Mixed cultures of Oenococcus oeni strains: A mathematical model to test interaction on malolactic fermentation in winemaking

The effect of mixed culture fermentation of Saccharomyces cerevisiae and Saccharomyces cerevisiae var. diastaticus on fermentation parameters and flavor profile

Belgian Saisons and Lambics are two well-known examples in the brewing industry of mixed fermentations, combination of two or more yeast and/or bacteria strains. The purpose of this study was to determine the impact different pitch rates of Saccharomyces cerevisiae (traditional brewing yeast) and S. cerevisiae var. diastaticus (a variant associated with Belgian styles) had on the fermentation kinetics and concentration of the volatile compounds in the finished beers. A series of brews were performed utilizing ratios of S. cerevisiae and diastaticus. The fermentations were heavily monitored, and a model was used to fit fermentation variables. It was found that mixed fermentations produced behaviors that were predictable and proportional to the mixture ratios. As expected, the pure cultural fermentations of diastaticus had a slower fermentation midpoint (M) at 45.45 h versus 28.28 h for S. cerevisiae with the mixed ones falling in between the two. Flavor and aroma play a key role in the acceptability of beer. The mixed fermentations showed a combination of the two different yeast strains aromatic profiles. When combined, there was a strong linearity between alcohols (R2  = 0.94), esters (R2  = 0.89), and the overall total (R2  = 0.91) volatile compounds. PRACTICAL APPLICATION: Modeling is a widely utilized tool in several different fields. The purpose of this research is to apply modeling techniques to describe the fermentation speed and flavor profile of a mixed fermentation between S. cerevisiae and diastaticus. The equations from this data can be used by brewers for product development purposes to make beers with certain flavor profiles within a desired timeframe.

Wine Microbiology and Predictive Microbiology: A Short Overview on Application, and Perspectives

Predictive microbiology (PM) is an essential element in food microbiology; its aims are the determination of the responses of a given microorganism combining mathematical models with experimental data under certain environmental conditions, and the simulation a priori of the growth/inactivation of a population based on the known traits of a food matrix. Today, a great variety of models exist to describe the behaviour of several pathogenic and spoilage microorganisms in foods. In winemaking, many mathematical models have been used for monitoring yeast growth in alcoholic fermentation as well as to predict the risk of contamination of grapes and grape products by mycotoxin producing fungi over the last years, but the potentialities of PM in wine microbiology are underestimated. Thus, the goals of this review are to show some applications and perspectives in the following fields: (1) kinetics of alcoholic and malolactic fermentation; (2) models and approaches for yeasts and bacteria growth/inactivation; (3) toxin production and removal.

Genome-Scale Reconstruction of the Metabolic Network in Oenococcus oeni to Assess Wine Malolactic Fermentation

Oenococcus oeni is the main responsible agent for malolactic fermentation in wine, an unpredictable and erratic process in winemaking. To address this, we have constructed and exhaustively curated the first genome-scale metabolic model of Oenococcus oeni, comprising 660 reactions, 536 metabolites and 454 genes. In silico experiments revealed that nutritional requirements are predicted with an accuracy of 93%, while 14 amino acids were found to be essential for the growth of this bacterial species. When the model was applied to determine the non-growth associated maintenance, results showed that O. oeni grown at 12% ethanol concentration spent 30 times more ATP to stay alive than in the absence of ethanol. Most of this ATP is employed for extruding protons outside of the cell. A positive relationship was also found between specific consumption rates of fructose, amino acids, oxygen, and malic acid and the specific production rates of erythritol, lactate, and acetate, according to the ethanol content of the medium. The metabolic model reconstructed here represents a unique tool to predict the successful completion of wine malolactic fermentation carried out either by different strains of Oenococcus oeni, as well as at any particular physico-chemical composition of wine. It will also allow the development of consortium metabolic models that could be applied to winemaking to simulate and understand the interactions between O. oeni and other microorganisms that share this ecological niche.