Bulletin of the Japanese Society of Scientific Fisheries 51(6), 1009-1013 (1985)

Bacterial Transformation of Paralytic Shellfish Toxins in Coral Reef Crabs and a Marine Snail

Yuichi KOTAI,*l Yasukatsu OSHIMA,*2and Takeshi YASUMOTO*2 (Accepted November 14, 1984)

Incubation of gonyautoxins with tissue extracts prepared from two species of crab, Atergatis floridus and Eriphia scabricuia, and from the turban shell, Turbo argyrostoma, confirmed trans- formation of gonyautoxins to saxitoxin through reductive elimination of C-11 hydroxysulfate and N-1 hydroxyl moieties. The reaction did not take place under bacteriostatic condition, indicating that bacteria rather than tissue enzymes were responsible for the reduction. The bacterial role was further confirmed by the fact that Pseudomonas sp. and Vibrio sp. isolated from the viscera of the above animals similarly converted gonyautoxins into saxitoxin. The bioconversion ex plained, at least partly, the discrepancy that saxitoxin and neosaxitoxin were predominant in the crabs and snails while their food alga, Jania sp., contained only gonyautoxin-I, -II, and -III. In- cubation with bacteria showed little effect on saxitoxin.

HASHIMOTOand colleagues were the first to Materials and Methods

report the occurrence of a potent toxin in coral Toxins reef crabs, Zosimus aeneus, Atergatis floridus, and GT1-3, were prepared according to the pre

Platypodia granulosa, and to identify saxitoxin viously described method from cultured cells of (STX) in Z. aeneus.l) The same toxin was also Protogonyaulax tamarensis, and neoSTX and STX

reported by YASUMOTOand KOTAII in the turban from the bivalve Spondylus butreli collected at shell Turbo argyrostoma having similar habitat.2) Palau.7) The following toxin preparations were

In later studies we also found a number of other used for incubation experiments: GTX1, with 13% crabs and snails to bear neosaxitoxin (neoSTX), of GTX2,3, a mixture of GTX2 and GTX3 (1:1),

decarbamoylsaxitoxin (dcSTX), gonyautoxin I- neoSTX containing 15% of STX, and STX. IV (GTX1-4), and an unknown toxin codenamed Bioassay of the toxins was performed according TST in the above animals.3-5) The primary to the method described in A. O. A. C.8) using male source of these toxins was presumed to be the mice of ddY strain weighing around 20g. Mouse

calcareous red alga Jania sp.6) However, in the units were converted to mol adopting the specific crabs and snails, strongly basic toxins such as activities of toxins reported by SHIMIZU et al. 9)

STX and neoSTX were always predominant over and HARADA et al. 10) their hydroxysulfate derivatives (GTX1-4), though details differed from species to species, while only Preparation of Viscera Extracts

GTX1,2,3 were detectable in Jania sp. This dis Two species of crab, Atergatis floridus and crepancy in toxin composition between the toxin- Eriphia scabricura, and the turban shell Turbo

accumulating animals and their food alga led us argyrostoma were collected respectively at Kabira to investigate the possibility of toxin conversion and Shiraho of Ishigaki Island, Japan, in March, during the accumulation process. 1983, and brought alive to the laboratory. The The present paper deals with bioconversion of viscera of the animals were homogenized with 3 GTX1-3 to STX through reductive elimination of volumes of 0.2M phosphate buffer (pH 7) and N-1 OH and C-11 OSO3 groups of GTXs by two centrifuged for 10min at 10,000rpm. species of bacteria, Pseudomonas sp. and Vibrio sp., isolated from the viscera of the crabs and the Incubation Tets turban shell. Conversion of toxins was carried out by leaving mixtures of 20 ƒÊl of toxin solution (5,000MU/ml)

1 Shokei Women's Junior College, Hachiman, Sendai, Miyagi 980, Japan(小 瀧 裕 一:尚絅 女 学 院 短 期 大 学). *2 Faculty of Agriculture, Tohoku University, Tsutsumi-dori, Sendai, Miyagi 980, Japan(大 島 泰 寛・ 安 元

健:東 北 大 学 農 学 部).

* 1010 KOTAKI, OSHIMA,and YASUMOTO

and 180ƒÊl of the viscera extracts at 25•Ž for 60h. In parallel experiments toluene was added to the

extracts at 2.5% level to prevent the bacterial

growth. Toxin conversion by isolated bacteria was tested by incubating the mixture of toxin solutions (100

MU of GTX1, GTX2,3 and STX, and 300MU of neoSTX respectively in 20ƒÊl) and 180ƒÊl of liquid

culture of bacteria at 25•Ž for 80h. Experiments were run under both aerobic and anaerobic condi

tions with use of Anaeromate (Nissui Seiyaku

Co.). To follow the progress of conversion a number of test solution were prepared. They

were taken out periodically, adjusted to pH 4 and centrifuged for 6min at 3,000rpm and then 50ƒÊl

of the supernatant was injected into the toxin analyzer.

Analyses of Toxins Toxin profiles of incubated solutions were analyzed by a liquid chromatographic fluorometric analyzer constructed by OSHIMA et al.,11) which separated toxins on a cation exchange gel column

(Hitachi gel 3011C) with citrate buffer (0.5M, pH 6.25) and then measured fluorescence inten sities of fluorophores produced by oxidation of the toxins with tert-butyl hydroperoxide in al kaline condition. With the above cited buffer separation of epimers, namely GTX1-GTX4, and GTX2-GTX3, and separation of dcSTX and TST from STX was unattainable. Therefore, supple- mental analyses by thin layer chromatography and electrophoresis were conducted after partial puri Fig. 1. Toxin profiles after incubation of GTX2,3 fication of incubation products on Bio-Gel P-2 with the extracts of the crabs (a, a', b, b') and the columns.3) turban shell (c, c') in the absence (a-c) and presence (a'-c') of toluene. Culture of Bacteria (a) A. floridus, (b) E. scabricula and (c) T. The viscera of the crabs and the midgut glands argyrostoma. of the turban shell were taken out under sterile conditions, suspended in sterile water and ino the extracts of crabs and the turban shell for 60h culated onto the surface of agar plates containing are shown in Fig. 1. In the absence of toluene, 1.5% agar, 0.25% yeast extract (Nissui Seiyaku conversion of toxins was evident as indicated by Co.), 0.5% peptone (Nissui Seiyaku Co.), 0.1% the decrease of GTX2,3 and appearance of STX, glucose (Wako Pure Chemicals) and 3% NaCl. though conversion rate differed remarkably de Viable bacterial count was taken by the spread pending on animal species; 98, 9 and 34%, re plate technique using the same agar plates. For spectively for A. floridus, E. scabricula and T. testing toxin conversion by isolated bacteria, a argyrostoma. On the other hand, GTX2 ,3 re liquid medium containing the same constituents mained unchanged under bacteriostatic conditions was used. Characterization of the isolated bac achieved by addition of toluene. teria was performed according to the systematics Two types of bacteria grew dominantly when proposed by SHIMIZU.12) suspensions of the viscera of the crabs and the turban shell and the washings of the red alga Results Jania sp. were inoculated to agar plates. The Toxin profiles after incubation of GTX2,3 with two species were code-named OK-1 and OK-2. Bacterial Transformation of PS Toxins 1011

Fig. 2. Toxin profiles after incubation of GTX2 ,3 with bacteria, Pseudomonas sp. and Vibrio sp., isolated from the crab and the snail.

respectively. OK-1 showed the following charac-

teristics: gram-negative rod with polar flagellum,

positive oxidase reaction, respiratory carbohydrate metabolism, and negative gas production. Fea

tures of OK-2 were similar to those of OK-1 except

that it fermented carbohydrates. On the basis of t hese results, OK-1 and OK-2 were assigned to

Pseudomonas sp. and Vibrio sp., respectively.

Both species were able to transform 30 and 12 of GTX2,3, respectively, into STX after incuba

tion for 60h, as shown in Fig. 2. The progress changes of various toxins during incubation with

Pseudomonas sp. are shown in Fig. 3. When a GTXI sample containing 13%. of GTX2,3 was

added to the bacterial culture, a significant de

crease of GTX, and increase of GTX2,3 took place at an earlier stage of incubation. Further trans-

formation into STX became noticeable after 40h

(Fig. 3a). When GTX2,3 were incubated, STX reached a detectable level after 20h (Fig. 3b).

NeoSTX was also converted to STX gradually Fig. 3. Time cource of the conversion of GTX1,

(Fig. 3c). STX was not affected during an in GTX2,3, neoSTX and STX during incubation with cubation period of 80h (Fig. 3d). Under an Pseudomonas sp. (ƒ¢); GTX1, (• ); GTX2,3, (•›); aerobic condition, cell density of Pseudomonas neoSTX, (•œ); STX. STX below 0.16 nmol was

sp. decreased remarkably from 3.1•~109 to 1.1•~ regarded as zero. 108cells/ml after 80h, while continuous growth of GTX2,3 and STX, and GTX2 the bacterium was observed under aerobic condi ,3 to STX, in ac tion. However, the conversion rate of GTX2,3 cordance with the results obtained with the liquid

to STX was much higher under anaerobic condition chromatographic analyzer. Formation of dcSTX and TST, which were indistinguishable from STX (29.8%) than under aerobic condition (2.2%). Obviously anaerobic condition favored the reduc by the liquid chromatographic analyzer, was not tive elimination of N-1 OH and C-11 OSO3. confirmed by TLC and electrophoresis . TLC and electrophoretic analyses of toxins Discussion before and after incubation with Pseudomonas sp. for 80h confirmed transformation of GTX1 to Our attempt to explain the discrepancy in toxin 1012 KOTAKI, OSHIMA, and YASUMOTO

Fig. 4. Reductive elimination of N-OH and -OSO3 moieties of paralytic shellfish toxins by Pseu domonas sp. Solid lines denote confirmed reaction and dotted lines denote posturated reac tion.

composition between toxin-accumulating coral at least to some extent, the discrepancy in toxin reef animals and their food alga proved the bio composition between the animals and their food conversion of GTX1-3 to STX during incubation alga. with viscera extracts of the animals. The con A significant difference was observed in toxin- version is accomplished by bacteria rather than by transforming capability of Pseudomonas sp. and tissue enzymes, as it did not proceed under bacteri Vibrio sp. isolated from the viscera. As the ostatic condition. Further support for the bac bacterial species and their population densities terial role was obtained by the isolation of two may differ among animal species, or even among species of bacteria capable of reductive elimina individuals of the same species, the actual toxin tion of N-1 OH and C-11 OSO3 groups. It is conversion may be more variable than shown by interesting to note that the reaction proceeded the in vitro experiments of this study. Involve much faster under anaerobic conditions than ment of the enzymes of the animals, especially under aerobic conditions' parallelling with the those strongly bound to tissues, cannot be ruled reductive nature of the reaction. When GTX1-3 out, as reported by other investigators. SHIMNZU were incubated with Pseudomonas sp., reduction and YOSHIOKA are the first to report toxin con of N-1 OH seemed to precede that of C-11 OSO3, version during incubation with scallop homo as indicated by the rapid increase of GTX2,3 and genates. 19) However, an ambiguity remains as the delayed appearance of STX in the medium to the possible involvement of bacteria, since com (Fig. 3a). However, neoSTX is judged to be more parison under bacteriostatic condition was not resistant than GTX, to the bacterial transforma conducted. SULLIVAN et al. showed that tissue tion despite its possession of N-1 OH (Fig. 3c). extracts of the littleneck clam specifically eliminate STX was little affected by incubation with bacteria. the carbamyol group of paralytic shellfish toxins, These results agree with the abundance of neoSTX even in the presence of a bacteriostatic agent. and STX in the crabs and snails. The visceral Further experiments are needed to assess the suspensions used for inoculation were obtained potential involvement of tissue enzymes of the from live specimens after sterilization of the outer crabs and the snails. Formation of dcSTX and surfaces with alcohol. Therefore, the bacteria TST present in crabs and snails, though in smaller must have originated in the digestive tract and it amount, was not confirmed in the present study and may be feasible to expect the bioconversion to remains to be elucidated in the future. take place in the digestive tract, accounting for, A scheme of bioconversion routes confirmed or Bacterial Transformation of PS Toxins 1013 presumed by the present study is shown in Fig. 4. Japan. Soc. Sci. Fish., 47, 957-959 (1981). Hydrolysis of aryl or alkyl sulfate of detergents by 5) T. YASUMOTO,Y. OSHIMA, M. TAJIRI, and Y. Pseudomonas sp. has been known,15) but hydro KOTAKI: Bull. Japan. Soc. Sci. Fish., 49, 633- genolysis of O-sulfate by marine bacteria seems to 636 (1983). be unreported. Further experiments are needed 6) Y. KOTAKI, M. TAJIRI, Y. OSHIMA, and T. YASUMOTO: Bull. Japan. Soc. Sci. Fish., 49, to explore the potential role played by bacteria in 283-286 (1983). other animal species, especially shellfish. 7) T. HARADA, Y. OSHIMA, H. KAMIYA, and T. YASUMOTO: Bull. Japan. Soc. Sci. Fish., 48, 821- Acknowledgements 825 (1982). The authors are greatly endebted to Drs. K. 8) W. HORWITZ (ed.): Official Methods of Analysis IZAKI of Tohoku University and P. J. SCHEUER of of the Association of Official Analytical Chemists, University of Hawaii for the valuable suggestion 13th ed., A. O. A. C., Washington, D. C., 1980, pp. for identification of bacteria and for the editorial 298-299. comments. Thanks are also due to Mr. A. TOMORI, 9) A. A. GENENAHand Y. SHIMIZU: J. Agric. Food Director of Yaeyama Branch, Okinawa Fisheries Chem., 29, 1289-1291 (1981). Experimental Station, for permitting collection of 10) T. HARADA, Y. OSHIMA, and T. YASUMOTO: Agric. Biol. Chem., 46, 1861-1864 (1982). specimens at Kabira Reef. The present study 11) Y. OSHIMA,M. MACHIDA,K. SASAKI,Y. TAMAOKI, was supported in part by a Grant-in-Aid for and T. YASUMOTO: Agric. Biol. Chem., 48, 1707- Scientific Research from the Ministry of Education, 1711 (1984). Science and Culture. 12) U. SHIMIZU: in "Kaiyo-Biseibutsu (Marine Microorganisms)" (ed. by N. TAGA), University References of Tokyo Press, Tokyo, 1974, pp. 45-64. 1) Y. HASHIMOTO: Marine Toxins and Other Bio 13) Y. SHIMIZU and M. YOSHIOKA: Science, 212, active Marine Metabolites, Japan Scientific 547-549 (1981). Societies Press, Tokyo, 1979, pp. 59-68. 14) J. J. SULLIVAN, W. T. IWAOKA, and J. LISTON: 2) T. YASUMOTOand Y. KOTAKI: Bull. Japan. Soc. Biochem. Biophys. Res. Commun., 114, 465-472 Sci. Fish., 43, 207-211 (1977). (1983). 3) Y. KOTAKI, Y. OSHIMA, and T. YASUMOTO: Bull. 15) J. M. CLOVES,K. S. DODGSON,D. E. GAMES,D. J. Japan. Soc. Sci. Fish., 47, 943-946 (1981). SHAW, and G. F. WHITE: Biochem. J., 167, 843- 4) T. YASUMOTO,Y. OSHIMA, and T. KONTA: Bull. 846 (1977).