No Ability To Produce Tetrodotoxin in Bacteria

Bacterial production of tetrodotoxin (TTX) was first reported by Yasumoto et al. (8). Detection of bacterially produced TTX has been primarily by high-performance liquid chromatography (HPLC) and gas chromatography-mass spectrometry (GC-MS). In September 1995, Applied and Environmental Microbiology published a paper indicating that both the HPLC and GC-MS methods lacked specificity for TTX (3). In that paper, I pointed out that the previously reported production of TTX by bacteria should be revised or reexamined.

In April 2000, Applied and Environmental Microbiology published data of Lee et al. (2) reporting that Vibrio strains isolated from the intestines of puffer fish produce TTX and its derivatives. Among the detection methods used by Lee et al. (2), thin-layer chromatography and electrophoresis had been commonly used through the 1970s. These methods have been little used since the 1980s because the separation of TTX is better accomplished by HPLC and GC-MS. Also, while Lee et al. also used HPLC and GC-MS, the detection of TTX by all four of these methods relies on alkali hydrolysis (2). Therefore, the analytical methods used by Lee et al. are nonspecific for TTX.

In the article of Lee et al. (2), a large TTX peak was shown in Fig. 1 while no toxicity was reported in Table 1. It is, however, well known that a good correlation is observed between TTX concentrations calculated by mouse bioassay and HPLC when TTXs from puffer fish and newt are used (7). Also, the apparent lower limit for TTX detection by HPLC is 10 ng 10 μl⁻¹ (= 1,000 ng ml⁻¹) (7) and that by mouse bioassay is 220 ng ml⁻¹ (1), indicating that TTX detected by HPLC can be easily detected by mouse bioassay. Therefore, the results of Lee et al.'s Fig. 1 and Table 1 may indicate false-positive results by the analytical methods. This possibility had been pointed out in the 1995 paper (3).

There has been no description of the determination of the structures of the toxins obtained from bacteria, although a large amount of TTX should easily be obtained by their cultivation. The chemical structures of TTXs from puffer fish, newt, and blue-ringed octopus have already been reported (6). It has, therefore, been thought even by Japanese scientists that the chemical structures of the substances obtained from bacteria were TTX, although no one has shown these structures. Hence, a structural determination should not be a hard task and is indispensable for confirmation of production of TTX by bacteria.

Production of TTX by bacteria has been presented to support the food chain origin of TTX in puffer fish (8). The food chain origin of TTX was first hypothesized from the observation that cultured puffer fish have no TTX, which suggested that puffer fish have no ability to produce TTX. It has, however, been demonstrated recently that cultured puffer fish have detectable TTX (4) and that these fish can produce TTX (5). The progress in TTX research, including the 1995 paper (3), is not referenced in the article of Lee et al. (2).

As the methods used to detect TTX lack specificity, I cannot agree with the conclusions of Lee et al. (2) that Vibrio strains produce TTX.