Comparison of LC-MS methods for the quantitation of ciguatoxins in fish - A collaborative study

Abstract

Ciguatoxins (CTXs) are marine biotoxins that can contaminate seafood and if consumed can result in ciguatera poisoning (CP). The analysis of CTXs is challenging, as they occur in complex tissue matrices, cause CP symptoms at trace amounts (<1 μg kg^-1), and certified reference materials are not available. Currently, no standard operating protocols exist for sample preparation or instrumental analysis. Five laboratories worldwide participated in this first-time method comparison study, which used sample extracts containing different CTX groups (CTX4A group, CTX3C group, C-CTX-1) to identify factors impacting CTX quantitation using LC-MS/MS and LC-HRMS. Matrix effects were found to significantly influence CTX quantitation, along with factors such as instrument, eluents, or selected precursor ion. CTXs were quantified using commercially available, non-certified CTX1B and CTX3C standards. Analogues of the CTX groups behaved differently with regard to matrix effects and suitable calibrants with differences between laboratories exceeding a factor of 10 in some cases.

Keywords: Ciguatoxins; Intercomparison exercise; Liquid chromatography; Mass spectrometry; Matrix effects; Toxin profile analysis.

Fig. 1

Structures of ciguatoxins representing the three different groups investigated in this study, stereochemistry according to FAO and WHO (2020).

Fig. 2

Calibration curves obtained for CTX1B (solid symbol) and CTX3C (open symbol) by (A,B) Lab A, (C,D) Lab B, (E,F) Lab C, (G,H) Lab D, (I, J) Lab E; panels show the curves for the commercial standards' concentration (left) or standards with qNMR adjusted concentration (right); data points show results of triplicate injections (except Lab D and 5 μg L⁻¹ standard for CTX3C: n = 2); scaling of the axes was kept constant between the left and right panels to better illustrate the impact of the concentrations on the slope of the calibration functions.

Fig. 3

Quantitative results obtained for sample 1 depending on the sample preparation (see Table 1 and Section 2.2 for details) and the laboratory (from left to right Lab A (grey), B (green), C (white), D (yellow), E (blue)) with quantitation based on the calibration functions obtained for (A) the concentration of the commercial standard and (B) the qNMR adjusted concentrations; bars represent the quantitative results obtained via the calibration of CTX1B (blank) and CTX3C (shaded); results are provided as mean, error bar shows standard deviation (n = 3); NA – no data available for Lab A and the CTX3C calibrant as the concentration of the standards was <LOD (11 μg L⁻¹); in panel A, different letters for each sample and calibrant represent a significant difference between the laboratories after pairwise comparison (One Way ANOVA, Tukey post hoc test, p < 0.01), letters are not provided in panel B to avoid redundancy; the reader is referred to the y-axis in panel B showing half the dimension compared to panel A; for detailed data see also Table S9. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 4

Quantitative results obtained for sample 2 depending on the sample preparation, the laboratory (for colour code see Fig. 3), and in the case of Lab E (blue bars) the ion transition used for quantitation with the loss of water ([M+H-H₂O]⁺ / [M+H-3H₂O]⁺, light blue) and the pseudo ion transition ([M+Na]⁺/[M+Na]⁺, dark blue); quantitation based on the calibration functions obtained for (A) the concentration of the commercial standard and (B) the qNMR adjusted concentrations; bars represent the quantitative results obtained via the calibration of CTX1B (blank) and CTX3C (shaded); results are provided as mean, error bar shows standard deviation (n = 3); NA – no data available for Lab A and the CTX3C calibrant as the concentration of the standards was <LOD (11 μg L⁻¹); different letters for each sample and calibrant represent a significant difference between the laboratories after pairwise comparison (One Way ANOVA, Tukey post hoc test, p < 0.01), letters are not provided in panel B to avoid redundancy; results for Lab E and the sodium adduct were not included in statistical data evaluation; the reader is referred to the y-axis in panel B showing half the dimension compared to panel A; for detailed data see also Table S9. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 5

Quantitative (left, in colour) and profile analysis results (right, grey scale) obtained for (A,B) sample 3 (fish curry, meal remnant), (C,D) sample 4 (fried fish fillet), (E,F) sample 5 (fried fish fillet); samples 3 and 4 were prepared according to (Spielmeyer et al., 2021), sample 5 was prepared according to (Murray et al., 2018) (see Table 1 and Section 2.2 for further details); bars in the left panels represent (from left to right) the quantitative results obtained by Lab A (grey), B (green), C (white), D (yellow), and E (blue) for the CTX1B calibrant; results are provided as mean, error bar shows standard deviation (n = 3); bars in the right panels represent profile analysis results obtained by each laboratory for the CTX1B calibrant, segments represent the mean portions (n = 3) of 54-deoxyCTX1B (grey), 52-epi-54-deoxyCTX1B (white) and CTX1B (black); data obtained using standards with qNMR adjusted concentrations as well as the CTX3C calibrant are provided in the Supporting Information (Fig. S1, Fig. S2, Table S10); different letters for each CTX analogue and the total content represent a significant difference between the laboratories after pairwise comparison (One Way ANOVA, Tukey post hoc test, p < 0.01); y-axes of panel C and E are adjusted to the same dimension to enable direct comparison of the results obtained for the same matrix (fried fish fillet) and different extraction methods (C: Spielmeyer et al. (2021), E: Murray et al. (2018)); second y-axes showing the total content are adjusted to twice the scale of the corresponding first y-axes. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

Fig. 6

(A,C) Quantitative and (B,D) profile analysis results obtained for sample 6 for (A,B) the concentrations of the commercial standard and (C,D) the qNMR adjusted concentrations; bars in (A,C) represent the quantitative results obtained by (from left to right) Lab B (green), C (white), D (yellow), and E (blue) via the calibration of CTX1B (blank) and CTX3C (shaded); results are provided as mean, error bar shows standard deviation (n = 3); second y-axes showing the total content are adjusted to twice the scale of the corresponding first y-axes; different letters for each CTX analogue and the total content represent a significant difference between the laboratories for the respective calibrants after pairwise comparison (One Way ANOVA, Tukey post hoc test, p < 0.01); bars in (B,D) represent profile analysis results obtained by each laboratory using the (B) CTX1B and (D) CTX3C calibrant, segments represent portions of 2,3-dihydroxyCTX3C (grey) and its 49-epimer (grey shaded), 51-hydroxyCTX3C (black), CTX3C (white) and its 49-epimer (white shaded); for each CTX analogue, different letters represent a significant difference between the laboratories after pairwise comparison (One Way ANOVA, Tukey post hoc test, p < 0.01); for Lab A, no results were provided as no peaks were detected by this participant; for detailed data see also Table S12.