A multilevel LC-HRMS and NMR correlation workflow towards foodomics advancement: Application in table olives.

ABSTRACT

Foodomics employs advanced analytical techniques to provide answers regarding food composition, authenticity control, marker identification and issues related to food quality and safety. Nuclear magnetic resonance (NMR) spectroscopy and chromatography hyphenated to mass spectrometry (MS) are the main analytical platforms used in this field. Nevertheless, they are rarely employed in an integrated manner, and even then, the contribution of each technique remains vague. Table olives (Olea europaea L.) are a food commodity of high economic and nutritional value with an increasing production tendency over the last two decades, which, however, suffers from extensive fraud incidents and quality determination uncertainties. Thus, the current attempt aims towards two axes with the first being the multilevel integration of LC-HRMS and NMR data of the same samples and table olives being the selected matrix. In more detail, UPLC-HRMS/MS-based analysis was compared at different stages within an untargeted metabolomics workflow with an NMR-based study and the complementarity of the two platforms was evaluated. Furthermore, statistical heterospectroscopy (SHY), rarely employed in foodomics, combining the spectroscopic with spectrometric datasets and aiming to increase the confidence level of annotated biomarkers was applied. Amongst these lines, the second parallel axis of this study was the detailed characterization of table olives' metabolome in search for quality markers considering the impact of geographical (from Northern to Southern Greece) and botanical origin (Kalamon, Konservolia, Chalkidikis cultivars), as well as processing parameters (Spanish, Greek). To that end, using deep dereplication tools including statistical methods, with SHY employed for the first time in table olives, different biomarkers, belonging to the classes of phenyl alcohols, phenylpropanoids, flavonoids, secoiridoids and triterpenoids were identified as responsible for the observed classifications. The current binary pipeline, focusing on biomarkers' identification confidence, could be suggested as a meaningful workflow not only in olive-based products, but also in food quality control and foodomics in general.