Occurrence of Cyanotoxins in Mineral Water Sources and Hot Springs from NW Iberian Peninsula ^†

Abstract

Balneotherapy can cause adverse reactions to the usual doses of application of treatments, and consists of a nonspecific clinical picture, the so-called “thermal crisis” or “balneointoxication”. Despite its clinical similarity (gastric discomfort, hepatic congestive outbreaks, cutaneous reactions, etc.) with that observed in acute cyanotoxin poisonings, a thermal crisis has never been associated with the abundant growth of potentially toxic cyanobacteria in mineral water sources. The aim of this work was to verify the hypothetical involvement of cyanotoxins in this clinical picture. Several samples (21) of 18 sources, representative of the different mineral–medicinal waters present in Galicia (northwest Spain) and mostly used for balneotherapy, were collected during September 2018. Samples were filtered and the algae retained were extracted with sonication using acidified methanol and analysed with LC-ESI-HRMS. The target analysis of the cellular matrix samples (limit of detection = 0.01–0.05 µg L⁻¹) did not show nodularin or any of the microcystins (MCs) for which standards are available (MC-dmRR, RR, dmLR, YR, LR, WR, LA, LY, LW and LF). The presence of other MCs, nodularins and related cyanobacterial peptides (CPs) was observed with a suspect screening analysis of cyanotoxins, based on an HRMS home-made database of 157 MCs, 10 nodularins, cylindrospermopsin and 29 CPs previously described in the literature. Signals not referenced in the literature were identified as CPs. Based on HRMS and restrictive criteria (accuracy, isotopic pattern, diagnostic fragments, elements considered, charge, ring plus double bond equivalents and nitrogen rule), the signals were confirmed. In summary, 12 MCs, 2 nodularins and 2 CPs were qualitatively detected. A list of all tentatively identified cyanopeptides in each sample was reported, including the retention times, ion signal type, proposed molecular formula, theoretical m/z, samples where each signal was detected, mass accuracy of measures and their isotopic pattern scores. CP-2 was detected in 14 samples, and it was confirmed as a signal from a cyanobacterial peptide, but with more unsaturations than analogous MCs. The presence of MC-OiaA and MC-OiaAba in three samples was also noteworthy. In addition, [seco-2/3]NOD-R was detected in five samples.