Figure 1. WINEXRAY's representative phenolic profile of a 21-day macerated Bordeaux varietal fermentation (The Future of Winemaking: Honoring the Vision of Professor Roger Boulton, 2022).

Protein-precipitable tannins (pTAN) react with saliva and are responsible for structure. Unlike free anthocyanins, they are very persistent and their mouthfeel can be shaped like clay using different enological controls. The Adams-Harbertson Assay was designed to mimic human sensory by using cow saliva to precipitate grape-derived condensed tannins. It’s important to note that this assay does not measure hydrolyzable oak tannins despite their astringent properties. Nevertheless, the vast majority of tannins in wine are from grape skins, seeds, and stems. They vary in concentration, size, composition, and polarity, all of which impact the resulting structure, astringency, and texture of the wine.

Concentration

The A-H assay’s protein-precipitable tannin measurement has the highest correlation to astringency (r2 = 0.82-0.90) when compared to high-performance liquid chromatography (HPLC) and methyl cellulose precipitation (MCP) (Mercurio & Smith, 2008). This being said, the characteristics of tannin chemistry are too diverse for concentration to represent astringency alone. At bound, we prefer to refer to protein-precipitable tannin concentration as a metric of wine structure because astringency is dependent on several other variables (bound anthocyanin incorporation in particular). Structure, on the other hand, is the phenolic backbone from which skin-macerated wines are built. Higher pTAN concentrations relate to a wider breadth of astringency on the palate. It is important to note that tannins are much more stable and persistent than free anthocyanins in wine. In practice, we see that protein-precipitable tannins decrease roughly 10-20% within the first year post-fermentation and persist thereafter at very stable concentrations. Another thing to consider is that extraction during fermentation is partial. Only 5-15% of grape tannins are extracted into the finished wine (WINEXRAY). This informs our fundamental understanding of tannins. They are abundant and they are persistent. The only exception to this rule is in the presence of extreme heat. Grapes nearing ripeness exposed to ambient temperatures above 43°C (110°F) show signs of cascading oxidation that deplete skin and seed tannin concentrations (about 30-50%). This type of heat stress also diminishes phenolic reactivity and stability in the finished wine with concentrations continuing to decrease in extended maceration and aging. WINEXRAY’s study with UC Davis in 2022 found an additional 60% reduction in extracted protein-precipitable tannins due to extended maceration of heat-damaged fruit. This is the greatest and most quantifiable consequence we have seen due to climate change.

Profiling tannin extraction during fermentation is especially useful because changes in sugar, alcohol, and acidity are all interfering with your ability to perceive astringency (Sáenz-Navajas et al., 2010). We also find protein-precipitable tannin in finished wines to be exceptionally useful in evaluating wines by style, vintage, and producer. Below are reference levels for protein-precipitable tannin measured in finished wines.

<500 ppm is a low amount of tannins found in Pinot Noir and commercial reds.
500-1000 ppm is a classic range for Bordeaux wines.
1000-1500 ppm is a classic range for Napa and Super Tuscan wines.
1500-2000 ppm is a range typical for reserve wines from Napa.
>2000 ppm is achieved by heavily extracting mountain Cabernet Sauvignon from Napa or highly tannic cultivars such as Sagrantino, Aglianico, and Nebbiolo.

Note: If you are targeting a protein-precipitable tannin concentration during fermentation, extract approximately 10-20% more than you want in your finished wine because the additional tannin will fall out within the first year.

Polarity

One interesting measurement developed by Dr. James Kennedy was the concept of tannin activity. Think of tannins like post-it notes, the more you restick them, the less sticky they become. This is precisely what Dr. Kennedy has set out to measure. By passing tannins through a hydrophobic HPLC column, he could quantify stickiness as a measure of enthalpy. Tannins decrease in stickiness the more they oxidize or adhere to other molecules. This is fundamentally due to a change in polarity, and this is especially the case for bound anthocyanin formation. Bound Anthocyanins increase tannin polarity, decrease tannin hydrophobic bonding, and make tannins more soluble and less astringent. This is of huge value to winemakers whose grapes are highly tannic. If the grapes also have high color, then anthocyanins will naturally refine tannins under the appropriate conditions. If the grapes don't have high color, then the tannins need to be softened through other means, primarily oxidation. Once you understand this, you see how cultural practices have evolved to balance tannins within these constraints. Extended barrel aging, oxidative winemaking, and warmer primary/secondary fermentation temperatures are all meant to integrate astringency into the mouthfeel of the wine.

Size

Perhaps the most common misunderstanding about wine tannins is the relationship between tannin size and astringency. We often hear speak of soft, polymerized tannins that develop as they age. This is misleading. Not only are larger tannins more astringent, but they also tend to decrease in size with wine age. The average size of tannin polymers in wine is expressed as the mean degree of polymerization (mDP). This includes monomers (mDP = 1), oligomers (2 ≥ mDP <5), and polymers (mDP ≥ 5). One study showed that 21 Bordeaux wines between 3 and 26 years old all had an mDP between 1 and 3 (Drinkine et al., 2007). This is in contrast to grape skin tannins which have mDPs ranging from 34-86 (Morata, 2018). Once extracted into wine fermentation, larger tannin molecules undergo acid catalysis creating small polymers (Smith, 2013). On the other hand, colloids are aggregates of bound anthocyanins, proteins, polysaccharides, and other subunits that snowball into larger and larger macromolecules as wine ages until they eventually precipitate out of solution. The incorporation of other molecules increases the polarity of the tannins thus decreasing their astringency.

Composition

Figure 1. Monomeric tannin structures common to V. vinifera

Monomeric tannin composition in grapes and wine consists primarily of the following 4 compounds: catechin, epicatechin, epigallocatechin, and epicatechin-3-O-gallate. This sounds simple, but the total number of unique combinations of these tannins is likely many multiples greater than 65,532 (Adams & Harbertson, 1999). Their combinations are magnificent in their complexity, and their respective concentrations in grapes and wines are largely due to variables outside of a viticulturist's or winemaker’s control. For example, the only significant way to selectively extract skin versus seed tannins is by controlling maceration length. Grape skins have more epigallocatechin which will rapidly extract at the start of fermentation and grape seeds have more catechin which requires more alcohol and time to extract. The ultimate luxury for a winemaker is to go dry on skins and to choose when to press off by evaluating the mouthfeel and tracking the tannin profile. It’s important to note that even though the wine has completed fermentation, there are still a lot of suspended solids that will mask the underlying astringency. This character is highly specific to site, cultivar, producer, etc. For this reason, we provide phenolic ratios as a reference point for winemakers. If protein-precipitable tannin is a measure of structure, then the ratio of bANT:pTAN is a measure of astringency.

Conclusion

In summary, the diversity of tannin polymers in wine is infinitely complex and surely one of the truest chemical markers of terroir that winemakers have little influence on. Winemakers can focus on managing protein-precipitable tannin concentration to establish structure, form bound anthocyanins to modulate astringency, and incorporate oxygen to decrease tannin activity. Applying this knowledge through and beyond the wine production process lends us powerful insight into its behavior. Rather than seeking to manipulate wine’s complexity, we can observe and work alongside it to reveal its potential. Tannins will always remain one of wine’s greatest mysteries, and the human palate wine’s greatest analytical tool. Below is our list of key variables involved in structural balance.

Timing: Tannins will not be extracted without alcohol present. Winemakers can selectively extract between skins and seeds by controlling maceration length. The degree of tannin polymerization will decrease with extended maceration.

Temperature: Higher temperatures increase the rate of total extraction as well Bound Anthocyanin formation.

Oxidation: Polymeric pigment formation increases tannin polarity (less astringent) whereas greater tannin mDP decreases polarity (more astringent).

Matrix: Higher concentrations of bound anthocyanins, alcohol, pH, sugar, mannoproteins, polysaccharides, and pectins all decrease astringency.

Movement: Short and frequent pumpovers can moderate temperature distribution and decrease extraction. Long and infrequent pumpovers will exaggerate temperature gradients and increase extraction.

References

Colantuoni, G., McLeod, S. WINEXRAY LLC. https://www.winexray.com/

Adams, D. O., & Harbertson, J. F. (1999). Use of Alkaline Phosphatase for the Analysis of Tannins in Grapes and Red Wines. American Journal of Enology and Viticulture, 50(3), 247–252. https://doi.org/10.5344/ajev.1999.50.3.247

Drinkine, J., Lopes, P., Kennedy, J. A., Teissedre, P.-L., & Saucier, C. (2007). Ethylidene-Bridged Flavan-3-ols in Red Wine and Correlation with Wine Age. Journal of Agricultural and Food Chemistry, 55(15), 6292–6299. https://doi.org/10.1021/jf070038w

Mercurio, M. D., & Smith, P. A. (2008). Tannin Quantification in Red Grapes and Wine: Comparison of Polysaccharide- and Protein-Based Tannin Precipitation Techniques and Their Ability to Model Wine Astringency. Journal of Agricultural and Food Chemistry, 56(14), 5528–5537. https://doi.org/10.1021/jf8008266

Morata, A. (2018). Red Wine Technology (1st ed.).

Sáenz-Navajas, M.-P., Campo, E., Fernández-Zurbano, P., Valentin, D., & Ferreira, V. (2010). An assessment of the effects of wine volatiles on the perception of taste and astringency in wine. Food Chemistry, 121(4), 1139–1149. https://doi.org/10.1016/j.foodchem.2010.01.061

Smith, C. (2013). Postmodern Winemaking: Rethinking the Modern Science of an Ancient Craft. University of California Press.

PROTEIN-PRECIPITABLE TANNINS