In Vitro Glucuronidation of Caribbean Ciguatoxins in Fish: First Report of Conjugative Ciguatoxin Metabolites

Abstract

Ciguatoxins (CTX) are potent marine neurotoxins, which can bioaccumulate in seafood, causing a severe and prevalent human illness known as ciguatera poisoning (CP). Despite the worldwide impact of ciguatera, effective disease management is hindered by a lack of knowledge regarding the movement and biotransformation of CTX congeners in marine food webs, particularly in the Caribbean and Western Atlantic. In this study we investigated the hepatic biotransformation of C-CTX across several fish and mammalian species through a series of in vitro metabolism assays focused on phase I (CYP P450; functionalization) and phase II (UGT; conjugation) reactions. Using liquid chromatography high-resolution mass spectrometry to explore potential C-CTX metabolites, we observed two glucuronide products of C-CTX-1/-2 and provided additional evidence from high-resolution tandem mass spectrometry to support their identification. Chemical reduction experiments confirmed that the metabolites were comprised of four distinct glucuronide products with the sugar attached at two separate sites on C-CTX-1/-2 and excluded the C-56 hydroxyl group as the conjugation site. Glucuronidation is a novel biotransformation pathway not yet reported for CTX or other related polyether phycotoxins, yet its occurrence across all fish species tested suggests that it could be a prevalent and important detoxification mechanism in marine organisms. The absence of glucuronidation observed in this study for both rat and human microsomes suggests that alternate biotransformation pathways may be dominant in higher vertebrates.

Figure 1.

Major CYP and UGT enzyme activities characterized in northern Gulf of Mexico fish liver microsomes using specific probe substrates. Heatmaps show the relative depletion rates for CYP and UGT substrates (A and C, respectively) as well as relative formation rates of the corresponding metabolites (B and D, respectively). Scale bars on each panel show the range of observed depletion/formation rates for the represented compound. Depletion rates are expressed as min⁻¹, while metabolite formation rates are expressed as ng mL⁻¹ min⁻¹. Rates were estimated from mean curves of duplicate incubations at 25 °C. Depletion rates could not be estimated for several phase II probe substrates, which are represented by an “X”.

Figure 2.

Formation of C-CTX-1/-2-GlcA conjugate over time during 60 min glucuronidation assay with L. campechanus (RSN1) microsomes. Figure shows an increase in peak area (arbitrary units) of C-CTX-1/-2-GlcA ([M−H]⁻, m/z 1315.6481, ± 5 ppm) following addition of 5.0 μL (black) or 1.5 μL (gray) of C-CTX-1/-2 reference stock solution. Note that substrate volumes are provided, since the concentration of C-CTX stock solution was not determined. Peak areas (arbitrary units) were obtained by manual integration of extracted ion LC-HRMS chromatograms using Instrument Method 1.

Figure 3.

Extracted ion LC-HRMS chromatograms (left) and FullMS spectra (right) of the [M−H]⁻ (top) and [M+NH₄]⁺ (bottom) ions for C-CTX-1/-2-GlcA. The spectra show the cumulative ions taken across both metabolite peaks, and the proposed identities, observed m/z, and mass accuracies (Δ_m) of the base peak for each isotope group are labeled on the figure.

Figure 4.

Formation of C-CTX-1/-2 glucuronide products in liver microsomes prepared from five northern Gulf of Mexico fish species, as well as microsomes from Atlantic salmon, rats, and humans. Figure shows extracted ion LC-HRMS chromatograms for [M−H]⁻ of C-CTX-1/-2-GlcA ([M−H]⁻ m/z 1315.6481, ± 5 ppm) conjugates 1 and 2, which eluted at 8.57 and 9.13 min, respectively, using Instrument Method 2. The intensities of the highest peak in each chromatogram are indicated in the upper right-hand corners (arbitrary units); the scale was fixed for RLM and HLM chromatograms. The (*) labeled peak in the chromatogram from the N. usta microsome incubation is from a double-charged interfering ion and not from a C-CTX-1/-2-GlcA conjugate.

Figure 5.

HRMS/MS spectrum of C-CTX-1/-2-GlcA ([M−H]⁻, m/z 1315.6481) using a collision energy of 90 eV. The HRMS/MS analysis revealed several fragments related to glucuronic acid ([M−H]⁻ m/z 193.0356), including sequential water loss ions (m/z 157.0141 and 175.0248). The spectra was taken across the LC-HRMS chromatogram peak for the major glucuronide product (metabolite 2; RT = 9.13); metabolite 1 (RT = 8.57) spectra (not shown) was confirmed to be similar. The chemical formulae, RDBE, fragment identities, and mass accuracies (in ppm) for the labeled fragment ions are listed in Table 4.

Figure 6.

Extracted ion LC-HRMS chromatograms (±7 ppm, left panels) for [M−H]⁻ of (A) C-CTX-1/-2-GlcA conjugates (m/z 1315.6481) from incubation of the toxin with L. campechanus liver microsomes and (B) the C-56 reduced product (m/z 1317.6637) following reaction with NaBH₄ using the same chromatographic method (Instrument Method 2). Chromatogram C shows separation of the reduction products using optimized chromatography and different stationary phase (Instrument Method 3). The intensities of the highest peak in each chromatogram are indicated in the upper right-hand corners (arbitrary units). The mass spectra (right panels) are from (D) the major isomer of unreacted C-CTX-1/-2-GlcA and (E) the major isomer of the +2H reduced C-CTX-1/-2-GlcA. Mass spectra represent cumulative ions taken across the representative peaks in the respective LC-HRMS chromatograms.

Scheme 1.. Molecular Structure of the Caribbean CTX Epimers C-CTX-1 and C-CTX-2 and Their C-56 Reduced Congeners C-CTX-3 and C-CTX-4^a

^aAdapted from Kryuchkov et al. 2020.