Abstract Misha Kwasniewski | Ryan J. Elias | Alex J. Fredrickson | Helene Hopfer | Yanxin Lin | Cynthia Loi | Ezekiel Warren

A Second Look at What We “Know” about Tannins in Wine and How They Are Perceived

Misha Kwasniewski,* Ryan J. Elias, Alex J. Fredrickson, Helene Hopfer, Yanxin Lin, Cynthia Loi, and Ezekiel Warren
*Penn State, 326 Rodney A. Erickson Food Science Building
University Park, PA, 16802 (mtk5407@psu.edu)

Despite significant advances in tannin research, we cannot consistently predict tannin sensory characteristics (e.g., roughness, drying, silkiness) in all wines or predict their extraction, retention, and elongation during vinification with any certainty or reproducibility. Some of these issues result from oversimplification of a complex system. From both a perceptual and a winemaking perspective, tannin interactions are often explained using the protein precipitation model, where all proteins (and tannins) are assumed to behave similarly, with any protein equally precipitating any tannin. However, protein and tannin structures do greatly matter. While specific proteins like F6HAU0 (a vascular invertase protein) and A0A438IRV9 (ß-fructofuranosidase) are reduced in wine by tannin additions, other prevalent proteins, such as those involved in Vitis lipid transfer, correlate poorly with tannin retention. Similarly, human saliva interacts with a different profile of tannins than common reagents used in precipitation assays (e.g., bovine serum albumin). This explains why, while these assays correlate with astringency, they often cannot explain sensory differences. Studying astringency in another high-phenolic product, cocoa, we found that test subjects perceived astringency differently and the interaction of specific salivary proteins with phenolics varied greatly by individual. These findings reinforce the idea that the complex structural differences within the tannin and protein categories contribute to their functionality. To better characterize tannin structure diversity, we have developed liquid chromatography with tandem mass spectrometry methods that can measure intact fragments of complex procyanidins post-chromatographic separation as well as elongation products caused by ethylene bridging. This comprehensive phenolic fingerprint allows us to differentiate wines and other high-phenolic foods with similar total phenolic concentrations. In the future, we expect to resolve the uncertainties in tannin evolution during winemaking and tannin perception by characterizing the diversity within tannins, proteins, and polysaccharides.

Funding Support: PA Wine Marketing and Research Program- Pennsylvania Liquor Control Board