Protein instability in wine is the appearance of an “amorphous haze or deposit” in the wine (1). Chemically, the haze is made up of complexes of proteins, polysaccharides, polyphenolics and metals (1). Normally, proteins in wine are dissolved, (coated in water molecules and not visible). Due to an overall positive charge on the protein at wine pH, they repel one another. The appearance of haze is associated with time, heat, or a change in chemistry that causes previously dissolved proteins to lose their three- dimensional shape, clump together, and come out of solution.

To understand how proteins are behaving in the wine matrix, it is helpful to understand a little bit about the molecular structure of proteins.

Figure 1: Three dimensional protein folding is determined by the sequence of amino acids. From https://doctorlib.info/genetics/medical-genetics/2.html

Proteins are polymers of amino acids bonded together in a row. In a liquid environment (wine), this string of amino acids folds in on itself again and again, like spaghetti in a bowl. Unlike spaghetti, the strand of amino acids is not the same along the whole length. There are 20 different kinds of amino acids found in living things, and each different kind of amino acid has a different side chain with different chemistry. Amino acid side chains can be positively charged, negatively charged, neutral, large and bulky, small and lean, water loving or water hating. As the string of amino acids folds back on itself, the side chains will interact to attract or repel one another depending on the chemistry of each side chain. The charge of the side chains is very important in folding, as side chains with the same charge will repel one another while side chains with different charges will attract one another. The water-loving nature is also important. The protein will fold in a way to hide the water-hating side chains in the middle of the protein, away from the surface, and put the water-loving side chains on the surface. Based on the sequence of amino acids for each protein (which is determined by genetics) each protein takes on a characteristic folding pattern that is determined by all of the interactions of the side chains (2).

Several changes in the environment can destabilize the folding of the protein (1). 

• Acid changes occur in wine through fermentation, malolactic fermentation, cold stabilization (the removal of tartaric acid), blending and aging. A slight acid adjustment (to a lower pH) shifts side chains to a slightly more positive state, so a small addition can help protein stability in the wine a great deal (4).
• If the wine is heated, all of the molecules in the wine start to move faster, and it takes more energy to hold the folding pattern of the protein together. If the interactions were weak to begin with, the protein might unfold. This is the basis of the heat-based stability tests that add a lot of heat to force all of the proteins to unfold. In practice, heat is most likely to occur under non-ideal storage or transport conditions.
• An increase in ethanol (due to fermentation, blending, or fortification) will increase the likelihood that the water-hating side chains in the middle of the protein will come out to the surface. This change in the folding pattern can destabilize the three dimensional structure and cause haze. 
• Over time, the protein itself starts to degrade and the amino acid strand breaks apart, leading to unfolding. When the proteins unfold, they are much more likely to interact with one another in solution, and can sometimes form a web, which leads to haze.
• Other components of the wine matrix maybe be involved in preventing protein haze (like mannoproteins) or in participating in the haze (like polysaccharides and phenols). Therefore, changes in the wine due to fining, filtration or additions can change protein stability.

What this means in practice:

Given the above considerations, when testing protein instability to determine a bentonite addition, testing must be done on the final wine. If you are preparing a blend, planning additions or cold stabilization, the protein stability of the wine might change with each of these. If you have to test early, make your bench blend as close to the final wine as possible. If you are planning to add CMC -based products for tartrate stability, it is safest to re-test the stability of the wine prior to addition but after all other operations have been completed. It is also possible to do a bench trial for CMC stability. (If you would like this protocol, feel free to email VaWrex@gmail.com with your request.)

References

(1) Zoecklein, B.; Fugelsang, K. C.; Gump, B. H.; Nury, F. S. Wine Analysis and Production; Springer: New York, 1995.

(2) Urry, L. A.; Cain, M. L.; Wasserman, S. A.; Minorsky, P. V.; Reece, J. B. Campbell Biology (11th Edition); Pearson: New York, 2017.

(3) Jackson, R. S. Wine Science: Principles and Applications, 4 edition.; Academic Press: Amsterdam, 2014.

(4) Zoecklein, B. Protein Stability Determination in Juice and Wine. Virginia Tech Online Enology Publications 1991.

Proteins are an important component of all living cells, including grapes and yeast, and are therefore plentiful in grape must and wine. Most of these proteins are not unstable. In fact, the protein instability in a wine is not strongly correlated to the overall amount of protein in the wine (1,2). Many of the proteins found in grapes are removed in the process of fermentation. Yeast enzymes break some of them down for food. Others become unstable when cells burst and they are exposed to the acidic environment of the juice. These precipitate and are removed as part of the lees. Others denature (unfold) as the alcohol levels build up, precipitate and are removed. By the end of fermentation the proteins that are left are those that are resistant to proteases, tolerant of wine pH and alcohol levels (1).

In recent years, the specific proteins responsible for protein instability have been identified, allowing for a better understanding of conditions leading to protein instability as well as development of targeted techniques for treatment. Most of the unstable proteins come from the grapes, not the yeast and belong to a specific family of proteins called the pathogenesis-related (PR) proteins (2,1). These proteins are expressed in grapes from veraison to harvest, with increasing expression as ripening progresses (increasing with increasing brix). Higher levels of expression are also seen with stress, wounding, or microbial attack such as Botrytis and powdery mildew (2).

Figure 1: Pathogenesis-related proteins from grapes are responsible for protein instability in wine. From Tian et al 20175”

In grapes, PR proteins protect grapes from microbial attack by damaging the cell walls of invaders. Depending on the circumstances, these proteins can make up 50-70% of the soluble protein content of grapes at maturity reaching levels of 50-800 mg/L depending on cultivar, season and level of environmental stress. They are localized to the skins of the grapes (2,1).

Chemically, as expected, they are resistant to proteases (so not broken down by yeast during fermentation), soluble at low pH (so not precipitated out during fermentation), and they have minimal tannin binding (not removed during fermentation). As a result, these proteins are the major protein left after fermentation. Unfortunately, this means they are also very difficult to remove.

The family of pathogenesis-related proteins includes thaumin-like proteins, chitinases and B-glucanase. The chitinases have been identified as responsible for the majority of protein instability in wine. Based on melting point studies of each of these proteins, chitinases were most likely to denature and form irreversible, strongly associated haze while the other proteins in this group showed less probability to denature, and greater probability to re-fold without haze formation. Using linear regression of melting behavior, it was estimated that chinases would denature in 6 minutes at 55C, 3 days at 35°C or 2 years at 25°C. This timeframe is consistent with protein instability seen in aging wine3. Also, unlike overall protein concentration, a correlation was found between chitinase concentration and haze formation (4).

Practically why this matters:

The identity of chitinase as the main protein in protein haze formation explains why total protein content is not a good predictor of instability and why our current test for instability likely overestimates the amount of bentonite fining needed to stabilize the wine. It also allows for better development of treatments for protein haze, as well as better testing for instability. These are currently under development. 

References

(1) Waters, E. J.; Muhlack, R. A.; Pocock, K. F.; Colby, C.; O’Neill, B. K.; Jones, P. Preventing Protein Haze in Bottled White Wine. Australian Journal of Grape and Wine Research 2005, No. 11, 215–225.

(2) Jackson, R. S. Wine Science: Principles and Applications, 4 edition.; Academic Press: Amsterdam, 2014.

(3) Robinson, E.; Scrimgeour, N.; Marangon, M.; Muhlack, R.; Smith, P.; Godden, P.; Johnson, D. Beyond Bentonite. Wine and Viticulture Journal 2012, No. November/December, 24–30.

(4) Ndlovu, T.; Divol, B.; Bauer, F. F. Yeast Cell Wall Chitin Reduces Wine Haze Formation. Appl. Environ. Microbiol. 2018, 84 (13)

(5) Tian, B.; Harrison, R.; Morton, J.; Jaspers, M.; Hodge, S.; Grose, C.; Trought, M. Extraction of Pathogenesis-Related Proteins and Phenolics in Sauvignon Blanc as Affected by Grape Harvesting and Processing Conditions. Molecules 2017, 22 (7)

The most common approach to removing unstable protein from wine is the use of bentonite. Bentonite is a volcanic montmorillonite clay that, in the US, is mined in Wyoming. There are several different kinds of bentonite based on the geographic origin with sodium and calcium based bentonites the two main categories on the market. Specifics of bentonite type will be covered in a separate newsletter, including experiments for efficacy and a discussion of efficiency.

Bentonite has a physical structure that looks like many overlapping sheets (Figure 1). These sheets are made up of crystals of tetrahedral silica dioxide and octahedral aluminum hydroxide. When hydrated in water, the sheets take on negative charges which causes them to repel one another. This can be seen in the physical swelling that occurs when bentonite is added to water in the winery. Once fully hydrated, the sheets orient themselves to minimize contact with other sheets that have the same charge. The flat ends of the crystals are positivity charged, which cause the whole structure to look like a house of cards (1).

Figure 1: Bentonite hydration, from Zoecklein 1998

Once it is swelled, bentonite has a high density of negatively charged particles. When added to wine, positively charged proteins bind to these negatively charged sheets. Charge neutralization causes proteins to clump to one another, and to the bentonite sheets, to form precipitates that drop to the bottom of the tank. Different proteins have different amount of positive charge on their outer surface, so will interact with bentonite in stronger or weaker ways. Unfortunately the most unstable proteins are also closest to their isoelectric points, meaning they are close to having a net charge of zero. Therefore the most unstable proteins are the least likely to bind to bentonite. 

Bentonite fining is based simply on charge, so anything that is positively charged will be bound. Potassium in wine is present in its positively charged form and may also bind to bentonite. This means that potassium is taking up binding sites on the bentonite, leading to higher rates needed to remove protein. If stability testing is done prior to cold stabilization, and fining is done after cold stabilization, the test may overestimate the amount of bentonite needed, as potassium is removed during cold stabilization. 

Anthocyanins are also positively charged at wine pH, and may bind to bentonite, which will strip color from the wine. This is one reason bentonite is rarely used on red wines. 

Despite the drawbacks, bentonite is still the most widely used fining agent for protein stabilization because it is thought to have little impact on sensory characteristics compared to other fining agents (2,3).

When, why and how bentonite is used

Bentonite can be used at several stages in the life of the wine including juice settling, fermentation and pre-bottling preparations. Pre-fermentation, the role of bentonite is mostly to settle juice lees. Jackson (2014) suggests adding bentonite after an initial settling period to help compact the lees. This is a useful approach for winemakers who wish not to use settling enzymes. Bentonite use on juice can also be used to remove laccase in Botrytis-affected fruit (2). Soluble protein content can increase by 50% during cold settling due to pulp exposure (3,4) and the use of bentonite to settle juice may limit the increase in protein extraction coming from must during long settling times (3). 

In a study testing the effects of bentonite pre-fermentation and during fermentation, Weiss and Bisson (2002) found that, at higher rates (36 g/hL), juice fining slowed fermentation and increased the total fructose in the final wine in some, but not all, of the lots they tested. Juice fining had no effect on amino acid composition nor YAN, though it reduced total protein in some cases. Lagging fermentation was not revived with nutrients, so they hypothesize an essential element such as sterols, fatty acids or phospholipids may have been taken out. Despite changes in overall protein content, juice fining did not change protein stability of the final wine (5).

Bentonite can also be added during fermentation with the intention of taking proteins out of solution with less effect on the aromatics of the wine. The rationale here is that many aromatics are still bound to precursor molecules during fermentation, so they are less susceptible to fining. Juice pH is generally lower than wine pH, which presumably increases the adsorption rate of proteins to bentonite when compared with the finished wine. Protein-bentonite adsorption has been shown to occur quickly, within 30 seconds of contact (6), and continued exposure of proteins in juice to bentonite that is mixed by the activity of fermentation itself provides continual accessibility to bentonite binding sites. Common rates for bentonite at this stage range from 25-40 g/hL, and it is commonly reported that addition of bentonite during fermentation can reduce bentonite addition rates later (3). Use of much higher rates is common in cold climate growing regions such as Germany (Maggie McBride Haek, Scottlabs, personal communication). Scottlabs has recently added a bentonite product specifically marketed to fermentation (Fermobent) with suggested rates of 200-300 g/hL for juice with high protein content.

Unlike their findings with juice fining, Weiss and Bisson (2002) found an increase in fermentation rate when wine bentonite was present during fermentation, possibly due to CO2 degassing. They also found a notable reduction in overall protein, especially at rates of 36 and 60 g/hL. However, as mentioned before, there is poor correlation between overall protein content and protein stability. This study did not test resulting protein stability.

Bentonite use both before and during fermentation is not uncommon in Virginia. In a survey of Virginia winemakers with 35 responses, 25% responded that they use bentonite during juice settling, 20% use it during fermentation and 8% replied they use it during both settling and fermentation.

The most common use of bentonite, however, is after the completion of fermentation, usually in the pre-bottling stage. At this time, bentonite addition rate can be determined by lab testing for each individual wine to be bottled. As mentioned above, it is important this testing occur after all other cellar operations have been completed, as any change in the chemical environment of the wine due to blending, pH adjustments, cold stabilization, fining or filtration can alter the protein stability of the wine. It is also important to do the bench test with the same type and lot number of bentonite that will be used to treat the wine, as bentonite can differ significantly among brands and even production lots. 

In a study testing the effect of fermentative vs. post-fermentative bentonite fining on aroma compounds in Sauvignon Blanc, Vela et al (2017) found that bentonite was more effective in removing proteins post-fermentation, meaning less bentonite was used overall. Despite high addition rates needed to achieve protein stability (100 g/hL), they found no significant differences in ethyl esters or acetates, and no impact on aroma in sensory analysis. There were differences, however, mainly in the thiol content. All three major thiol components known to characterize aromas of Sauvignon blanc (Table 1) were decreased with bentonite use, and the decrease was larger when bentonite was used during fermentation compared to its use on the finished wine. These results led the authors to hypothesize that bentonite interacts more strongly with thiol precursors than finished thiols (7). 

Despite its widespread use, there is still much debate as to the best time to use bentonite and the potential impacts bentonite fining has one wine. Effects may be different for different wines. Therefore, each winemaker must decide for him or herself what is the best approach.

References

(1) Zoecklein, D. B. Wine Proteins and Protein Stability. Virginia Tech Enology Notes Winemaking Topics.

(2) Zoecklein, B.; Fugelsang, K. C.; Gump, B. H.; Nury, F. S. Wine Analysis and Production; Springer: New York, 1995.

(3) Jackson, R. S. Wine Science: Principles and Applications, 4 edition.; Academic Press: Amsterdam, 2014.

(4) Waters, E. J.; Muhlack, R. A.; Pocock, K. F.; Colby, C.; O’Neill, B. K.; Jones, P. Preventing Protein Haze in Bottled White Wine. Australian Journal of Grape and Wine Research 2005, No. 11, 215–225.

(5) Weiss, K. C.; Bisson, L. F. Effect of Bentonite Treatment of Grape Juice on Yeast Fermentation. American Journal of Enology and Viticulture 2002, 53 (1), 28–36.

(6) Blade, W.; Boulton, R. Adsorption of Protein by Bentonite in a Model Wine Solution. Am. J. Enol. Vitic 1988, 39.

(7) Vela, E.; Hernández-Orte, P.; Castro, E.; Ferreira, V.; Lopez, R. Effect of Bentonite Fining on Polyfunctional Mercaptans and Other Volatile Compounds in Sauvignon Blanc Wines. American Journal of Enology and Viticulture 2017, 68 (1), 30–38. 

Despite the potential drawbacks, bentonite is the most commonly used fining agent for proteins. In some cultivars grown in Virginia, such as Sauvignon Blanc, Pinot Gris, Petit Manseng and Traminette, large doses of bentonite (100 g/hL, 7.5 lbs/1000 gallons) are needed to achieve stability. This leads to concerns over flavor stripping as well as overall loss of wine due to large amounts of lees. Other options are being explored to work in conjunction with bentonite to improve its effectiveness, or to replace bentonite altogether:

Figure 1 Structural rendering of Chitinase, the principle unstable protein in wine. From https://www.creative-enzymes.com/similar/chitinase_122.html

Protease enzymes

One approach to reducing protein instability is the use of a targeted enzymes to break unstable proteins into smaller pieces to either take them out of solution earlier or to make them less likely to form a haze. However, several considerations make this difficult. The pathogenesis-related proteins responsible for protein instability are generally resistant to proteases, and, most protease enzymes don’t work well at wine pH and temperature (Robinson et al 2012, Waters 2005)(1,2).

Despite these challenges, Australian Wine Research Institute (AWRI) has reported work on the use of a targeted protease to treat protein instability in wine2. Treatment involves first heating juice to 70-75°C (158-167°F) to cause the proteins to unfold, then treating with proctase enzyme, which has a good activity at juice pH. Since the proteins have unfolded they are more susceptible to protease attack. Heating the juice also increases enzyme activity. When tested at both the laboratory and production scale, treatment with proctase showed efficacy comparable to bentonite in removing chitinase proteins. In a triangle test, a sensory panel was not able to distinguish wines fined to stability with bentonite from those treated with heat and proctase. It is difficult to assess if this treatment achieves protein stability, since the most commonly used test, heating the wine to 80°C for six hours, is really a measure of overall protein content. By its nature, proctase is not removing all proteins, only the unstable ones. AWRI is working on developing a test that is better suited to assess true protein stability in the winery (2).

Proctase is not yet widely available, and most wineries do not have the equipment needed to heat the wine to the prescribed temperatures for treatment. However, many some pectinase enzyme mixtures have some protease activity. Use of pectinase mixtures with known protease activity has been suggested to decrease help break up unstable proteins in juice. Practically, this may not reduce protein content enough to achieve stability on its own but may be included in an additive strategy along with tannins and bentonite during fermentation to achieve stability in the finished wine (Eglantine Chauffour, personal communication). This approach is one of several interventions suggested by the Enartis Total Protein Stability protocol, tested by Ingleside in 2018.

Tannin addition is another component in an additive strategy. In this approach, tannin addition is used to react with proteins at the juice stage. Tannin is often used as part of rot protocol to remove laccase (a protein) from juice. It has been shown that hybrid grapes have lower levels of tannin due to high protein content indicating an active binding of protein with tannin in grape juice and must3. Addition of selected exogenous tannin has been suggested as a possible mechanism to remove overall protein load from juice to reduce instability later4. Enological tannins are added to the press pan or at juice settling and left behind when juice is racked prior to inoculation. This has the added benefit of antioxidant protection during settling. Tannin addition is also part of the Enartis Total Protein Stability protocol.

Chitin and Chitosan

Chitinase, the primary offender in protein instability, is an enzyme meant to act against chitin. Chitin is a component of the cell walls of organisms in the Kingdom Fungi (chitinase breaks these down to kill invading fungi) but is also found in the exoskeletons of crustaceans and insects. It is the second most abundant polysaccharide found in nature5. The most direct way to reduce chitinase in wine would be use chitin as a fining agent. When tested in bench trials, the addition of chitin resulted in dose-dependent removal of chitinase with up to 80% reduction in haze with much less reduction in overall protein (6). However, organoleptic and chemical effects of fining were not evaluated. Also, chitin is not currently a legal additive to wine.

Chitosan is a legal additive to wine. It is produced industrially by the de-acetylation of chitin, and therefore retains a structure similar to chitin. It is also found naturally in the cell walls of some fungi5. Trials using 1 g/L chitosan to stabilize Muscat showed an almost complete reduction in chitinase and near stabilization of the wine (NTU of treated wine was 2.10) as well as removing potassium, calcium and iron (Ndlovu et al 2019). Removal of these metals can aid in tartrate stabilization and reduction on oxidative capacity. However, treatment also showed a reduction in acidity (tartaric acid decreased by 0.65 g/L, malic acid decreased by 0.46 g/L) as well as reductions in free Terpenols, important aroma compounds in Muscat. In addition, chitosan is currently very expensive ($1/gram), so treatment of wine at this level would be cost prohibitive.

Figure 2: Yeast cell walls stained for the presence of chitin. (a) S. cerevisiae (b) S. paradoxus. From Ndlovu et al (2018)

As members of Kingdom Fungi, yeast also have chitin in their cell walls. An alternative to using exogenous chitin or chitosan is to increase the chitin already found in the cell walls of yeast. In a study of yeast cell wall chitin, Ndlovu et al (2018) measured chitin content in yeast cell walls, as well as the ability of yeast lees to reduce protein instability. They found that cell wall chitin was variable based on yeast species as well as the environmental conditions in which the yeast were raised. Saccharomyces paradoxus strains had much higher levels of chitin and produced wines with less chitinase after fermentation. Post-fermentation fining with S. paradoxus yeast hulls also reduced chitinase concentration in the resulting wine. Cell wall chitin levels were higher in cooler fermentations (15°C vs. 37°C) as well as when calcium was added to the growth medium (7). These findings lead to the potential of raising S. paradoxus strains specifically for chitin expression to be used in chitinase fining.

The role of polysaccharides and polyphenols

Protein haze is not entirely due to protein, but rather a complex of proteins, polysaccharides and polyphenols. In a study of the effect of wine composition on protein stability, Mesquita et al 2001 showed that the polysaccharide fraction of wine affected the characteristic behavior of a wine to become turbid under moderately high temperature (40-50°C, 104-122°F)(8). Pectin has been shown to be important in haze formation1. In an approach that alters the protein and polysaccharide fraction of the wine, Spada (2019) outlines a protocol using of Scottzyme KS, a mixture of protease and pectinase enzymes developed to settle juices that were hard to filter.  The study posits that heat unstable proteins are present in a colloidal state and therefore less available to bentonite, and suggests treatment of wine with 0.079 ml/L Scottzyme KS, 12-24 hour incubation, followed by bentonite addition. In this trial, treatment reduced instability by 82% relative to bentonite only.  KS is a broadly acting enzyme mixture, so it may affect aromas and flavor. No information is given on the effect of KS treatment on chemistry or organoleptic properties of the treated wines4. 

Mannoproteins, another type of polysaccharide, have been shown to decrease the size of protein precipitates, thus making them less visible1. This effect, along with chitin in cell walls of yeast, is likely why wines that have been aged on lees have lower levels of protein instability than younger wines or those not aged on lees.

When protein precipitates in wine are analyzed, they often contain proteins, polysaccharides and polyphenols. In fact, 50% of proteins in wine have been shown to be associated with phenolics. The hydrophobic nature of phenolics may stabilize hydrophobic portions of proteins that become exposed during unfolding holding them open and making haze more likely(1). Waters et al (2005) hypothesize that phenolic compounds may be a required element in haze formation (1).

References

(1) Waters, E. J.; Muhlack, R. A.; Pocock, K. F.; Colby, C.; O’Neill, B. K.; Jones, P. Preventing Protein Haze in Bottled White Wine. Australian Journal of Grape and Wine Research 2005, No. 11, 215–225.

(2) Robinson, E.; Scrimgeour, N.; Marangon, M.; Muhlack, R.; Smith, P.; Godden, P.; Johnson, D. Beyond Bentonite. Wine and Viticulture Journal 2012, No. November/December, 24–30.

(3) Springer, L. F.; Sacks, G. Protein-Precipitable Tan- Nin in Wines from Vitis Vinifera and Interspecific Hybrid Grapes (Vitis Ssp.): Differences in Concentration, Extract- Ability, and Cell Wall Binding. Journal of Agricultural and Food Chemistry 62, 7515–7523.

(4) Spada, P. A Modern Look at Protein Stability. The Grapevine Magazine 2019, April.

(5) O’Kennedy, K. Chitin, Chitinase, Chitosan ... Wineland Magazine, 2019.

(6) Vincenzi, S.; Polesani, M.; Curioni, A. Removal of Specific Protein Components by Chitin Enhances Protein Stability in a White Wine. American Journal of Enology and Viticulture 2005, 56 (3), 246–254.

(7) Ndlovu, T.; Divol, B.; Bauer, F. F. Yeast Cell Wall Chitin Reduces Wine Haze Formation. Appl. Environ. Microbiol. 2018, 84 (13).

(8) Mesquita, P. R.; Piçarra-Pereira, M. A.; Monteiro, S.; Loureiro, V. B.; Teixeira, A. R.; Ferreira, R. B. Effect of Wine Composition on Protein Stability. 2001, 7.