Untargeted High Resolution Mass Spectrometry for discriminating authentic saffron

Saffron is a valuable and highly appreciated spice derived from the dried red stigmas of the flowers of the cultivated plant Crocus sativus L. It is commonly used as a coloring and flavouring agent in food preparation. In addition, it is a good source of flavonoids, proteins, sugars, vitamins, amino acids, mineral matter, gums, and other chemical compounds that make it a health promoting spice. To produce saffron spice, harvested stigmas are submitted to a mild food processing peculiar of the region of production thus it could be considered a traditional product, which aroma and chemical composition strictly depend on the geographic location of production. Due to its high costs of production, saffron is one of the most expensive spice commercialized across the world, and often susceptible of adulteration. Different plant-derived adulterants have been discovered, and the most frequently involve cut and/or dyed Carthamus tinctorius L. petals and Curcuma longa L. powdered rhizomes (Kanti et al., 2011). Several analytical methods have been reported in literature for the detection of plant adulterants in saffron samples based on different techniques (e.g. Maggi et al., 2011, Zalacain et al., 2005, Tarantilis et al., 2004, Zougagh et al., 2005) although official methods (ISO 3632-1; ISO 3632-2) exist. Very recently, high-performance liquid chromatography coupled to high resolution mass spectrometry (HRMS) was successfully proposed for saffron authentication/traceability according to the geographical origin based on untargeted metabolic fingerprinting (Rubert et al., 2016). In this work, we investigated two HRMS based approaches one using ESI ionization coupled to LC separation and the other based on DART ionization before HRMS detection to assess saffron authenticity by an untargeted metabolic approach. Pure saffron samples and saffron spiked with different amounts of Carthamus and Curcuma adulterants were extracted with a mix of ethanol and water to extract the majority of metabolites and final MS fingerprints obtained were used for authenticity assessment and/or adulteration detection. The produced spectra were then processed via the commercial software Compound Discoverer v.2.1 SP1 (Thermo Fisher Scientific). The detection and grouping of the unknown compounds by setting a mass accuracy <= 5ppm was further accomplished, together with a preliminary statistical analysis of the integrated peak areas. As grouping factor "Type of adulterant" (Carthamus and Curcuma) was set and data were further pretreated by aligning the extracted chromatograms on the respective retention time. In order to filter the compound list to the species most suitable in discriminating pure saffron from the adulterated one, Volcano plots combining the statistical significance of the identified compounds and magnitude of change in the extracted peak areas, were investigated. Edited compounds list constrained by means of p-value thresholds was then subjected to statistical evaluation by PCA. The present communication will provide a comparison of results by using both approaches demonstrating that HRMS in positive polarity combined with a proper statistical treatment might discriminate between authentic and adulterated saffron.