Effect of uncontrolled factors in a validated liquid chromatography-tandem mass spectrometry method question its use as a reference method for marine toxins: major causes for concern.

Chromatographic techniques coupled to mass spectrometry is the method of choice to replace the mouse bioassay (MBA) to detect marine toxins. This paper evaluates the influence of different parameters such as toxin solvents, mass spectrometric detection method, mobile-phase-solvent brands and equipment on okadaic acid (OA), dinophysistoxin-1 (DTX-1), and dinophysistoxin-2 (DTX-2) quantification. In addition, the study compares the results obtained when a toxin is quantified against its own calibration curve and with the calibration curve of the other analogues. The experiments were performed by liquid chromatography (LC) and ultraperformance liquid chromatography (UPLC) with tandem mass spectrometry detection (MS/MS). Three acetonitrile brands and two toxin solvents were employed, and three mass spectrometry detection methods were checked. One method that contains the transitions for azaspiracid-1 (AZA-1), azaspiracid-2 (AZA-2), azaspiracid-3(AZA-3), gimnodimine (GYM), 13-desmethyl spirolide C (SPX-1), pectenotoxin-2 (PTX-2), OA, DTX-1, DTX-2, yessotoxin (YTX), homoYTX, and 45-OH-YTX was compared in both instruments. This method operated in simultaneous positive and negative ionization mode. The other two mass methods operated only in negative ionization mode, one contains transitions to detect DTX-1, OA DTX-2, YTX, homoYTX, and 45-OH-YTX and the other only the transitions for the toxins under study OA, DTX-1, and DTX-2. With dependence on the equipment and mobile phase used, the amount of toxin quantified can be overestimated or underestimated, up to 44% for OA, 46% for DTX-1, and 48% for DTX-2. In addition, when a toxin was quantified using the calibration curve of the other analogues, the toxin amount obtained is different. The maximum variability was obtained when DTX-2 was quantified using either OA or a DTX-1 calibration curve. In this case, the overestimation was up to 88% using the OA calibration curve and up to 204% using the DTX-1 calibration curve. In summary, the correct quantification of DSP toxins by MS detection depends on multiple factors. Since these factors are not taken into account in a validated protocol, these results question the convenience of having MS/MS as a reference method for protecting consumers of marine toxins, moreover if toxicity of each group is considered independently and total toxicity is not summed anymore as it is in the MBA.