Liquid chromatography coupled to high-resolution mass spectrometry for nut classification and marker identification

Abstract

Fraud in nut and seed products poses an economic deception and a threat to human health because of their allergens. This study comprehensively evaluated the metabolomic diversity of ten different nut types through non-targeted liquid chromatography coupled to high-resolution mass spectrometry (LC−HRMS). First, LC−HRMS fingerprints were subjected to partial least squares regression-discriminant analysis (PLS-DA), and the developed multi-class model reached a classification accuracy of 100% after external validation. Then, variable importance in projection (VIP) scores obtained from two-input class PLS-DA models (i.e., a specific nut type against all the other samples) allowed the selection of 136 discriminant compounds that were tentatively annotated/identified through HRMS data. Finally, as a case of study, successful detection and quantitation of almond-based products adulteration (with hazelnut or peanut) was achieved through a targeted LC−HRMS study, using some of the found markers and partial least squares (PLS) regression. In this context, new profiling approaches could be further implemented based on the reported markers using cheaper techniques.