Some species of puffer fish have been reported to possess both of tetrodotoxin and saxitoxin, which share one binding site on sodium channels. We purified a novel soluble glycoprotein that binds to these toxins from plasma of the puffer fish, Fugu pardalis, and named puffer fish saxitoxin and tetrodotoxin binding protein (PSTBP). PSTBP possessed a binding capacity of 10.6 ± 0.97 nmol·mg⁻¹ protein and a K_d of 14.6 ± 0.33 nm for [³H]saxitoxin in equilibrium binding assays. [³H]Saxitoxin (10 nm) binding to PSTBPs was half‐inhibited by the presence of tetrodotoxin and saxitoxin at 12 µm and 8.5 nm, respectively. From the results of gel filtration chromatography (200 kDa) and SDS/PAGE (104 kDa), PSTBP was suggested to consist of noncovalently linked dimers of a single subunit. PSTBP was completely deglycosylated by glycopeptidase F, producing a single band at 42 kDa. Two highly homologous cDNAs to each other coding PSTBP (PSTBP1 and PSTBP2, the predicted amino‐acid identity 93%), were obtained from a cDNA library of F. pardalis liver. These proteins consisted to two tandemly repeated homologous domains. The predicted amino‐acid sequences of PSTBP1 and 2 were not homologous to that of saxiphilin, a reported saxitoxin binding protein, or sodium channels, but their N‐terminus sequences were homologous to that of the reported tetrodotoxin binding protein from plasma of Fugu niphobles, which has not been fully characterized. The partially homologous cDNA sequences to PSTBP1 and 2 were also found in expressed sequence tag clones of nontoxic flounders liver. Presumably, PSTBP is involved in accumulation and/or excretion of toxins in puffer fish.

Even though tetrodotoxin (TTX) is a widespread toxin in marine and terrestrial organisms, very little is known about the biosynthetic pathway used to produce it. By describing chemical structures of natural analogs of TTX, we can start to identify some of the precursors that might be important for TTX biosynthesis. In the present study, an analog of TTX, 5,11-dideoxyTTX, was identified for the first time in natural sources, the ovary of the pufferfish and the pharynx of a flatworm (planocerid sp. 1), by comparison with totally synthesized (−)-5,11-dideoxyTTX, using high resolution ESI-LC-MS. Based on the presence of 5,11-dideoxyTTX together with a series of known deoxy analogs, 5,6, 11-trideoxyTTX, 6,11-dideoxyTTX, 11-deoxyTTX, and 5-deoxyTTX, in these animals, we predicted two routes of stepwise oxidation pathways in the late stages of biosynthesis of TTX. Furthermore, high resolution masses of the major fragment ions of TTX, 6,11-dideoxyTTX, and 5,6,11-trideoxyTTX were also measured, and their molecular formulas and structures were predicted to compare them with each other. Although both TTX and 5,6,11-trideoxyTTX give major fragment ions that are very close, m/z 162.0660 and 162.1020, respectively, they are distinguishable and predicted to be different molecular formulas. These data will be useful for identification of TTXs using high resolution LC-MS/MS.

Silencing two cytochome P450 genes reduces steroidal glycoalkaloid content and stops tuber sprouting.

Bufonid anurans of the genus Atelopus contain both steroidal bufadienolides and various guanidinium alkaloids of the tetrodotoxin class. The former inhibit sodium-potassium ATPases, whereas the latter block voltage-dependent sodium channels. The structure of one guanidinium alkaloid, zetekitoxin AB, has remained a mystery for over 30 years. The structure of this alkaloid now has been investigated with a sample of ≈0.3 mg, purified from extracts obtained decades ago from the Panamanian golden frog Atelopus zeteki . Detailed NMR and mass spectral analyses have provided the structure and relative stereochemistry of zetekitoxin AB and have revealed that it is an analog of saxitoxin. The proposed structure is characterized by richness of heteroatoms (C ₁₆ H ₂₅ N ₈ O ₁₂ S) and contains a unique 1,2-oxazolidine ring-fused lactam, a sulfate ester, and an N -hydroxycarbamate moiety. Zetekitoxin AB proved to be an extremely potent blocker of voltage-dependent sodium channels expressed in Xenopus oocytes. The IC ₅₀ values were 280 pM for human heart channels, 6.1 pM for rat brain IIa channels, and 65 pM for rat skeletal muscle channels, thus being roughly 580-, 160-, and 63-fold more potent at these channels than saxitoxin.