Classification of Stevia rebaudiana Chemotypes Based on Steviol Glycosides and Their Identified Malonyl Forms Using Liquid Chromatography Coupled to High-Resolution Mass Spectrometry

The sweetness of Stevia rebaudiana (Bertoni) leaves is attributed to steviol glycosides (SGs). A method to analyze more extensively and decipher the structures of SGs is needed to understand (i) the biochemical determinism of SG diversity and (ii) the variability of their sensorial attributes. SGs from 20 genotypes of S. rebaudiana (Bertoni) were extracted via liquid extraction and tentatively identified by using LC-ESI-HRMS untargeted metabolomics, enabling the annotation of 114 SGs at different confidence levels. We identified the presence of original malonylated-SGs (up to 36%), acetylated-SGs, and other unidentified families, and confirmed the structure of malonyl-Reb A through 1D and 2D-NMR analyses after purification. Targeted identification and absolute quantification of 16 SGs were performed using analytical standards. Statistical analyses were applied to SG-based classification of stevia within 5 classes. This approach, combining targeted and untargeted metabolomics, may facilitate genotype selection for breeding programs and industrial applications based on SG proportions.