Comparative Study on the Performance of Three Detection Methods for the Quantification of Pacific Ciguatoxins in French Polynesian Strains of Gambierdiscus polynesiensis

Abstract

Gambierdiscus and Fukuyoa dinoflagellates produce a suite of secondary metabolites, including ciguatoxins (CTXs), which bioaccumulate and are further biotransformed in fish and marine invertebrates, causing ciguatera poisoning when consumed by humans. This study is the first to compare the performance of the fluorescent receptor binding assay (fRBA), neuroblastoma cell-based assay (CBA-N2a), and liquid chromatography tandem mass spectrometry (LC-MS/MS) for the quantitative estimation of CTX contents in 30 samples, obtained from four French Polynesian strains of Gambierdiscus polynesiensis. fRBA was applied to Gambierdiscus matrix for the first time, and several parameters of the fRBA protocol were refined. Following liquid/liquid partitioning to separate CTXs from other algal compounds, the variability of CTX contents was estimated using these three methods in three independent experiments. All three assays were significantly correlated with each other, with the highest correlation coefficient (r² = 0.841) found between fRBA and LC-MS/MS. The CBA-N2a was more sensitive than LC-MS/MS and fRBA, with all assays showing good repeatability. The combined use of fRBA and/or CBA-N2a for screening purposes and LC-MS/MS for confirmation purposes allows for efficient CTX evaluation in Gambierdiscus. These findings, which support future collaborative studies for the inter-laboratory validation of CTX detection methods, will help improve ciguatera risk assessment and management.

Keywords: Gambierdiscus polynesiensis; ciguatera poisoning; ciguatoxins; fluorescent receptor binding assay; liquid chromatography tandem mass spectrometry; neuroblastoma cell-based assay; risk assessment.

Figure 1

Binding dose–response curves of Gambierdiscus polynesiensis TB92 sample extracts using the fluorescent receptor binding assay with an incubation step at 37 °C for 30 min (grey) versus 4 °C for 90 min (black). Data represent the mean ± standard deviation (SD) (each concentration run in duplicate wells) of four independent experiments run on different days. The mean ± SD of EC₅₀ values obtained for TB92 were 349 ± 21 and 359 ± 26 cell mL⁻¹ (n = 4) with coefficients of variation (CV) of 6% and 7% incubated at 37 °C for 30 min and 4 °C for 90 min, respectively.

Figure 2

Binding dose–response curves of CTX3C using the fluorescent receptor binding assay (fRBA). Data represent the mean ± standard deviation (SD) (each concentration run in triplicate wells) of seven independent experiments run on different days (incubation 30 min at 37 °C, synaptosomes at 0.1 mg mL⁻¹, and Bodipy PbTx2 at 1 nM). The mean ± SD of EC₅₀ value for CTX3C was 2.10 ± 0.16 ng mL⁻¹ (n = 7) with coefficient of variation (CV) of 7.6%.

Figure 3

Binding dose–response curves of Gambierdiscus polynesiensis using the fluorescent receptor binding assay (fRBA). (a) RG92-b sample (▲). (b) NHA4-b sample (◆). Data represent the mean ± standard deviation (SD) (each concentration run in triplicate wells) of three independent experiments run on different days. The dotted vertical line corresponds to the maximum concentration of cell equivalent (MCE = 16,000 cell mL⁻¹) for matrix interference. The mean ± SD of EC₅₀ values obtained for RG92-b and NHA4-b were 1926 ± 237 and 255 ± 16 cell mL⁻¹ (n = 3) with coefficients of variation (CV) of 14% and 8%, respectively (incubation for 30 min at 37 °C).

Figure 4

Cytotoxic dose–response curves of CTX3C under OV+ condition (100/10 µM) using the neuroblastoma cell-based assay (CBA-N2a). Data represent the mean ± standard deviation (SD) (each concentration run in triplicate wells) of six independent experiments run on different days. The mean ± SD of EC₅₀ value for CTX3C was 1.50 ± 0.23 pg mL⁻¹ (n = 6), with coefficient of variation (CV) of 15.6%.

Figure 5

Cytotoxic dose–response curves of Gambierdiscus polynesiensis in OV+ conditions (100/10 µM) using the neuroblastoma cell-based assay (CBA-N2a). (a) RG92-b sample (▲). (b) NHA4-g sample (◆). Data represent the mean ± standard deviation (SD) (each concentration run in triplicate wells) of three independent experiments run on different days. The maximum concentration of cell equivalent (MCE = 1904 cell mL⁻¹) for matrix interference was not indicated in both graphs, being outside the concentration range tested for these samples. The mean ± SD of EC₅₀ values for RG92-b and NHA4-g were 2.08 ± 0.46 and 0.15 ± 0.04 cell mL⁻¹ (n = 3) with coefficients of variation (CV) of 4 and 17%, respectively.

Figure 6

Calibration curves of CTX3C using liquid chromatography tandem mass spectrometry (LC-MS/MS). Data represent the mean ± standard deviation (SD) (each concentration run in triplicate) of three independent experiments run on different days (n = 3).

Figure 7

LC-MS/MS chromatograms of ciguatoxin (CTX) standards and Gambierdiscus polynesiensis using the most intense MRM transition for each toxin. (a) Mix of CTX standards; (b) RG92-b sample; (c) NHA4-d sample. Data represent toxin profiles from three independent experiments run on different days (n = 3). As peaks (1) to (5) share the same MRM transitions as CTX3C (Table S3), they are identified as CTX3C isomers.

Figure 8

Comparison of CTX contents in 30 Gambierdiscus polynesiensis samples; each sample tested by the fluorescent receptor binding assay (fRBA), neuroblastoma cell-based assay (CBA-N2a), and liquid chromatography tandem mass spectrometry (LC-MS/MS) in three independent experiments (n = 3) for each method. The distribution of CTX contents measured by the three methods is presented in (a–c), the scatter plots and regression lines of CTX contents between two methods in (d–f), with corresponding correlation coefficients (r²) and slopes listed in each graph.