A data independent acquisition all ion fragmentation mode tool for the suspect screening of natural toxins in surface water

Abstract

Among natural freshwater pollutants, cyanotoxins, mycotoxins, and phytotoxins are the most important and less studied. Their identification in surface water is challenging especially cause of the lack of standards and established analytical parameters. Most target methods focus one or a single group of compounds with similar characteristics. Here we present an AIF fast method for the tentative identification of natural toxins in water. Respect to the previous method [1] , it offers higher performances for the acquisition of unknown compounds at low levels for higher number of analytes. The key aspects of the method are: • The qualitative screening DIA-AIF workflow using Q Exactive Orbitrap . Both targeted and suspect screening bases have been combined with online databases and suspect list to retrieve candidates as suspect natural toxins and their metabolites or degradation products. • The in-silico analysis of mass spectrums allowed a fast structural characterization. • The workflow has been finally applied to real samples coming from the Czech Republic, Italy, and Spain allowing the determination of 17 suspect natural toxins, 4 of them confirmed. None toxin passed the limit of 1 µg/L taken from the legislation applied for microcystin LR and arbitrarily extended to all toxins.

Comprehensive suspect screening for the identification of contaminants of emerging concern in urine of Flemish adolescents by liquid chromatography high-resolution mass spectrometry

“…Other techniques such as iterative MS/MS expand on the number of fragmentation spectra generated but increase the analysis time by at least 3-fold (Koelmel et al, 2017). For an increase in the annotation of compounds at CL2 or better further improvements of the available reference mass spectral libraries or of the available standards are needed (Picardo et al, 2021). Furthermore, the application of novel approaches in data processing, such as in silico deconjugation methods, could allow resolving the above-described challenges within the identification of glucuronidated metabolites (Huber et al, 2022).…”