Abstract: Toxicity analysis of the puffer fishes Lagocephalus wheeleri and Lagocephalus sceleratus from the East Coast Water of Peninsular Malaysia was carried out. The presence of tetrodotoxin (TTX) in fish tissue and cultures of bacteria isolated from the liver of the fish were determined. Detection of TTX was carried out by mouse bioassay, Thin Layer Chromatography (TLC) and High Performance Liquid Chromatography (HPLC). The bacteria Shewanella sp. was isolated from the liver of L. wheeleri while, Exiguobacterium sp. and Staphylococcus sp. were isolated from the liver of L. sceleratus. Mouse bioassay showed that tissue extracts of L. wheeleri and culture supernatants of Shewanella sp. were positive for TTX. Tissue extracts of L. sceleratus and culture supernatants of Exiguobacterium sp. and Staphylococcus sp. exhibited non-lethal toxicity to mice but the symptoms were not typical of TTX poisoning. The symptoms in mice, coupled with TLC and HPLC analysis indicated that the toxic factor in L. wheeleri and Shewanella sp. was TTX. This study has confirmed the toxicity of the puffer L. wheeleri and some of this toxicity may be attributed to symbiotic bacteria.
INTRODUCTION
There are 185 species and 28 genera of puffer fish in the family Tetraodontidae (Oliveira et al., 2006). According to Sabrah et al. (2006) and Froese and Pauly (2007), these fishes typically occur in tropical seas and estuarine environments. Among the 185 species, Lagocephalus wheeleri and Lagocephalus sceleratus are the two most common puffer fish species in Malaysia and the former are eaten by some locals. These have resulted in several poisoning cases including a few fatalities (Kan et al., 1987). Many puffer fish species contain the neurotoxin tetrodotoxin (TTX) which is one of the most potent marine toxin (Yang et al., 1995). Fugu veKanrmicularis vermicularis is a well known producer of TTX and anhydro-TTX (Jensen and Fenical, 1994; Noguchi et al., 1987; Sugita et al., 1987). Levels of TTX in puffer fishes may vary according to season, geographic location, sex and organ tissue (Noguchi et al., 1997; Nuñez-Vázquez et al., 2000; Yu and Yu, 2002). Most of the toxin is concentrated in the liver, gonads, intestines and skin of these fishes and probably contributed to approximately 60% of fatal cases (Ellenhorn and Barceloux, 1988). However, TTX is not restricted to puffer fish and is widely distributed among various animal classes such as the Californian newt Tarichi torosa (Mosher et al., 1964), the goby Gobius criniger (Noguchi and Hashimoto, 1973), Atelopus frogs (Kim et al., 1975), the gastropod mollusks Charonia sauliae (Narita et al., 1981) and Babylonia japonica (Noguchi et al., 1981; Yasumoto et al., 1981), the xanthid crab Atergatis floridus (Noguchi et al., 1983), the blue-ringed octopus Octopus maculosus (Sheumack et al., 1978), Astropecten starfishes (Maruyama et al., 1984, 1985; Noguchi et al., 1982), the frog shell Tutufa lissostoma (Noguchi et al., 1984) and the small gastropod mollusks Zeuxis siquijorensis (Narita et al., 1984) and Niotha clathrata (Jeon et al., 1984). The occurrence of TTX in these various taxa suggested that the toxin could either be acquired through the food web or is the product of bacterial symbionts in these species (Kosuge et al., 1985; Matsui et al., 1989). Many studies have been carried out to determine the origin of TTX in marine animals (Lee et al., 2000). Results from one study showed that Vibrio fischeri isolated from the xanthid crab Atergatis floridus and Vibrio alginolyticus and Shewanella sp. isolated from the puffer fish were able to produce TTX in pure cultures (Noguchi et al., 2006). The exogenous origin of TTX in puffer fish was first proposed by Matsui et al. (1981) in feeding experiments with artificially bred larvae. Later studies suggested that all TTX-bearing organisms harbour symbiotic bacteria capable of producing TTX (Matsui et al., 1985; Mosher and Fuhrman, 1984) and was later confirmed by isolation of TTX-producing bacteria from various TTX-bearing animals. Puffer fishes are very common in Malaysian water are often caught in large numbers by trawlers or line fishing. Lagocephalus wheeleri is consumed in some parts of the country and puffer fish poisoning is quite frequent (Kan et al., 1987; Kanchanapongkul, 2001; Zaki et al., 2005). However, very few studies have been conducted on the toxicity of these fishes and their symbiotic bacteria. In this study, we tested the toxicity of some fish specimens originating from the East Coast Water of Peninsular Malaysia and also the toxicity of pure bacteria cultures isolated from the fishes.MATERIALS AND METHODS
Fish samples: A total of 232 puffer fish comprising Lagocephalus wheeleri (84) and Lagocephalus sceleratus (148), were collected in September 2006 from the East Coast Water of Peninsular Malaysia. Samples were collected using a standard bottom trawl net. Details of the trawling operations are shown in Table 1. After hauling, the catch was removed and sorted into species groups. The puffer fish specimens were dissected and the body, gonad and liver weight was recorded. The dissected parts were then pooled together for further analysis in the laboratory.Symbiotic bacteria isolation and culture:Each pool of puffer fish organ was homogenized in a sterile blender and suspended in 5 mL of sterile seawater. Serial dilutions were made to a final concentration of 0.05 g tissue mL-1. Isolation of bacteria was achieved by spreading 0.5 mL of the diluted samples on Marine Agar plates. The plates were incubated at 28°C for 5-14 days to permit bacterial growth. Pure colonies were obtained by the streak-plate method onto new Marine Agar plates. The pure cultures were subcultured for several times to ensure purity prior to bacterial identification by morphological, physiological and biochemical tests. The tests used included gram staining, API 20 E, as well as oxidase and catalase tests.
Bacteria culturing for toxicity test: Bacteria cultures for toxicity tests were grown in sterilize Ocean Research Institute (ORI) medium (Simidu and Tsukamoto, 1985).
| Table 1: | Sampling locations, number of trawls, duration, average speed and trawling ranges |
The medium contained (g L-1) protease peptone No. 3 (Difco), 2 g; bacto-yeast extract (Difco), 2 g; phytone peptone (BBL), 1 g; sodium thiosulphate, 0.4 g; sodium sulphite, 1 g; iron citrate, 0.08 g; seawater, 750 g and made up to 1 L with distilled water and adjusted to pH 7.6. Cultures were grown in 500 mL volumes and incubated at 25°C for 10 days in darkness without aeration (Yasumoto et al., 1986).
Isolation and purification of TTX from bacteria cultures: The broth culture was centrifuged at 3000 rpm for 15 min to obtain a cell-free supernatant. The supernatant was transferred into new centrifuge tubes. The supernatant was evaporated in reduced pressure to a volume of about 100 mL and was then mixed with an equal volume of water-washed activated charcoal under agitation and filtered through a Buchner funnel. The charcoal on the funnel was thoroughly washed with distilled water and the TTX adsorbed was eluted 3 times with 3 volumes of 1% acetic acid in 20% aqueous ethanol. The elute was evaporated using a rotary evaporator. After the volume was reduced to 5 mL the sample was lyophilized, followed by addition of 10 mL 0.03 M acetic acid. This sample was used for detection of TTX by HPLC following the methods of Yu et al. (2004).
Processing of the fish organs for toxicity tests: From the frozen samples, 5 g each of gonad and flesh obtained and placed into separate tube containing 5 mL 0.03 M acetic acid and mashed using an ultrasonicator probe on ice. The slurry was centrifuged at 6000 rpm for 5 min to obtain the supernatant. The collected supernatant was later used for TTX detection.
Mouse bioassay: The mice used for the bioassay were female pure-bred Wistar strain each weighing around 20 g. One milliliter of the bacteria and fish extracts were injected intraperitoneally. Two mice were used for each sample. Symptoms were noted and the death time was recorded at the last gasping breath of the mouse (Yu et al., 2004).
Thin-Layer Chromatography (TLC): Thin-Layer Chromatography (TLC) was performed according to (Hwang et al., 1989). Briefly, samples were spotted on 5x20 cm silica gel 60 F254 precoated plates (Merck) and eluted with two solvent systems, pyridine: ethyl acetate: acetic acid: water (15:5:3:6) and 1-butanol:acetic acid: water (2:1:1). Toxins were visualized as a yellow fluorescent spot under UV (365 nm) after spraying the plate with 1% H2O2 and 10% KOH.
High-Performance Liquid Chromatography (HPLC): High-Performance Liquid Chromatography (HPLC) analysis was performed on a Shimadzu VP system with a Pickering post-column reaction system. Separation was performed using an Agilent Hipersil AA-ODS column (2.1 mm i.d. x200 mm). The mobile phase was 2 mM sodium 1-heptanesulfonic acid and 1% methanol in 0.05 M potassium phosphate buffer (pH 7.0) at a flow rate of 1 mL min-1. Sample injection volume was 10 μL. The post-column reagent was 4 N NaOH. The column temperature was kept at 37°C for all runs, while the post column temperature was set at 100°C. Detection wavelengths were set at 380 nm excitation and 505 nm emissions. All samples were analyzed twice (Hwang et al., 1989).
RESULTS
Three pure bacteria cultures were isolated from the livers of L. wheeleri and L. sceleratus and designated as isolates 1, 2 and 3. The morphological and biochemical characteristics of the three bacteria isolates are shown in Table 2. The results suggested that isolate 1 from L. wheeleri was probably a Shewanella species while isolates 2 and 3 were probably Exiguobacterium and Staphylococcus species, respectively.
Tissue extracts of L. wheeleri and culture supernatants of bacterial isolate 1 from the fish liver were determined to be toxic in the mouse bioassay. Positive symptoms were observed between 5-7 min after injection resulting in death of the mice. In contrast L. sceleratus tissue extracts and culture supernatants of bacteria isolated from the fish did not kill mice. Toxicity effects were present but the symptoms were not typical of TTX poisoning. Detailed observations from the mouse bioassays are shown in Table 3.
Using the pyridine: ethyl acetate-acetic acid: water as solvent system, spraying the plate with 1% H2O2 and 10% KOH revealed a single spot with an Rf value of 0.78 for both bacterial supernatant (sample 1/isolate 1) and tissue extracts (sample 2, flesh) of L. wheeleri (Fig. 1). In contrast, sample 3 (isolate 3, bacterial supernatant of L. sceleratus) did not exhibit any spots. When the solvent system was replaced by 1-butanol: acetic acid: water (2:1:1), spraying the plate with 1% H2O2 and 10% KOH revealed fluorescent spots for sample 1 and 2 apart from sample 3.
| Fig. 1: | Thin layer chromatography of the TTX fraction. (1: Sample 1, bacterial supernatant of L. wheeler (isolate 1), 2: Sample 2, tissue extract L. wheeleri (flesh), 3: Sample 3, bacterial supernatant of L. sceleratus (isolate 3) |
| Table 2: | Morphological and biochemical characteristic of the three bacterial isolates |
| Table 3: | Toxicities of Lagocephalus wheeleri and Lagocephalus sceleratus tissue extracts and isolated bacterial supernatants determined by mouse bioassay |
| +: Positively response, -: Negatively response, LW: Lagocephalus wheeleri, LS: Lagocephalus sceleratus, min: Minute | |
| Fig. 2: | Lagocephalus wheeleri. High-performance liquid chromatography HPLC of (a) sample 1 (isolate 1, bacterial supernatant) and (b) sample 2 (tissue extract, flesh) |
In both cases for sample 1 and 2, the main spots, with an Rf value of 0.22, was similar to that of TTX (Fig. 1).
During HPLC analysis (Fig. 2a, b), only sample 1 and 2 displayed 2 peaks between the Retention times (Rt) -1.383 and 15.617 min, respectively. According to Shida et al. (1984), Hwang et al. (1989), Lee et al. (2000) and Jang and Yamashita (2006) synthesized of TTX and its compounds are optimal within the retention times of 14-20 min. Thus, the second peaks were a result of TTX-synthesis, whereas the first peaks were probably due to lipid, fat or other compounds.
DISCUSSION
Since 1986, the TTX-producing bacteria have been isolated form several phyla of TTX-bearing organisms in searching for the true origin of TTX in the marine and land animals. There are more than 60 indigenous species of puffer fish in Chinese seas (Zhu, 1998), but few studies have been conducted to investigate TTX-bearing animal and TTX-producing bacteria (Yu et al., 2004). To present knowledge, this is the first study in which TTX-producing Shewanella sp. and non toxin producing Exiguobacterium sp. and Staphylococcus sp. were isolated from L. wheeleri and L. sceleratus in the East Coast Water of Peninsular Malaysia, which possess moderate to extreme toxic livers and flesh.
In the present study, Shewanella bacteria are dominant in the liver of L. wheeleri which has not been observed in L. sceleratus. This is not in agreement with earlier reports in which most TTX-producing strains are the members of genus Vibrio, Pseudomonas, Alteromonas, Flavobacterium and micrococcus (Miyazawa and Noguchi, 2001; Yu et al., 2004). Nevertheless, Noguchi et al. (2006) confirmed that Shewanella sp. can produce TTX. Moreover, present results seemed to be partially in agreement with the results reported by Ellenhorn and Barceloux (1988) in terms of liver as a toxic organ, even though tissue extract of L. sceleratus and culture supernatant of Exiguobacterium sp. and Staphylococcus sp. exhibited some toxic effects not typical of TTX and did not kill mice.
TLC and HPLC analysis indicated that the toxic factor in L. wheeleri and Shewanella sp. was TTX based on the peaks between retention time ranges published earlier by Shida et al. (1984), Hwang et al. (1989), Lee et al. (2000) and Jang and Yamashita (2006). This positive result was only apparent from the tissue extract of flesh and bacterial supernatant from the liver of L. wheeleri, whereas Exiguobacterium sp. and Staphylococcus sp., which were isolated from the L. sceleratus liver did not show any positive results. In mouse bioassay it might be accountable for fluffy and fat appearance of the injected mice; however in HPLC tests it did not exhibit any peak between 14-20 min retention times. Noguchi et al. (1997), Nuñez-Vázquez et al. (2000) and Yu and Yu (2002) stated that in fish species concentration of TTX may vary due to season, geographic location, sex and organ tissue but in the present study only organ and tissue were taken into account.
In this study, we have shown that L. wheeleri specimens from the East Coast Waters of Peninsular Malaysia were toxic. We also showed that a bacteria isolate from the liver of the fish produced toxicity factor(s) that resembled the fish toxin characteristics. While it can not be definitely proven that the toxic factor was TTX, the symptoms, TLC and HPLC analysis strongly suggest that the toxin was TTX. Based on the results of this study it is recommended that further studies should be conducted on puffer fishes and their associated bacteria for Malaysian waters in order to determine toxicity patterns related to season and geographic localities.ACKNOWLEDGMENTS
This study was funded by the Malaysian government through UKM Science Fund Grant No. 04-01-02-SF0124. The authors would like to express their heartiest thanks to all dedicated laboratory technicians for their help in collecting fish used in this study. The technical assistance swiftly provided by Puan Hajah Mimi Nora, Cheng Kok Wah and Mei Ling in the microbial and HPLC analysis is deeply acknowledged. Special thanks to Rabi’a Ryklief at the Nelson Mandela Metropolitan University, South Africa for critical reading of the manuscript. We gratefully acknowledge the valuable comments of four anonymous reviewers. All field sampling and laboratory protocols followed and complied with the current laws of Malaysia. This study represents here was part of first author’s MS work.