Isolation of Caribbean Ciguatoxin-5 (C-CTX5) and confirmation of its structure by NMR spectroscopy

Abstract

Ciguatera poisoning occurs throughout subtropical and tropical regions globally. The Virgin Islands in the Caribbean Sea is a known hyperendemic region for ciguatera and has been associated with Caribbean ciguatoxin (C-CTX) contamination in fish. An algal C-CTX (C-CTX5) was identified in Gambierdiscus silvae and G. caribeaus isolated from benthic algal samples collected in waters south St. Thomas, US Virgin Islands. The highest CTX-producing isolate, G. silvae 1602 SH-6, was grown at large-scale to isolate sufficient C-CTX5 for structural confirmation by NMR spectroscopy. A series of orthogonal extraction and fractionation procedures resulted in purification of approximately 40 μg of C-CTX5, as estimated by quantitative NMR. A suite of 1D and 2D NMR experiments were acquired that verified the structure originally proposed for C-CTX5. The structural confirmation and successful isolation of C-CTX5 opens the way for work on the stability, toxicology and biotransformation of C-CTXs, as well as for the production of quantitative reference materials for analytical method development and validation. The strategies developed for purification of C-CTX5 may also apply to isolation and purification of CTXs from the Pacific Ocean and other regions.

Keywords: Caribbean ciguatoxin; Ciguatera; LC−HRMS; NMR; structure confirmation.

Figure 1.

Structures of C-CTX1 (Lewis et al., 1998) and of C-CTX5 as proposed by Mudge et al. (2023), as well as of the 56-methyl ketal derivative of C-CTX5 used during purification of C-CTX5.

Figure 2.

LC−HRMS total-ion chromatogram (TIC) of a 1-L trial for C-CTX5 extraction showing: (A) MTBE extract of the filtered medium; (B) the extract from the cellular material on the filter, and; (C) the SPE C18 extract of the medium after the MTBE extraction.

Figure 3.

LC−HRMS TIC (black lines, mass range: m/z 1000–1250) of: (A) the MTBE extract of the G. silvae 1602 SH-6 filtrate, and; (B) purified C-CTX5 following the isolation procedure. The red and blue lines are summed intensities of all isotopologue peaks exceeding 1% intensity relative to the monoisotopic [M + H]⁺ ion at the m/z values (± 5 ppm) of [M + H – 2H₂O]⁺, [M + H – H₂O]⁺, [M + H]⁺, [M + NH₄]⁺, and [M + Na]⁺ for C-CTX5 and C-CTX1, respectively.

Figure 4.

¹H–¹H-coupled spin systems (blue) identified from COSY and TOCSY NMR spectra in the ring-A–D region of C-CTX5, and the observed HMBC correlations.

Figure 5.

Modelled 3-D structure of C-CTX5 showing the ring-A–D ring region of C-CTX5, with the ¹H (red) and ¹³C (blue) chemical shift assignments (ppm) and their observed NOESY correlations.

Figure 6.

¹H chemical shift differences (ppm) between C-CTX1 (Lewis et al., 1998) and C-CTX5 (Table 1), for each assigned resonance.

Figure 7.

Edited HSQC NMR spectrum of C-CTX5 (blue, CH; red, CH₂; 3.0–4.7 ppm) overlaid on the HSQC NMR spectrum of C-CTX1 (black) from the supporting information of Lewis et al. (1998). Black atom numbers are for C-CTX1 from Lewis et al., red and blue atom numbers are shown for correlations of C-CTX5 that differ significantly from those of C-CTX1. Both spectra were from d₅-pyridine calibrated to residual ¹H and ¹³C solvent signals at 7.21 and 123.5 ppm. Adapted with permission from Lewis et al. (1998). Copyright 1998 American Chemical Society.