ORIGINAL ARTICLE Novel aspects of grape berry ripening and post-harvest withering revealed by untargeted LC-ESI-MS metabolomics analysis Ketti Toffali • Anita Zamboni • Andrea Anesi • Matteo Stocchero • Mario Pezzotti • Marisa Levi • Flavia Guzzo Received: 1 September 2010 / Accepted: 7 November 2010 / Published online: 26 November 2010 Ó Springer Science+Business Media, LLC 2010 Abstract We established a step-by-step, experiment- guided metabolomics procedure, based on LC-ESI-MS analysis, to generate a detailed picture of the changing metabolic profiles during late berry development in the important Italian grapevine cultivar Corvina. We sampled berries from four developmental time points and three post- harvest time points during the withering process, and used chromatograms of methanolic extracts to test the perfor- mance of the MetAlign and MZmine data mining pro- grams. MZmine achieved a better resolution and therefore generated a more useful data matrix. Then both the quan- titative performance of the analytical platform and the matrix effect were assessed, and the final dataset was investigated by multivariate data analysis. Our analysis confirmed the results of previous studies but also revealed some novel findings, including the prevalence of two spe- cific flavonoids in unripe berries and important differences between the developmental profiles of flavones and flava- nones, suggesting that specific individual metabolites could have different functions, and that flavones and flavanones probably play quite distinct biological roles. Moreover, the hypothesis-free multivariate analysis of subsets of the wide data matrix evidentiated the relationships between the various classes of metabolites, such as those between anthocyanins and hydroxycinnamic acids and between flavan-3-ols and anthocyanins. Keywords Grape metabolomics Á O2PLS-DA Á Grape berry ripening Á Grape berry withering 1 Introduction The quality and organoleptic properties of wine are strongly influenced by the accumulation of secondary metabolites during berry ripening. The economic value of wine has therefore encouraged many researchers to study secondary metabolite profiles in ripe berries, with GC-MS used to analyze volatiles and LC-MS used to analyze non- volatiles. Most studies in the latter category have focused on one or a small group of related compounds, such as anthocyanins (Wang et al. 2003), procyanidins (Wu et al. 2005), or anthocyanins and flavonols (Mattivi et al. 2006; Downey and Rochfort 2008). A small number of more comprehensive studies have been reported, including one encompassing anthocyanins, flavonols and hydroxycin- namic acids (Guerrero et al. 2009) and another cover- ing anthocyanins, procyanidins, flavonols and stilbenes (Cavaliere et al. 2008). The latter is probably the broadest investigation carried out thus far, involving the simulta- neous detection of 40 different compounds in the berry skins of three table cultivars. Although the static metabolic profiles of ripe berries have received the greatest attention, less studies have looked at how metabolic profiles change during development and ripening. The more complete qualitative-quantitative investigation has been performed on stilbene and viniferin profiles (Gatto et al. 2008), while for other classes of compounds mainly the quantitative Electronic supplementary material The online version of this article (doi:10.1007/s11306-010-0259-y) contains supplementary material, which is available to authorized users. K. Toffali Á A. Zamboni Á A. Anesi Á M. Pezzotti Á M. Levi Á F. Guzzo (&) Department of Biotechnology, University of Verona, Strada le Grazie 15, 37129 Verona, Italy e-mail: flavia.guzzo@univr.it M. Stocchero S-IN Soluzioni Informatiche, Via Salvemini 9, 36100 Vicenza, Italy 123 Metabolomics (2011) 7:424–436 DOI 10.1007/s11306-010-0259-y