The goal of this study is the metabolomic analysis of ten commercial wine yeast strains that show differences in fermentation behavior and in final ethanol yield. Five of these strains tend to give lower and five higher ethanol yields during grape juice fermentation. Since non-sterile cold-settled juices were to be used to mimic wine production conditions, genetically marked strains of each commercial strain were generated. These strains were marked with a drug resistance gene that encodes resistance to kanamycin or G418. Initial experiments compared the fermentation performance of these strains to their parental strains and there were no differences noted. The kanamycin resistance gene was inserted into the HO gene which is required for yeast mating and not expressed in diploid cells thus it presents an attractive location at which to insert foreign marker DNA. The strains were grown in six replicate fermentations for the metabolomic analysis. Ethanol yields and dominance of the fermentation by the strains were monitored. One strain was not able to complete the fermentation of this juice but it was still taken for metabolomic analysis as this may reveal why it eventually arrested fermentation. Comparisons of the metabolomics data for high and low ethanol-producing strains suggests that low ethanol producers maintain active growth longer than high ethanol producers thus directing more sugar carbon towards biosynthetic intermediates and cellular components than high producers. High ethanol producing strains also showed evidence of the accumulation of degradation products of plant cell wall polysaccharides. This has not been observed before and offers the possibility that some of these components may be catabolized as well at this time which would boost yields of ethanol even further. Analysis of the metabolite pool levels of strains with intermediate ethanol production is currently underway.

It is generally assumed that bottled wine is a dynamic system. A proper wine bottle closure will allow a dynamic and healthy gas exchange between the wine the air. The amount of oxygen ingression through the wine closure can directly affect wine aging and flavor development. Too much oxygen ingress can oxidize the wine during post-bottle aging, whereas too little oxygen ingress may slow flavor development. It is a concern that screw caps with too low oxygen permeability may even induce ?reduced? sulfur off-flavor due to accumulation of volatile thiols. To test this hypothesis, natural cork (up and down), synthetic, and screw caps with Saran-Tin, Saranex and low density polyethylene (LDPE) lines with various oxygen permeability were investigated on both Pinot noir and Chardonnay wines over two years storage. For Chardonnay wine, the LDPE screw cap gave the highest dissolved oxygen, lowest free SO2, and highest absorbance at 420 nm. The Saran-Tin screw cap gave the lowest dissolved oxygen. Similar trends were observed for Pinot noir wines, with Saran-Tin screw cap gave the lowest dissolved O2 and highest free SO2 and total SO2. Decreasing trend was observed for some sulfur compounds during 2 years of storage. Methanethiol had the highest decrease in wines with LDPE screw cap and synthetic closures, correlated with high oxygen permeability of these closures and low SO2 content in the wines. Similar pattern was observed for H2S. On the other hand, a decreasing trend was observed for ethylthiol acetate due to hydrolysis, and was not affected by closure types. Other sulfur compounds remained stable or decreased slightly, and there was no elevated sulfur compound for any screw cap closures after two years of storage. In addition, we did not detect any significant amount of dimethyl disulfide and dimethyl trisulfide in any of the experimental wines. Expert sensory panel did not pick up sulfur off-flavor in the experimental wines, and trained sensory panelist did not pick up any aroma difference between the Saran-Tin and LDPE screw capped wines. These results suggest the original wine status was probably more important, and the data does not support the hypothesis that dimethyl disulfide could converted to methanethiol under low oxygen conditions.

Very good progress has been made in understanding the Fenton reaction. Data collected in Denmark in collaboration with Skibsted revealed the identity of major free radical in wine, demonstrating that the Fenton reaction is important and the role of sulfite in slowing this reaction. Studies of additives to prevent oxidation are currently underway. This research has stimulated many other efforts in wine oxidation around the globe.

This project attempts to clarify the consequences for wine flavor that result from harvesting fruit at different maturities. It aims to do a complete study (from a sensory and chemical analyses point of view) from the vineyard to the resultant wines. During the current funding period we did the descriptive sensory analysis of the 2006 wines, did chemical analyses of the 2006 wines and had a winemaker panel evaluate the quality of a subset of the 2006 wines. We harvested the 2007 grapes from a single vineyard at six time-points over a fourteen week period (August 23 to November 28, 2007). The samples spanned maturity levels from what would be considered early/normal harvest (average Brix=21) to late harvest (average Brix=30.8). The fresh grapes were evaluated by a trained descriptive sensory panel and by Napa Valley winemakers. The trained panel found subtle but consistent differences among grapes harvested at different times and these results were very similar to those found in 2006. Specifically, lower Brix grapes were more sour in skin and pulp, had more vegetative skins and had greener, more bitter seeds, whereas higher Brix grapes were sweeter, more fruity, more squishy with browner and nuttier seeds and more raisin-flavored skins. The Napa winemakers each used their own method to determine whether the sample grapes were ready to be harvested and in 2006 as well as 2007 they all agreed that grapes from harvest 1 through 3 were not ready to be picked. However, for harvests 4 through 6 there were some differing clusters of winemakers. Additionally, grape berry samples were prepared and frozen for chemical analyses. There were differences between the 2006 and 2007 grapes in the concentration of tannins, anthocyanins and 2-methoxy-3isobutyl pyrazine concentrations. The 2007 wines were made and will be evaluated by a trained sensory descriptive panel (in April) and Napa Valley winemakers (in June). The descriptive analysis of the 2006 wines showed that wines made from grapes with higher initial sugar content were higher in dark fruit, dry fruit and spicy flavors and lower in sourness as well as lower in fresh and cooked vegetative flavors. Additionally, for higher initial sugar concentration musts adding water prior to fermentation had a larger effect on the wine sensory attributes than adding water after fermentation. The winemakers thought that wines made with grapes harvested between 24.7 and 26 Brix were of better quality, and would receive the highest retail price. Wine chemical analyses of 2006 and 2007 wines are on-going.

There was extensive progress towards unraveling the key reactions in the oxidation of ethanol with hydrogen peroxide. Iron and its oxidation state are proving to be critical to the reaction. Paradoxically, the addition of oxygen to a simple Fenton reaction nearly stops the oxidation of ethanol. When sulfites are present, phenolics seem to enhance the oxidation, also unexpected because phenolics act as antioxidants on their own. The new understanding of oxidation revealed here is beginning to suggest means to better control oxidation and the next few months of this project should give some more direct guidance.