Pathogenesis-related proteins in Sauvignon Blanc grapes and the influence of their extraction on resultant juice composition and wine protein stability

Abstract

Protein stabilization of white wine is a process whereby proteins which may later give rise to a haze are removed prior to bottling. This is normally achieved by fining with bentonite, a clay material that has a strong affinity for proteins and other larger molecules. It is now well-established that pathogenesis-related (PR) proteins, which are originally derived from grape berries, are mostly responsible for haze formation. In this study, PR proteins in specific Sauvignon Blanc grape tissues were identified and quantified. Changes in the PR protein composition of grape skin and pulp during ripening and in response to environmental stresses (UV exclusion and powdery mildew infection) were determined, followed by evaluation of the impact of these changes on the extraction of PR proteins into juice. In addition, as grape harvesting and processing conditions can influence the extraction of proteins and other grape components that may interact with them, the effects of harvesting and grape processing on juice composition were investigated and protein heat instability in corresponding wines determined. Finally, the effect of skin contact on juice composition was examined in an experiment designed to investigate how the interactions between proteins and phenolics affect the extraction of grape components.

Using liquid chromatography tandem mass spectrometry (LC-MS/MS) technology, the main soluble PR proteins in wine, thaumatin-like proteins (TLPs) and chitinases, were identified as present in both Sauvignon Blanc grape skin and pulp, but not in the seed. The majority of identified proteins were involved in metabolism and energy. Proteins identified in grape seed were less diverse than those identified in grape skin and pulp. The sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) analysis of fractions of protein extracts separated by high performance liquid chromatography (HPLC) indicated the individual elution time of TLPs and chitinases. Thus, the area of these peaks was accordingly used for relative quantification of TLPs and chitinases in protein extracts of grape tissues.

Ripening of Sauvignon Blanc grapes coincided with the increase in phenolic content on a per berry basis in both grape skin and pulp, but the phenolic concentration on per kg fresh weight basis showed a decreasing trend in grape skin during ripening. This suggested that the accumulation of phenolics in skin during ripening was less than the dilution effect caused by berry growth. Tannin was only detected in grape skin and the changes in tannin content was similar to total phenolics. Accumulation of proteins along with the increase of °Brix during ripening was observed in both grape skin and pulp. PR proteins were synthesized and accumulated in both skin and pulp from véraison until harvest. This increase of PR proteins in grape berries was also reflected in the corresponding juice. In addition, the concentration of PR proteins in grape berries showed a positive correlation with the concentration of total proteins.

UV exclusion and powdery mildew infection resulted in decreased phenolic concentration and increased PR protein concentration, respectively, which was in agreement with previous studies. Furthermore, the results showed that the concentration of PR proteins in the resultant juice was predominantly determined by the concentration of PR proteins in grape pulp, but the concentration of PR proteins in the skin and interactions of proteins with phenolics/tannin during the juicing process may modulate the final concentration of PR proteins in juice. The UV exclusion treatment compared to the control treatment resulted in a lower concentration of PR proteins in grape skin, but not in the pulp. However, the corresponding juice showed a higher concentration of PR proteins from UV exclusion treatment suggesting the loss of PR proteins as a result of phenolic-protein interactions in UV exclusion treatment was less than control as the concentration of phenolics in grapes was significantly reduced by UV exclusion. Significantly higher concentration of PR proteins in grapes from highly scarred berries with powdery mildew infection was reflected in the resultant juice. In addition, the lower concentration of tannin in juice from these berries also suggested the loss of tannin as a result of interactions between tannin and proteins.

The harvesting and grape processing experiment showed that the year and vineyard site had a major impact on juice and wine composition. For grapes harvested in the same vineyard and vintage, machine harvesting resulted in a lower concentration of proteins, particularly PR proteins, than hand harvesting with destem, crush and 3 h skin contact. These juices and wines therefore required less bentonite addition for protein stabilization as bentonite requirement had been observed to be positively correlated with concentration of chitinases in juice and wine. The results suggested that the phenolic-protein interaction and juice oxidation, which was greater in mechanical harvesting than hand harvesting may have resulted in the loss of PR proteins. Furthermore, the potential increase in PR proteins as a result of 3 h skin contact was likely having little impact on final concentration of PR proteins in juice. Pressing treatments also affected juice composition since desteming, crushing and 3 h skin contact resulted in greater juice yield and higher concentration of phenolics and proteins (including PR proteins) than whole bunch pressing. Greater extraction of phenolics was also found in juice obtained at higher pressing pressure.

Investigation of the effects of skin contact showed that juice treated with 24 h skin contact had the highest concentration of total phenolics and chitinases using both fresh berries and berries that were chilled for 24 h. Thus, longer skin contact will increase the protein concentration in juice, in particular the haze forming PR proteins, but the degree of this increase might be reduced along with more phenolic compounds co-extracted from skin as a result of phenolic-protein interactions. The effect of such interaction on reducing protein concentration was reflected in decreased concentration of bovine serum albumin (BSA) in extractants after skin extraction. The concentration of phenolic and tannin in skin extracts by using extractants added with BSA was lower than by using extractants without addition of BSA, also indicating the occurrence of phenolic-protein interactions. In addition, compared to mechanically pressed juice in which no tannin was observed, hand-squeezed juice had considerable amount of tannin but lower protein concentration. This suggested that hand squeezing was likely more effective in extracting grape skin components (mainly phenolics) through repeatedly squeezing and rubbing the mixture of skin and pulp, while mechanical pressing was likely more effective to extract grape pulp components (mainly proteins) through continuously applying high pressing pressure.

This PhD project has revealed the distribution of haze forming TLPs and chitinases in Sauvignon Blanc grape skin and pulp. The concentration of PR proteins in juice is predominantly determined by their concentration in grape pulp, but the extraction of phenolics from grape skin might modulate the final concentration of PR proteins in juice through the phenolic-protein interaction during juicing process. The wine protein heat instability has been found positively correlated with the concentration of chitinases. Thus, any practices that can reduce extraction of chitinases into juice are important to improve the protein stabilization in final wine products.