlated with decreased susceptibility to haze. In addition, we have completed the third year of a comparison of bentonite treatment of juice to bentonite treatment of wine.to determine if less bentonite is necessary with a juice treatment. We have found that indeed bentonite treatment of some juices is more efficient than treatment of the resulting wines, but this is very juice specific. We have not been able to reproduce the haze-protecting effect reported for yeast polysaccharide or mannoprotein material of yeast cell walls. However, we have not yet truly duplicated the original observation by purifying the polysaccharide material from wine yeast lees. Finally, chemical analysis of heat and ethanol induced hazes revealed that these hazes are distinct. Ethanol hazes contain a much greater level of polysaccharide material and less protein than heat hazes. There are striking varietal specific haze problems as well.
Bentonite is the most effective, reliable treatment to achieve protein stability of wine. However, bentonite has a tendency to over-fine and strip wine of constituents important for complexity and can result in a significant loss of wine volume as lees. Disposal of waste bentonite is also a problem. For these reasons, an alternative to bentonite is desirable. Acid proteases should be effective at the selective degradation of proteins in wine, would not generate any lees and will not strip wine of positive characteristics. Commercially available proteases are not effective agents for protein stability. We have been examining the activity against wine proteins of proteases produced by yeasts with a normal habitat of grape juice. These proteases should have evolved towards being active against the proteins of their natural environment. In addition, we have been characterizing the proteins causing haze, as a specific subset of proteins, not total or all wine proteins, are responsible for haze formation. This grant is a continuation of these studies with the specific goals of purifying and characterizing haze causing proteins from Chardonnay, examining the effectiveness of bentonite treatment of grape juice rather than finished wine, exploring the nature of the polysaccharides causing polysaccharide haze and genetic construction of a strain of Saccharomyces secreting an acid protease.
Use of bentonite as a protein stabilizing agent for wine generates significant lees and results in the loss of flavor or aroma characteristics. An alternative to bentonite such as acid protease treatment would not lead to significant lees nor would non-proteinaceous materials be affected. The biochemical characteristics of proteins leading to haze formation are not yet known. Haze could result as a consequence of insolubility due to proteins with isoelectric points at or near the pH of wine, or to hydrophobic interactions among denatured proteins. Ethanol may reduce the solubility of glycosylated proteins. Ethanol also results in loss of solubility of polysaccharide material, also causing a haze. Hazes may arise for different reasons in different wines. Proteins of varying isoelectric points as predicted from retention time on ion exchange chromatography can lead to haze in wine. This was determined from an analysis of the protein composition of the haze material as well as from analysis of the haze forming potential of proteins separated from wines then subjected to heat treatment. Proteins with an isoelectric point near to the pH of wine were previously thought to be largely responsible for haze formation. Earlier results from our laboratory agreed with this finding. However, our studies of this past year suggest otherwise, but need to be confirmed using other juices. It is likely that the factors leading to haze formation vary depending upon the juice and seasonal variation. In any event, haze formation is a complex chemical process.
There are many causes of hazes in wine and two predictive tests: heat and ethanol. In our experience, ethanol-induced hazes are largely polysaccharide while heat induced hazes are proteinaceous. The protein profiles of numerous experimental and commercial wines were compared using HPLC at differing stages of bentonite treatment. Soluble proteins remaining in the wine, following heat treatment were also analyzed. Distinct HPLC peaks were associated with heat-haze. Total protein was not correlated with either heat or ethanol haze in agreement with many previous studies. Fermentation of Chardonnay in the presence of bentonite reduced requirement for bentonite in one wine, but had no impact on a second. Excessive bentonite treatment increased subsequent heat haze in some, but not all wines tested, in agreement with commercial winery observations.