Extended Targeted and Non-Targeted Strategies for the Analysis of Marine Toxins in Mussels and Oysters by (LC-HRMS)

Abstract

When considering the geographical expansion of marine toxins, the emergence of new toxins and the associated risk for human health, there is urgent need for versatile and efficient analytical methods that are able to detect a range, as wide as possible, of known or emerging toxins. Current detection methods for marine toxins rely on a priori defined target lists of toxins and are generally inappropriate for the detection and identification of emerging compounds. The authors describe the implementation of a recent approach for the non-targeted analysis of marine toxins in shellfish with a focus on a comprehensive workflow for the acquisition and treatment of the data generated after liquid chromatography coupled with high resolution mass spectrometry (LC-HRMS) analysis. First, the study was carried out in targeted mode to assess the performance of the method for known toxins with an extended range of polarities, including lipophilic toxins (okadaic acid, dinophysistoxins, azaspiracids, pectenotoxins, yessotoxins, cyclic imines, brevetoxins) and domoic acid. The targeted method, assessed for 14 toxins, shows good performance both in mussel and oyster extracts. The non-target potential of the method was then challenged via suspects and without a priori screening by blind analyzing mussel and oyster samples spiked with marine toxins. The data processing was optimized and successfully identified the toxins that were spiked in the blind samples.

Keywords: LC-HRMS; marine toxins; method characterization; non-targeted analysis; suspects screening; targeted analysis.

Figure 1

Processing workflow with optimized parameters for (1) quantitative targeted analysis, (2) suspect screening and (3) non-targeted screening of unknowns using liquid chromatography coupled with high resolution mass spectrometry (LC-HRMS) (adapted from Krauss et al. [26] and Nürenberg et al. [28]).

Figure 2

Separation of the different marine toxins in the optimized chromatographic conditions, analyzed on a 5600 quadrupole time of flight mass spectrometer (QTOF) (a) in positive ionization and (b) negative ionization mode.

Figure 3

Matrix effects observed for the tested toxins in mussel and oyster extracts. The dashed line representing the 100% response corresponds to the results obtained in MeOH, used as a reference.

Figure 4

Z-scores obtained after analyzing, by LC-HRMS, the five samples provided by the EURLMB as part of a proficiency testing scheme for lipophilic toxins and DA.

Figure 5

(a) Scores plot of a supervised PCA-DA analysis of the data generated after analyzing contaminated and non-contaminated extracts (MeOH, mussel, and oyster) by LC-HRMS in ESI⁺ (in red: L0; green: L1 and L2; orange: L3 and L4; blue: L5 and L6); (b) Zoom at the corresponding loadings plot: ions circled in blue are the representative features (green dots) of the most contaminated level (L6). Ions corresponding to the toxins of interest are flagged with yellow stars. The three figures reported next to each green dot (such as 508.3/4.3 (378)) represent, respectively, the exact mass, the retention time, and the feature’s area (in brackets).

Figure 5

(a) Scores plot of a supervised PCA-DA analysis of the data generated after analyzing contaminated and non-contaminated extracts (MeOH, mussel, and oyster) by LC-HRMS in ESI⁺ (in red: L0; green: L1 and L2; orange: L3 and L4; blue: L5 and L6); (b) Zoom at the corresponding loadings plot: ions circled in blue are the representative features (green dots) of the most contaminated level (L6). Ions corresponding to the toxins of interest are flagged with yellow stars. The three figures reported next to each green dot (such as 508.3/4.3 (378)) represent, respectively, the exact mass, the retention time, and the feature’s area (in brackets).