Tetrodotoxin and Its Analogues in the Pufferfish Arothron hispidus and A. nigropunctatus from the Solomon Islands: A Comparison of Their Toxin Profiles with the Same Species from Okinawa, Japan

Pufferfish poisoning has not been well documented in the South Pacific, although fish and other seafood are sources of protein in these island nations. In this study, tetrodotoxin (TTX) and its analogues in each organ of the pufferfish Arothron hispidus and A. nigropunctatus collected in the Solomon Islands were investigated using high resolution LC-MS. The toxin profiles of the same two species of pufferfish from Okinawa, Japan were also examined for comparison. TTXs concentrations were higher in the skin of both species from both regions, and relatively lower in the liver, ovary, testis, stomach, intestine, and flesh. Due to higher TTX concentrations (51.0 and 28.7 µg/g at highest) detected in the skin of the two species from the Solomon Islands (saxitoxin was <0.02 µg/g), these species should be banned from consumption. Similar results were obtained from fish collected in Okinawa, Japan: TTX in the skin of A. hispidus and A. nigropunctatus were 12.7 and 255 µg/g, respectively, at highest, and saxitoxin was also detected in the skin (2.80 µg/g at highest) and ovary of A. hispidus. TTX, 5,6,11-trideoxyTTX (with its 4-epi form), and its anhydro forms were the most abundant, and 11-oxoTTX was commonly detected in the skin.


Introduction
Tetrodotoxin (TTX) is a well-studied neurotoxin known for its distribution in pufferfish [1] and marine invertebrates such as snails [2], crabs [3], starfish [4], blue-ringed octopus [5], and sea slugs [6]. Recently, TTX was also detected in bivalve mollusks: New Zealand clams [7], mussels and oysters in England [8], and also in Greek shellfish [9]. Generally, TTX blocks the voltage-gated sodium ion channels, incapacitating nerve conduction and muscle action potentials, causing progressive paralysis and death due to failure of the respiratory system [10,11]. TTX, as the primary agent of pufferfish poisoning, is reported to be produced by marine bacteria and accumulated in the pufferfish via the food chain [12,13]. TTX in newts was reported to be an excellent defense strategy to ward off predators [14], and TTX is used to capture mobile prey by flatworms in Guam [15]. We have found various TTX analogues ( Figure 1) which can be further classified into (1) hemilactal type analogues, (2) 5-deoxy-10,7-lactone type analogues, (3) 4,9-and 4,4a-anhydro type analogues and (4) tetrodonic acid type analogue [16,17]. Based on the structures of these analogues, we predicted a stepwise oxidative pathway of TTX in marine animals [17]. Additionally, we recently found C5-C10 directly bonded TTX analogues in newts, suggesting a monoterpene origin of TTX [18].
Past incidences of pufferfish food poisoning were reported in several countries including Thailand [25], Bangladesh [26], Hong Kong [27], Singapore [28], Japan [29], Suez Canal region [30], Eastern Mediterranean Sea [31], and several cases in the United States originating mainly from Asian dried pufferfish believed to be illegally imported [32,33]. Reports on these global occurrences and its potential expansion indicate that proper research on the specific toxic species and their distribution is of great importance. However, unlike ciguatera and other seafood poisoning, pufferfish poisoning is not well documented in the South Pacific region although several incidences matching its description have been reported in localized rural areas [34]. The Solomon Islands in particular does not have clear records of victims suffering from pufferfish poisoning. Past undocumented incidents may have triggered a precautionary knowledge passed down through successive generations of the potential dangers of pufferfish consumption. However, because fish and other seafood are a source of protein to the citizens of this island nation, and together with socioeconomic pressures that may affect food security and possibly alter food selectivity, we believe this investigation is imperative as well as informative to the citizens regarding potential risk of pufferfish poisoning.
In this study, the pufferfish Arothron hispidus and A. nigropunctatus collected in the Solomon Islands were investigated for TTX and its analogues, as well as STXs for the first time, using high resolution hydrophilic interaction chromatography (HR-HILIC) LC-MS. The toxin profiles of the same two species from Okinawa, Japan, were also examined and compared with those from the Solomon Islands. In addition, two specimens of Diodon holocanthus from the Solomon Islands, were also tested for the presence of TTXs, because the skin, flesh, and testis of D. holocanthus are officially allowed for consumption in Japan [39].

Estimation of the Concentrations of TTX and Its Analogues in Each Organ of A. hispidus, A. nigropunctatus, and D. holocanthus from the Solomon Islands
For three specimens of A. hispidus and A. nigropunctatus, and two specimens of D. holocanthus from the Solomon Islands, TTX and its analogues both in the ethanolic solutions and in the soaked organs were separately quantified using HR-LC-MS after prepurification. The results were combined together, according to solvent quantity and tissue weight, to estimate the real concentrations of TTXs in the organs. 5,6,11-TrideoxyTTX and 4,9-anhydro-5,6,11-trideoxyTTX were present relatively higher in the ethanolic solutions compared with the organs. The results of A. hispidus and A. nigropunctatus (three specimens for each species) from the Solomon Islands are summarized in Tables 1 and 2, respectively. TTX, 5,6,11-trideoxyTTX, 4,4a-anhydro-5,6,11-trideoxyTTX, and 4,9-anhydro-5,6,11-trideoxyTTX were the most abundant analogues in all organs in these pufferfish species, and 11-oxoTTX which is considered a more potent analogue than TTX [23,24] was detected in the skin of all specimens. TTX concentrations in the skin were generally higher in all specimens, and relatively lower in the liver, ovary, testis, stomach, and intestine. The highest TTX concentrations in the skin of A. hispidus and A. nigropunctatus were 51.0 and 28.7 µg/g, respectively. Additionally, we confirmed that TTXs in D. holocanthus from the Solomon Islands were less than the limit of detection (LOD; S/N > 3, 0.01 µg/g) in all the organs tested (skin, liver, gonad, stomach, intestine, and flesh), even though there were only two specimens (a male and a female).   For 11-oxoTTX and 4-epi-11-oxoTTX: LOD = 0.04 µg/g (S/N > 3), LOQ =0.14 µg/g (S/N > 10), for other analogues: LOD = 0.01 μg/g (S/N > 3), LOQ = 0.03 μg/g (S/N > 10). 5,6,11-TrideoxyTTX was measured as the mixture with its 4-epi form.

The concentrations of TTX and Its Analogues in Each Organ of A. hispidus and A. nigropunctatus From Okinawa, Japan
The pufferfish specimens from Okinawa were transported live to Sendai and immediately frozen at −25 °C. TTXs in each organ of three specimens of A. hispidus and four specimens of A. nigropunctatus were extracted and pre-purified with activated charcoal, and were analyzed using HR-LC-MS. The data obtained for A. hispidus and A. nigropunctatus are shown in Tables 3 and 4, respectively. The mass chromatograms and organ distribution of toxins in these Arothron species from Okinawa, Japan, were very similar to those from the Solomon Islands. However, one specimen of A. nigropunctatus from Okinawa (No.1 male specimen) showed very high levels of TTX and 11-oxoTTX in the skin, 255 and 42.4 µg/g, respectively. TTX concentration in the stomach of this specimen (25.5 µg/g) was higher than other organs. 11-OxoTTX was detected in the skin of all specimens of both species, same as those from the Solomon Islands.

Analysis of Saxitoxin (STX) and Its Analogues in A. hispidus and A. nigropunctatus from the Solomon Islands and Okinawa, Japan
The presence of STX, neoSTX, and dcSTX was investigated in the skin and ovary of the pufferfish from the Solomon Islands and Okinawa, Japan, using HR-LC-MS. The concentration of STX was below the LOD (0.02 µg/g) in the skin and ovary of all pufferfish from the Solomon Islands and all four specimens of A. nigropunctatus from Okinawa, Japan. However, STX was detected in the skin (0.09, 2.80, 0.34 µg/g) of all three specimens of A. hispidus from Okinawa, Japan. A female specimen A. hispidus (No. 2) from Okinawa, which possessed highest STX (2.80 µg/g) and dcSTX (0.66 µg/g) in the skin, also contained trace amount of STX (0.04 µg/g: between LOD and LOQ) in the ovary. In other samples, dcSTX was not detected. NeoSTX was not detected in all samples. LOD (S/N > 3) and LOQ (S/N > 10) for STX were 0.02 µg/g and 0.07 µg/g, neoSTX, 0.05 µg/g and 0.18 µg/g, and dcSTX, 0.07 µg/g and 0.25 µg/g, respectively.

Discussion
We estimated the concentrations of TTX and its analogues in the pufferfish, A. hispidus and A. nigropunctatus, from the Solomon Islands and Okinawa, Japan, using HR-LC-MS. The pufferfish organs from the Solomon Islands were transported in ethanol-water (7:3, v/v) at room temperature to Japan. Based on the LC-FLD analysis, the amounts of tetrodonic acid type analogs, likely derived from TTXs under such condition, was estimated to be almost 2/3 of total TTX. Thus, we should note that the amount of TTX in the samples from the Solomon Islands (Tables 1 and 2) could be underestimated. However, we believe that the data obtained in this study must be reported to warn of the high toxicity of these species in the Solomon Islands. The flesh of only one specimen of A. hispidus from the Solomon Islands was analyzed, due to limited number storage tubes and volume of ethanolic solution available during pufferfish collection. TTX in that flesh was low (0.07 µg/g), however, the levels of TTX in the skin of A. hispidus and A. nigropunctatus were high in all specimens of both species. In addition, TTX was detected in the flesh of all specimens of these species from Okinawa, Japan (A. hispidus from 0.23 to 0.61 µg/g; A. nigropunctatus from 0.19 to 4.26 µg/g). Based on these data, these species, including flesh, from the Solomon Islands and Okinawa, Japan, should be banned from consumption. Khora et al. [47] and Teruya et al. [48] also indicated high toxicity in the flesh of these Arothron species from Okinawa, Japan, based on the result of a mouse bioassay.
The distribution of TTXs in the organs of A. hispidus and A. nigropunctatus from the Solomon Islands and Okinawa, Japan, revealed that the skin is the major location of TTX accumulation. The skin of A. hispidus and A. nigropunctatus from the Solomon Islands contained 7.20 to 51.0 µg/g and 8.51 to 28.7 µg/g, respectively, while those from Okinawa, Japan, were 4.26 to 12.7 µg/g (A. hispidus), and 12.6 to 255 µg/g (A. nigropunctatus). These data suggested that the concentration of TTX is highly variable among specimens. Our findings agree with previous reports that showed higher toxicity to mice in the skin tissues of the same species collected in Okinawa area, Japan [47]. A study on the Hawaiian A. hispidus suggested high levels of TTX producing bacteria isolated from the skin [49], although the details have not yet been clarified. The levels of TTX in the liver and gonads are usually high in common Japanese pufferfish species in genus Takifugu (Fugu) [50], but they were low in the Arothron species compared with those of the skin. We assume that different regions, with their different biotic and abiotic factors, may play an important role in the TTX accumulation ability of pufferfish. More specimens are needed to compare the TTX distribution profile in the organs of the pufferfish from the Pacific region. The difference of toxin distribution pattern between Takifugu and Arothron is probably caused by the difference of some proteins which are implicated in TTX accumulation system. We previously found a glycoprotein, named as PSTBP (puffer fish saxitoxin and tetrodotoxin binding protein), that binds to TTX and STX in the blood plasma of Takifugu pardalis [51], and also, it was reported that PSTBP was commonly detected in several Takifugu species [52,53]. We are planning to examine the presence of PSTBP-like protein in Arothron species to get a hint that will help explain the difference of TTX-accumulation tissues between Takifugu and Arothron genus.
We also analyzed TTX and its analogues in several organs of two specimens of D. holocanthus from the Solomon Islands, and confirmed that the concentrations of all TTX analogues were below the LOD (S/N > 3, 0.01 µg/g). We will further analyze more specimens for confirmation in the future. In Japan, the skin, flesh, and testis of D. holocanthus collected in Japan are officially allowed for consumption [39].
The presence of STX and its analogues were also tested in the skin and ovary of A. hispidus and A. nigropunctatus from the Solomon Islands and Okinawa, Japan. STX was detected in the skin of all three specimens of A. hispidus from Okinawa (0.09, 2.80, 0.34 µg/g), as well as in the ovary of one specimen, but not detected in the pufferfish from the Solomon Islands (< 0.02 µg/g). It is notable that the concentration of STX equivalents (STX + dcSTX) was 3.46 µg/g in the skin of an A. hispidus specimen from Okinawa. This is higher than the global regulatory limit of 800 μg of STX equivalents per kg of shellfish meat (e.g. Codex, EU, US, Japan) [60][61][62][63]. Analysis of STX in more specimens from both regions is needed, because Nakashima et al. [36] reported that STX and dcSTX were detected in the ovary of Arothron firmamentum collected in Japan at high concentration. From the results obtained in this present study, we believe this investigation is informative and also important for the people of Solomon Islands regarding potential risk of pufferfish poisoning.

Standards and Reagents
Semi-purified TTXs mixture prepared from the ovary of Takifugu poecilonolus using a charcoal column [21] was used as the standard for TTXs as described previously [17]. The authentic 11-oxoTTX was prepared from TTX by chemical oxidation [24]. TTX used to draw the standard curve for HR-LC-MS was highly purified by our group. The standard mixture of STX, dcSTX, and neoSTX were prepared (approved under AOAC guidelines) by Oshima [64]. All of the solvents for LC and the reagents were purchased from Wako Pure Chemical Industries (Osaka, Japan), while ammonium formate for LC-MS, heptafluorobutyrate acid, and acetonitrile for LC-FLD were purchased from Sigma-Aldrich (St. Louis, MO, USA).

Sampling and Test Materials
Three specimens of Arothron hispidus, three specimens of A. nigropunctatus and two specimens of

Preliminary Analysis of TTXs in the Skin of A. hispidus from the Solomon Islands Using LC-FLD for TTXs
An aliquot (0.5 mL) of the ethanol-water (7:3, v/v) solution used to soak the skin of A. hispidus three specimens from the Solomon Islands was acidified by acetic acid and dried in vacuo, and then dissolved in 0.5 mL of 0.05 M acetic acid. After centrifugation, a part of the supernatant was diluted ten-fold with 0.05 M acetic acid. An aliquot (2 µL) of this solution was applied to LC-FLD for TTXs measurement. Similarly, the skin from these specimens was removed from the ethanolic solution, and a part (0.25 g) was extracted with 1.25 mL of 0.2 M acetic acid (v/v) in boiling water. The extract was centrifuged at 15,000× g and an aliquot (2 µL) of the supernatant was applied to LC-FLD [41,42].

From the Ethanolic Solution Used to Soak the Pufferfish Organs from the Solomon Islands
An aliquot (0.5 mL) of ethanol-water (7:3, v/v) solution (from a total of 14.9 mL for the skin of A. hispidus No.3 for example) was mixed with 5 µL of acetic acid, and dried in vacuo. The obtained residue was dissolved in 0.5 mL of 0.05 M acetic acid and neutralized with 1 M NaOH aqueous solution, and then applied to activated charcoal (column volume: 250 μL, for chromatography, Wako, Osaka, Japan) packed in a glass pipette equilibrated with water. After washing charcoal with 0.75 mL of water, TTXs were eluted with 1.5 mL of acetic acid-ethanol-water (2:50:48, v/v). An aliquot (50 µL) of this eluate was dried in vacuo and dissolved in 50 µL of 0.05 M acetic acid, and then filtered through Cosmospin Filter H (Nacalai Tesque, Inc., Kyoto, Japan). After dilution of the filtrate with 0.05 M acetic acid, an aliquot of this solution was used for HR-LC-MS. The recovery of TTXs in the organs from the activated charcoal was estimated at approximately 70% using LC-FLD. Therefore, the final TTXs concentrations in the organs were recalculated based on this recovery ratio.

From the Organs of Solomon Islands Pufferfish Soaked in Ethanolic Solution, and from the Frozen Organs of Pufferfish from Okinawa, Japan
The organ was homogenized and an aliquot (0.25 g) was mixed with 1.25 mL of acetic acid-water (1:99, v/v). The mixture was boiled for 10 minutes in a 1.5 mL micro tube, and centrifuged at 15,000× g for 15 minutes at 4 °C. Half of the supernatant was neutralized with 1 M NaOH, and then applied to the charcoal (column volume 250 µL). The sample was treated by the same procedure as described above Section (4.4.1).

LC-FLD for TTXs
Post-column LC-fluorescent detection (LC-FLD) for TTXs was performed as described previously [41,42]. Briefly, The LC condition was as follow: Develosil C30 UG-5 (0.46 cm i.d. × 25 cm) (Nomura Chemical, Seto, Japan) with 30 mM ammonium heptafluorobutyrate buffer (pH 5.0) and 10 mM ammonium formate buffer (pH 5.0) containing 1% (v/v) acetonitrile at a flow rate of 0.4 mL/min; for the post column reaction 4 M NaOH at a flow rate of 0.7 mL/min to be heated at 105 °C in the stainless tube (0.46 mm i.d. × 3.5 m). The chromatography was performed at 20 °C. The derived fluorophores by post column reaction were detected by a Jasco FP2025 plus fluoromonitor (Jasco, Tokyo, Japan) setting excitation wavelength at 365 nm and emission wavelength at 510 nm.

Conclusions
The concentrations of TTX and its analogues in the skin was generally higher in both A. hispidus and A. nigropunctatus both from the Solomon Islands and Okinawa, Japan, and relatively lower in the liver, ovary, testis, stomach, intestine, and flesh. Because of high concentration of TTX in the skin of these two species from the Solomon Islands (51.0 and 28.7 µg/g at highest) and Okinawa, Japan (12.7 and 255 µg/g at highest), their consumption should be banned. 11-OxoTTX, which is considered a more potent analogue than TTX by some authors, was commonly detected in the skin of all specimens of both Arothron species from these two regions, an occurrence that should warn citizens of a potentially higher toxicity. STX was detected in the skin of all three specimens of A. hispidus from Okinawa, Japan. The level of STX equivalents (STX + dcSTX) in the skin of one specimen was 3.46 µg/g, more than the global regulation level. The toxicities of other Arothron species from these regions should be continuously studied.