A colorimetric screening method to evaluate the presence of pyrrolizidine Alkaloids and their N-Oxides in bee pollen

Bee pollen is consumed as a food supplement due to its nutritional properties. However, depending on the botanical sources visited by the honeybees, natural toxins such as pyrrolizidine alkaloids can be present in bee products. Pyrrolizidine alkaloids (PAs) and their N-oxides (PANOs) are synthesised by a variety of plant species, mainly from the families Asteraceae, Boraginaceae and Fabaceae, as a mechanism of defence against herbivores. Since PAs/PANOs can cause acute and chronic toxicity in humans after metabolic activation, it is important to determine their presence in bee products. Thus, the aim of this study was to develop a colorimetric method that could allow establishing a possible relationship between the presence of PAs-PANOs and bee pollen colour. Bee pollen samples acquired from supermarkets and online stores (n=32) were analysed by liquid chromatography coupled to mass spectrometry (LC-MS/MS) with a validated method to quantify 17 PAs/PANOs (1). The CIE-L*a*b* colour coordinates of the bee pollen samples in grains and ground grains were recorded in dark conditions using a Konica Minolta CD-600 visible spectrophotometer. The Spearman rank correlation coefficients (rs) between PAs/PANOs and colour coordinates were calculated. Furthermore, a receiver operating characteristic (ROC) analysis, was carried out to define the capability of b* (yellowness) to rank the presence of PAs/PANOs according to a cut-off of 35 μg kg-1 (PAs/PANOs ≤ 35 negative, PAs/PANOs >35 positive). PAs/PANOs were present in 24 (75%) of the samples analysed by LC-MS/MS. The total PAs/PANOs concentration ranged from 1.7 to 3148.9 μg.kg-1. A significant correlation (rs= -0.53; P= 0.002) between b* PAs/PANOs was detected on ground bee pollens. The ROC analysis evidenced that a threshold of b* ≤ 38 could be used as an indicator to predict the presence of a total PAs-PANOs content > 35 µg kg-1 with accuracy, sensitivity and specificity equal to 0.78, 1.00, 0.61, respectively. The colorimetric analysis showed to be simple, fast, non-destructive, highly sensitive and accurate to predict the presence of PAs/PANOs in ground bee pollens. Acknowledgements tothe University of padua for funding this Project CPDA 15889