Differentiation of Geographical Origins for Cabernet Sauvignon Wines

Food adulteration and mislabeling may pose potential health risks and trust issues to consumers, particularly for high-valued wine products. The current analytical methods and anticounterfeit labeling techniques are insufficient to determine the identity and point of origins for these products. This application note describes a metabolomic profiling method to trace the origins of wine products based on the work reported by Liang, et al. [1]. The reference Cabernet Sauvignon wines collected from five wineries (two from the US and three from China) were initially analyzed using an Agilent UHPLC-Q-TOF/MS platform under accurate TOF/MS scanning mode. The raw data obtained were subjected to data mining using find-by-molecular-feature extraction. The results were imported into Agilent Mass Profiler Professional (MPP) chemometric software to find the characteristic compounds among the groups. Principle components analysis and clustering analysis using the obtained differential compounds demonstrated an ability to separate the two groups of US wines from wines made in China. A partial least square differentiation analysis (PLSDA) model based on them can predict the wine groups with high accuracy. Using the customized polyphenol compounds database and library for wine and other available Agilent PCDLs, twenty-three compounds were tentatively identified, and most of them were endogenous metabolites in grapes, suggesting that the grape metabolites contribute to the major characteristics of wines from different points of origin. The work demonstrates that the metabolomic profiling approach by combining UHPLC-Q-TOF/MS with chemometric analysis is promising in tracing the geographical origins of grape wines.