Palytoxin in Twospecies of Xanthid Crab from the Philippines Takeshi Yasumoto, Daisuke Yasumura, Yasushi Ohizumi,*

Agric. Biol. Chem., 50 (1), 163-167, 1986 163

Palytoxin in Two Species of Xanthid Crab from the Philippines Takeshi Yasumoto, Daisuke Yasumura, Yasushi Ohizumi,* Masami Takahashi,* Angel C. Alcala** and Lawton C. Alcala** Faculty of Agriculture, Tohoku University, Tsutsumidori, Sendai 980, Japan * Mitsubishi-Kasei Institute for Life Science, ll Minamiooya, Machida-shi, Tokyo 194, Japan ** Marine Laboratory, Silliman University, Dumaguete City 6501, The Philippines Received July 23, 1985

Twospecies of xanthid crab, Lophozozymuspictor and Demania alcalai, knownto cause fatal poisoning when ingested, were collected on southern Negros, Philippines. They were highly fatal by mouse assays. The toxin in both species was identified as palytoxin,'a highly lethal toxin of zoanthids, by chromatographic and mass spectrometric analyses. The toxin was found in all tissues but the concentrations were higher in the gills, viscera, and eggs.

In tropical Pacific areas, widespread rumors thi.ds.11'12) As a resemblance between the crab exist regarding the occurrence of toxic crabs.1} toxin and palytoxin has been suggested in The species most frequently implicated in hu- pharmacological studies, emphasis was made man poisoning is Zosimus aeneus, in which the to prove the identity of the two toxins by occurrence of saxitoxin analogues1 ~4) and tet- chromatographic and spectral evidence. rodotoxin5) has been confirmed. However, the toxic principle(s) in other crabs implicated in MATERIALS AND METHODS poisoning cases has remained unidentified. Materials. Four specimens of Lophozozymuspictor and Teh and Gardiner first reported the presence two specimens of Demania alcalai were collected on of a potent toxin in Lophozozymuspictor and Negros Island, Philippines, during the period from suggested the toxin to be different from both November 1982 to May 1983. They were frozen im- saxitoxin and tetrodotoxin on the basis of mediately after capture and kept at -20°C until used for dose-death time relationships and gel per- extraction. A reference palytoxin was generously donated meation chromatography.6) The incidence of a by Professors D. Uemura of Shizuoka University and Y. human fatality due to ingestion of this species Hirata of Meijo University. was reported in Negros Island in Philippines by Alcala's group.7'8) In the same area, a Purification. Pulverized crabs were extracted 5 times with twice the volume (v/w) of MeOHat room tempera- human fatality resulting from the ingestion of ture, and the combined extracts were evaporated to dry- another crab, Demaniatoxica, also took place ness. The residue was suspended in water, freed of lipids and a related species, Demania alcalai, was with diethyl ether, and the toxin was extracted with shown to be highly lethal.8~10) BuOH.The toxic residue obtained after evaporating the BuOHwas successively treated on columns of TSK Wenow report the isolation of a toxin from G3000S (Toyosoda Co., 2.8 x 5cm), DEAE Sephadex A- specimens of L. pictor and D. alcalai, and the 25 (Pharmacia Co., 2.8 x 40cm) and CM Sephadex C-25 identification of the toxin as palytoxin, a high- (Pharmacia Co., 1 x 20cm), following the method for pu- ly potent toxin known to occur in zoan- rification ofpalytoxin described by Hirata et al.n) After re- 164 T. Yasumoto et al.

peating the DEAESephadex and CMSephadex chroma- (A) ! tography, the toxin was further purified by high perform- ance liquid chromatography (HPLC) on a Develosil ODS-10/20 column (Nomura Kagaku Co., 2x25cm) with alternative use of MeOH-O.lNHOAc(8:2) and MeCN-0.05N HOAc(5:5) solvents. Separation of the s\* I W_ Z?.alcalai toxin toxin was monitored by mouse assays and with a UV- \ x^ Palytoxin spectromonitor (Japan Spectroscopic Co., UVIDEC-100- II) at 263nm. A comparison between the toxin and the /V__^J X^L^ pictortoxin reference palytoxin by HPLCwas carried out on an ERC 0 10 20 (min) ODS-1282 column (Erma Optical Works Co., 0.6 x 25 cm) with the aforementionedtwo solvents, as well as on a TSK (B) G3000SW column (Toyosoda Co., 0.75x60cm) using 0.03n HOAcas the eluant. Thin layer chromatography (TLQ. TLC was car- ried out on Silica gel 60 and NH2F254s plates (Merck Co.) with two different solvent systems: (A) pyridine-H2O-BuOH-HOAc (10: 12: 15: 13) and (B) 1- l D. alcalai toxin pentanol-pyridine-H2O (7 : 7 : 6). The toxin on the Silica j \ Palytoxin gel 60 plates was detected by heating the plates after J \^_^ L. pictor toxin spraying with H2SO4, and that on the NH2F254s plates by 0 10 20 (min) exposure to UV light (254nm).

Bioassay. A bioassay method proposed by Teh and (O ] Gardiner6) for the toxin in L. pictor was employed. Groups of 3 male mice of the ddY strain with body weight of 17g were injected intraperitoneally with suitable volumes of sample solutions and their death times were noted. The D.alcalaitoxin j \\^ toxin amount in mouse units (MU)was calculated from Palytoxin J \\ the equation 7=225.19X~° ", where Yand Xdenote MU L.pictor toxin J ^-v. and death time in min, respectively. 0 10 20 (min) Spectra. UVspectra were taken in an aqueous solution with a Hitachi 124 spectrophotometer. FABmass spectra Fig. 1. Chromatograms of Crab Toxin and Palytoxin. were taken with a JMX-HX110 spectrometer using glyc- (A) column, ERC ODS-1282 (0.6x25cm); solvent, erol as the matrix. A cluster ofcesium iodide was used to MeOH-O.lN HOAc (8:2). (B) column, ERC ODS-1282 calibrate the massscale. (0.6x25cm); solvent, MeOH-0.05N HOAc (5:5). (C) column, TSK G3000SW (0.75x60cm); solvent 0.03n HOAc. Flow rate, 0.9 ml/min for all systems; monitoring, RESULTS 263 nm. In all the chromatographic experiments, the toxin of either L. pictor or D. alcalai was was indistinguishable from the reference indistinguishable from palytoxin. Chromato- palytoxin by TLC as shown in Table I. The grams of the crab toxin on columns of DEAE UVspectrum of the crab toxin was similar Sephadex and CMSephadex were superim- to that of palytoxin in having absorption posable on those of palytoxin. Separation of maxima at 233 and 263 nm. Molecular extinc- the toxin from buffer salts was achievable on tion coefficients at 263 nm were 23,600 for the the TSK G3000S column in a similar man- toxin ofL. pictor and 19,600 for the toxin ofD. ner to that described for palytoxin.U) As alcalai. These values are compatible with the shown in Fig. 1, the retention times for ei- reported value (22,000) for palytoxin.12) The ther crab toxin on an ERCODS-1282 column measured masses of the (M+ +H) ions of the and on a TSK G3000SW column were iden- toxin ofL. pictor and D. alcalai were 2680 and tical with those of palytoxin. The crab toxin 2678, respectively (Fig. 2). After calibrating Palytoxin in Xanthid Crabs 165

Table I. Comparison Between Crab Toxin and Palytoxin by Thin Layer Chromatography R f P la te S o lv e n t P a ly to x in L -T o x in a D -T o x in ft

Silica ge l 60 A c 0. 35, 0.60 0.3 5, 0 .60 0. 35, 0.60 Silica g el 60 B d 0.5 0 0 .5 0 0. 50 N H 2 F 254s A 0. 19, 0.20 0.1 9, 0 .20 0 .19, 0.2 0 N H 2 F 254s B 0.7 6 0 .76 0. 76

Toxin from L. pictor. Toxin from D. alcalai. l-Pentanol-pyridine-H2O (7 : 7 : 6), reported to give two spots for pure palytoxin. Pyridine-HOAc-BuOH-H2O (10 : 13 : 15 : 12).

Table II. Anatomical Distribution of Palytoxin in Crabs T o xin levels (M U /g tissue)

Specim en s C h elip ed C arap ace Vi scera G ills E ggs C u ticle F lesh

L. victor Mal e 16 5 g 2,000 4 0 0 1 0 0 1 0 , 0 0 0 2,200 M al e 14 5 g 300 2 2 0 9 4 6 , 0 0 0 2,400 F em ale 2 18 g 1,400 1 7 0 5 0 6 , 0 0 0 3,000 F em ale 120 g 160 40 20 200 180 700

D. alcalai F em ale 129 g 500 4 0 0 2 4 0 4 , 0 0 0 16,000 F em ale lO3 g 500 160 80 5,400 2,400

the mass scale, the latter ion was confirmed to DISCUSSION have a mass unit of2679.52, which agrees with the molecular composition of palytoxin The results obtained in this study confirm C129H223N3O54, 2679.487 daltons. Thus, all that the toxic principle in L. pictor and D. the chromatographic and spectral evidence alcalai is palytoxin. To the best of our knowl- supports that the toxin from the two species of edge, the present study is the first to present crab was palytoxin. evidence of the occurrence of palytoxin in The toxin contents of both crabs and the crabs and to confirm its implication in human anatomical distribution of the toxin in the poisoning. It is not certain at present whether body are summarized in Table II. All the the toxin in the crabs is of endogenous or specimens tested were lethal. One specimen of exogenous origin. The most plausible expla- D. alcalai contained as much as 1.2x 106 MU nation is that the crabs accumulate the toxin of toxin in the body. The toxin was found in all by feeding on Palythoa spp., which were com- tissues and organs, but was present in higher monly seen in the sampling areas. However, concentration in the gills, livers and gonads. the assumption remains to be verified in future Chelipeds, especially the flesh, were low in studies, as microscopic examination of the lethal potency. The eggs of L. pictor showed stomach contents of the four specimens of the the highest toxin level in the body. crab did not reveal the presence of the zoan- thids. The extremely high lethal potency of the 166 T. Yasumoto et.al.

l(9Bi «-

SB L. piotor toxin

60 C68B

1500 2000 £300 3000 3300

B300 1O0B 130B Z00B 2=00 300B 350B

Fig. 2. FABMass Spectra of Crab Toxin. Spectra were taken with a JMXHX110 spectrometer using glycerol as the matrix. crabs is worth noting and disseminating so REFERENCES that the public is informed of the potential danger of the crabs. The case report on a 1) Y. Hashimoto, "Marine Toxins and Other Bioactive Metabolites" Japan Scientific Societies Press, Tokyo, human fatality caused by the ingestion of D. 1979. toxica9) suggests that the causative toxin was 2) T. Yasumoto, Y. Oshima and T. Konta, Bull. Jpn. also palytoxin. Effort is being exerted to obtain Soc. Sci. Fish., 47, 957 (1981). 3) K. Koyama, T. Noguchi, Y. Ueda and K. specimens of D. toxica for analysis. Hashimoto, Bull. Jpn. Soc. Sci. Fish., 47, 965 (1981). 4) U. Raj, H. Haq, Y. Oshima and T. Yasumoto, Acknowledgments. The authors wish to thank Toxicon, 21, 547 (1983). Professor D. Uemura of Shizuoka University and 5) D. Yasumura, Y. Oshima, T. Yasumoto, A. C. Professor Y. Hirata of Meijo University for their gifts of Alcala and L. C. Alcala, submitted to Agric. Biol. the reference palytoxin and TSKG3000S gel. Thanks are Chem. also due to Drs. Y. Itagaki and I. Nojima, Application 6) Y. F. Teh and J. E. Gardiner, Toxicon, 12, 603 Center, JEOLLtd., for the measurement of mass spectra. (1974). The present study was supported by a Grant-in-Aid for 7) R. B. Gonzales and A. C. Alcala, Toxicon, 15, 169 Scientific Research (Overseas Scientific Survey) from the (1977). Ministry of Education, Science and Culture of Japan. 8) A. C. Alcala, Toxicon, Suppl., 3, 1 (1983). 9) A. C. Alcala and B. W. Halstead, Clin. ToxicoL, 3, Palytoxin in Xanthid Crabs 167

609 (1970). 12) R. E. Moore, R. F. Dietrich, B. Hatton, T. Higa and 10) E. E. Cammbana, A. C. Alcala and E. P. Ortega, P. J. Scheuer, /. Org. Chem., 40, 540 (1975). Silliman J., 23, 265 (1976). 13) S. Kimura and Y. Hashimoto, Publ. Seto Mar. Biol. ll) Y. Hirata, D. Uemura, K. Uedaand S. Takano, Pure Lab., 20, 713 (1973). Appl. Chem., 51, 1875 (1979).