A multidisciplinary Approach to Evaluate Pyrrolizidine Alkaloids and their N_oxide Metabolites presence in Bee pollen: Conventional and Innovative methods

Bee pollen production is gradually increasing due to demand for dietary supplements and therapeutic natural products. However, bee products might be unsafe for human consumption, if they originates from plants (families Asteraceae, Boraginaceae and Fabaceae) producing secondary defense metabolites as pyrrolizidine alkaloids (PAs) and their N-oxide metabolites (PANOs). Monitoring PAs-PANOs in pollen, not required by the European legislation, could be useful to protect human health, in particular children and strong consumers. The aim of the study was to evaluate plants involved in the PAs-PANOs presence in bee pollen, applying classical technologies, as liquid chromatography coupled to mass spectrometry (LC-MS/MS) and melissopalynological analysis, with modern techniques as the DNA barcoding approach. Bee pollen samples, purchased from beekeepers of the Veneto region (North Italy), supermarkets or online stores (n=84), were analysed by LC-MS/MS) with a validated method to quantify 17 PAs/PANOs (1). In order to identify the plant source of PAs and PANOs a melissopalynological analysis and a ITS2 region DNA barcoding approach were applied on 36 and on 55-bee pollen samples, respectively. The results obtained by LC-MS/MS analysis indicate different PAs-PANOs contamination profile reflecting the different botanical and geographical origins of bee pollen samples. Most samples from Veneto region contained Senecio-type PAs-PANOs (Asteraceae family, i.e. Senecio vulgaris). On the other hand, Lycopsamine-type PAs-PANOs (Boraginaceae family, i.e. Borago officinalis, Echium plantagineum), were present in most samples from the great retail market. The microscopic analysis of 36-bee pollen confirmed the presence of Echium and Borago pollen grains in 39% of the samples classified as positive by LC-MS/MS and characterized by a Lycopsamine-type alkaloid contamination panel. The 61% of the LC-MS/MS positive samples was classified as negative, false negative samples. On the other hand, only the 3% was false positive samples. The DNA metabarcoding approach showed a good correlation between PAs-PANOs detected by LC-MS/MS analysis and the presence of DNA of producing plant belonging to Asteraceae and Boraginaceae family, with a correct attribution of the positive samples equal to 79%, with a false negative rate equal to 21%. Preliminary results obtained suggest DNA barcoding approach as a promising tool to identify PAs-PANOs in pollen samples. This information could be useful in order to alert beekeepers about the presence of dangerous plants in their area of farming. Acknowledgements to University of Padua for funding this Project CPTA 158894