Revealing the yeast modulation potential on amino acid composition and volatile profile of Arinto white wines by a combined chromatographic-based approach

Abstract

The importance of yeasts in aroma production during wine fermentation is a significant concern for obtaining a wine that appraises a broad number of consumers. For wine producers, wine aroma modulation is an essential issue where the yeasts used during the winemaking process represents a feasible way to improve the complexity and enhance wines specific characteristics. During the fermentation process of wines, yeasts convert grapes sugars into alcohol, carbon dioxide and a large number of secondary metabolites, depending on yeast metabolism, affecting the wine composition, namely its aroma and amino acids (AAs) composition. So, the present work aims to study the effect of different Saccharomyces-type yeasts on the AAs composition and volatile profile of Arinto white wines. To pursue this goal, four white wines from Arinto grapes were fermented with three different commercial yeasts (Saccharomyces bayanus EC1118, Saccharomyces cerevisiae CY3079, Saccharomyces bayanus QA23) and one Native yeast. Arinto wines AAs composition was quantified by HPLC-DAD, after a derivatization step to obtain the aminoenone derivatives. The volatile content of Arinto wines was determined by GC/MS, after an HS-SPME extraction. Results showed significant differences among the AAs content and volatile profile in the Arinto wines. The higher AAs content was found in the Arinto wines fermented with the CY3079 yeast (470.74 mg•L−1), and the lowest content of AAs in the Arinto wines fermented with EC1118 yeast (343.06 mg•L−1). Native yeast results in wines with a volatile profile richer in esters compared to the other sample wines. Principal component analysis (PCA) obtained with combined data of AAs and volatile compounds, after normalization, for each Arinto wine samples, shows a clear separation of wines fermented with Native and CY3079 yeasts in relation to QA23 and EC1118 fermented wines . The first and second principal components are responsible for 44.40% and 32.20%, respectively, of the system’s variance, which clearly showed a differentiation among wines.