- Email: [email protected]
Local differences in toxin composition of a xanthid crab Atergatis floridus inhabiting Ishigaki Island, Okinawa
Twicat, Vol . 2I, No. 7, pp. 705-711, Printed in Croat ariuin.
0041-olova6s5.ao+ .ao
1986.
Perpmon Journals Ltd.
LOCAL DIFFERENCES IN TOXIN COMPOSITION OF A XANTHID CRAB ATERGATIS FLORIDUS INHABITING ISHIGAKI ISLAND, OKINAWA TAMAO NOGUCHI, OSAMU ARAKAWA, KINUE DAIGO Snd KANEHISA HASHIMOTOw Laboratory of Marine Biochemistry, Faculty of Agriculture; University of Tokyo, Bunkyo, Tokyo 113, Japan (Acceptedjorpublirntion 4 Mmrh 1986) T. Nooucltl, O. ARAICAWA, K. D.~loo and K. HASHIIrtOTO . Loral differences in toxin composition of a xanthid crab Atergatisjloridus inhabiting Ishigaki Island, Okinawa. Toxicon 24, 705-711, 1986 .-Spaimens of a xanthid crab AtergatisJloridus were collected from two adjacent aresß in Kabira Hay, Iahigaki Island, Okinawa and compared with respect to toxicity and toxin composition . 'Reef specimens', which were collected from the reefs of Kabira Bay, showed an average toxicity score of 380 MU/g as paralytic ahellfulh poison, aad a toxin composition consisting mainly of saJdtoxin and a neosaxitoxin-related substance. On the other hand, 'Kojima specimens', which were collected from a small island in the bay, showed an average toxicity of 38 MU/g as tetrodotoxin, and a toxin composition consisting mainly of tetrodotoxin and related silbatance(s) .
INTRODUCTION
Atergatisfloridus is known to contain high levels of paralytic toxins (HASHIMOTO, 1979),
which are composed mainly of paralytic shellfish poison (PSP) in Okinawa specimens (KOYAMA et al., 1981) and composed of tetrodotoxin (TTX) in specimens inhabiting the Pacific coasts of Japan Proper (NOßUCHI et al., 1983). Recently we have found that this crab, as well as another xanthid crab Zosimus aeneus, releases appreciable amounts of toxin when wiped with gauze (NOGUCHI et al., 198 . This suggests that the toxin could be a biological defense agent or an essential substance for these crabs to survive. This is supported by evidence that its resistance to PSP or TTX is several thousands times higher than that of non-toxic species of crabs (KOYAMA et al., 1983). The mechanism by which crabs, including A. floridus, Z. smear and Platypodia granulosa, become toxic has not yet been fully elucidated. In this connection, KOTAKI et al. (1983) consider a calcareous alga Jania sp. to be the origin of crab toxin. The present paper deals with a comparison of the toxin compositions of A. floridus specimens coliected from two environmentally close areas on Ishigaki Island, as part of studies on the mechanism by which crabs become toxic. MATERIALS AND METHODs
CYnb specimens In May 1984 11 specimens (S males and 6 females; mean body weight t S.E ., 12 t 2 g) of A. Jloridus were collected on reefs near the entrance to Kabira Hay, Ishigaki Island, Okinawa Irefature, and S I specimens from rI'o whom reprint requests should be addressed. 70S
70 6
TAMAO NOGUCHI n al.
a small island, Kojima, which is separated by about 100 m from one of the reefs by astrait . The specimens were immediately frotta, transported to our laboratory and kept frozen at -20°C until use. A portion of each of the 11 `reef specimens' and 10 of the 'Kojima spedmrns' (S males and 5 females; 19 t 3 g) was assayed for toxicity, and the remaining portions, along with the rest of the `Kojima specimens', were used for toxin purification, as described below. Amy of toxicity The appendages were removed from each crab and toxicity determined by the official assay method for PSP (KAWAHATA, 1978). The toxicity of 'reef spaimrns' was calculated as PSP, that of `Kojima specimens' as TTX. Purjjication ojAtergatis toxins The method applied was essentially the same as that previously reported (Noaucxt et aL, 1983). Briefly, appendages and carapaces of the crab specimens were combined, and extracted with S0~ ethanol acidified to pH 2 with HCI. The extract was defatted with dichloromethane, and the toxin was purified by a procedure which consisted of activated charcoal treatment, followed by several types of chromatography using a Bio-0el P-2 column (6 x SS cm) and a Bio-Rex 70 (H` form) column (0 .8 x 94 cm) (Fig . 1). In Bio-Rex70 chromatography toxins were doted by a two-aRep linear gradirnt, the Eirat step 0-O.OS M acetic acid and the second 0.05 -1 .S M acetic add (Fig . 2) . The toxic fraction (6000 MU) from `reefspedmena' was separated into two peaks, which were designated Sr-I and -II. Fractions indicated by the arrow: were combined and analyzed as described below. The toxic fraction (6000 MU) from `Kojima specimens' was separated into two aasymmetrical peaks. Fractions as indicated by the arrows were combined sad designated Gk-I, -II and -III . Fradiona 31-S8 from 'reef apedmens' were, though apparently non-toxic, combined and concentrated to aff~d a toxic solution, which was designated Gr. Similarly, fractions 154-174 from `Kojima spaimena' were combined and concentrated, resulting in fraction Sk . The 7 fractions thus provided (Gr, Sr-I, -II, Gk-I, -II, -III and Sk) were analyzed for toxin composition a: described blow . laectropJtorrsis Electrophoresis was performed on S x 18 cm cellulose acetate strips (Chanetron, Milano, Italy) in 0.08 M Tris-HCl buffer (pH 8.7) at 0.8 mA/cm width for 30 min. After the run, the strlpa were sprayed with IY~ H,O followed by heating at 110°C for S min. The toxins were visualized under u.v . (365 nm). The strips were then sprayed with 10~ KOH, heated for 10 min and observed similarly. Thin-layer ehrontatogiaphy
Thin-layer chromatography was performed on silica gel LHP-K plates (whatmaa, Clifton, U.S .A .) with a solvent system of pyridine-ethyl acetate-acetic acid-water (15:5 :3 :4). After devdopmrnt the toxins were detecrted in the same manner as in datrophoreais . perjorrnareoe lfquld chrontatog+aPky Reverse phase high performance liquid dtronuttography (HPLC) was carried out on so AM-314 (YMC, Kyoto, Japan) cohmm (0.6 x 30 cm), using heptatkaulfonic add as an ion-pairing reagent, with 19s McOH-O .OS M potassium phosphate (pH 7 .0) for Gr and Ok-I- IiI and with 20s1s McOH-0.03 M potassium phosphate (pH 7.0) for Sr-I, -II and Sk (Nwa+stt~wtw et aL, 1986). The periodate reagent used to detect PSP was prepared according to themethod of Stn.uvwN et al. (1983). The fluorogenic reaction was performed at 63°C for 0.7 min. The ratio of column duale to periodate reagent volume is the mixing tee was 1:1 . The fhorophors formed were monitored at 388 nm with 344 am excitation . TTX was dilated by repLdng the paiodate reagent with 3 N NaOH, batting the reaction mixture at 100°C for 0.4 min and monitoring the fluorescence at SOS am with 38l nm excitation . Hish
RESULTS
Results of the lethal potency assays of the A. floridus specimens were as follows. Nine out of the 11 `reef specimens' were toxic. The lethal potency ranged from 10 to 1300 MU/g (mean t S.E ., 380 f 136 MU/g). All of the 10 `Kojima specimens' assayed were toxic, the lethal potency ranging from 11 to % MU/g (38 t 8 MU/g). The lethal potency did not depend upon body weight nor sex. Yields of toxin at each step of purification are shown in Fig. 1 . In Bio-Rex 70 cohimn chromatography of the `reef specimens' most of the lethal potency (4700 MU as PSP, 9~4a7a) was eluted in fractions Sr-I and -II, while only a small portion (310 MU as PSP,
Toxin Composition of a Xanthid Crab Material
707
(R = 80 gs R = 380 q)
HomogQnized Extracted with BOi EtOH (pH 2 .0) Extract Defatted with CH 2 C1 2 Aqueous layer ~(R = 12,000 Mo* ; R = 6,000 MU**) Activated charcoal treatment Eluted with l~ Ac08 in 20i EtOH Elutee
(R = 8,000 MU* : R = 6,600 MO**)
Hio-Gel P-2 Eluted with 0 .1 M AcOH Elutee (R = 6,000 MU*= R = 6,000 MU**) Hio-Rex 70
(H + )
Eluted with 0-0 .05 M AcOH, 0 .05-1 .5 M AcOH
0-0 .05 M Ac08 eluatn
0 .05-1 .5 M ACOH eluate
( "GT7C fraction")
( "STX fraction")
(R = 310 MU* ; R = 4,900 MO**)
(R = 4,700 MO*t R ~ 140 MU*)
Fic . 1 . PaocenuASS rnoa pu~cwnoN oF Ateraatis roxuvs . R = 'reef apecimens ; K = 'Kojima apaimena' . " Calculated as PSP . "" Calculated as TTX .
6%) was in fraction Gr. The elution profile is shown in Fig. 2. In strong contrast, the 'Kojima specimens' gave a lethal potency ratio of Sk to Gk (I + II + III) of 3% (140 MU as PSP):97% (4900 MU as TTX). The 7 fractions (Gr, Sr-I, -II, Gk-I, -II, -III and Sk) thus obtained were submitted to the following analyses . Electrophoresis of fraction Gr (Fig. 3) gave five spots, three of which corresponded in mobility to GTX,_, . One yellow spot with R. 0.69 was hardly detected before spraying with 10% KOH, suggesting that it was similar to TTX. Fraction Sr-I was composed of three unknown components, whereas Sr-II was composed of STX and two unknown components, whose mobilities agreed well with those of two unknowns from Sr-I . Fraction Gk-I showed two yellow spots, one of which (R= 0.59) was toxic . These spots were visualized after 10% KOH was sprayed, suggesting their similarity to TTX and not to PSP. Fraction Gk-II also showed two yellow spots which were visualized by spraying with 10% KOH. One of them (R= 0.63) was judged to be TTX from the R value and yellow fluorescence . Fraction Gk-III showed the presence of TTX. It also gave two faint spots corresponding to C3TX2,~ and, in addition, several unidentified yellow spots which
708
TAMAO NOGUCHI et al.
FIa. 2. BIO-REX 7O COLlJh4d CHAOMATOOAAPHY OF THE TOXIC FRACTION FROM `REEFSPECIMENS' AND `COJIMA SPECIMENS' . The toxic fraction was obtained from the extract of 'reef specimens' or 'Kojima specimens' of Ater~atis cab by Bio-0el P-2 treatment (ii= . 1), and subjected to Bio-Rex 70 column (0.8 x 94 cm) chromato~aphy, usins a two-step linear ~adient of 0-0.05 M and 0.05 -1 .5 M AcOH . The rate of flow was I ml/min. Thetoxic fraction (6000 MU) from 'reefspaimens' was separated into two peals, which were desisnated Snl and -II. Fractons indicated by the arrows was combined and analysed . The toxk fraction (6000 MU) from 'Koj{ma specimens' was separated into two assymmarical pealu. Fractions w indicted by the arrows was combined and designated Gk-I, -II and -III. Fractions 31- 58 from `reef specimens' were combined and concentrated to afford a toxic solution, which was de:;gnated Gr. Similarly, fractions 154-174 fmm `Kojima specimens' were combined and concentrated, resulting in fraction Sk. A total of acwen fractions (Gr, Sr-I, -II, Gk-I, -II, -III and Sk) were analysed by ekctrophoraic and other techniques . " " `Reef specimens' ; p.-Q `Kojima specimens' .
were visualized with 109o KOH . Fraction Sk was composed of the same components as Sr-I and -II from the `reef specimens', except for an additional component with an R of 0.33. The results with thin-layer chromatography of the 7 fractions were essentially consistent with the electrophoretic results (Fig. 4). . Upon high performance liquid chromatography, fraction Gr gave four peaks with the periodate reagent (for PSP), as shown in Fig. 5A. Two large peaks and the small slowest one were identified as GTX,_,, respectively, from their retention times. On the other hand, a small peak with the same retention time as TTX (10.9 min) also appeared when 3 N NaOH (for TTX) was used instead of the periodate reagent (data not shown) . Fraction
Toxin Composition of a Xanthid Crab
Reference standards
6TX4.1 .3 .2 neoSTX.STX ~
Fr . ~r 'Reef 0 Fr . $j-1 saeclmens' Fr . ,~-II
'KO11n:o
speclss;ns'
0
~
e
~
709
Q
e
0 il 0 ®
e
Fr, §j;-I Fr . Ç~k-II
Fr . if ~-m Fr . $Iç (
~
~
~
.
I
.
~
,
.
I
Fta. 3. Et .ecraorttoat~rs of Atergatis ~rowN trawcnoNS, wtAtvo wrrx wtrrxe:vrt~ ~rowrrs. Electrophoresis was performed on S x 18 cm cellulose soetate strips (Chemetron) in 0.08 M Tris-HCl buffer (pH 8.~ at 0.8 mA/cm width for 30 min. After the run the Grips were spayed with 1 ß~ H,O,, followed by heating at 110°C for S min. The toxins were visualized under u.v . (365 ). The strips were en sprayed with l0ß` KOH, heated for 10 min and observed~m~1arlY. Blue fluorecena ; greenish yellow fluorescence, both visualized with lß~ H,O,. " Yellow fluorescence, visualised with lß4 H,O, and 10ßi KOH. Sr-I showed a main peak with a shoulder on its preceding edge followed by a smaller peak (Fig . SC). The retention times of the peaks did not correspond to STX or neoSTX . The shoulder, however, was identified as neoSTX from its retention time (8 .9 min) . Fraction Sr-II showed a large peak, which was identified as STX by its retention time (11 .9 min), with a shoulder on its preceding edge (Fig . SC).
Reference standards
0 0 ~ O"
~- 67X3.2 "4, t STX.n~STX
r"c s o " Gr
Fr, fiL 'Reef I Fr, ~-I speclss;ns" ~ Fr, ~-II Fr . ¢g-I 'Ko11nq Fr . ~k-II snect:sens' Fr, I1lç-m Fr, Slç
r N sis~" o " m I 0
.
.
~
~
I , 0.5
~ "
Py :AcOEt :AcOH :H20 (15:5 :3 :4)
I ~ 1,0
Fta. 4. Twx-urea cxao~wTOanwrFtr of Ateraatis TowN . Thin-layer chromatography was Performed on silica gd LHP-K plats (Whatman) using a solvent system of pyrWine-ethyl acetate-acedc acid-water (15 :5 :3:4). After devdopmeat toxins were detected as in elatmphoresb (see legend in Fig. 3) .
710
TAMAO NOCiUCHl d al.
A B
_ rr . ~-a
0
nwB~S
0
10 ~o Il~t~ntSoa sir (~ial
o
m a~bntion tir (~Sa1
0
oa
10 1~t~nsim tir l~Inl
FIC . S . RICH PERFORMANCE LIQUID CHROMATOORAPI{Y OF SOME FRACTIONS OF Atergatis ToxINS. Reverse-phase high performance liquid chromatography wag arried out on an AM-314 (YMC) column (0 .6 x 30 cm), using heptanesulfonic add as an ion-pairing reagent, with 1~ McOH-0 .03 M potassium phosphate (pH 7.0) for Gr and Gk-I-III, and with 20% McOH-0.05 M potassium phosphau (pH 7.0) for Sr-I, -II and Sk. The periodate reagent (SULLIVAN et d., 1983) wag used to detect PSP. The fluorogenic reaction wsa performed at 65°C for 0.7 min. The fluorophors formed wen monitoredat 38B nm with 344 nm exdtation. TTXwag detected by replacing the periodau reagent with 3 N NaOH, heatingthe reaction mixtureat 100°C for 0.4 min and monitoring the fluorescence at SOS nm with 381 nm exdtation.
As shown in Fig. SB, fraction Gk-I gave two peaks, each with a trailing shoulder, when 3 N NaOH was used for detection. Both of these peaks remain unidentified, but the shoulder of the main peak was identified as TTX from its retention time (10.9 min) . Fraction Gk-II showed a dominant peak, whose retention time (10.9 min) agreed well with that of TTX (Fig. SB). Fraction GK-III showed several peaks, one of which (retention time 11 .1 min), in spite of a little lag of retention time, was judged to be TTX from cochromatography (data not shown) . In addition, small amounts of GT7C,.., were detected when the periodate reagent was used (Fig. SA). In fraction Sk neoST7{ and STX were detected, along with several unknown components (data not shown) . DISCUSSION
`Reef specimens' were found to contain ST?C and a neoSTX-related substance as major components, along with GT7C's and a small amount of a TTX-related substance as minor components, by electrophoresis, thin-layer chromatography and high performance liquid chromatography . We previously obtained (NOOUCHI et al., 1984) essentially the same results for `Ishigaki specimens' of A. floridus. On the other hand, it was also shown that `Kojima specimens' possessed TTX and unidentified TTX-related substances) as major
Toxin Composition of a Xanthid Crab
711
toxins, along with some PSP components, such as neoSTX; STX and GTX's, as minor toxins . This toxin composition is similar to that of A . floridus inhabiting the Pacific coasts of Japan proper (NOGUCHI et al., 1983). Thus the toxin compositions of the specimens collected from two adjacent areas almost simultaneously were extremely different from each other . This supports the view that the toxin of A. fTorzdus is derived from an exogenous origin, such as its foods or symbiotic microorganisms, rather than being of endogenous origin . The present results, along with other toxicity data, suggest that TTX might be an essential constituent for this crab and that the crab could be additionally toxic due to the harboring of a PSP-producing organisms) which lives indigenously on coral reefs. In this connection, KOTAKI et al. (1983) detected PSP in a calcareous alga Jania sp. inhabiting tropical and subtropical seas and presume its exclusive involvement in the intoxication of xanthid crabs. The intoxication of those crabs, however, can not be unambiguously accounted for by this alga alone, because the PSP content of this alga generally is very low and because A. floridus also often contains TTX as the major toxin, as demonstrated here and as reported previously (NocucHl et al., 1983). Acknowledgemarts-The authors express sincere thanks to Diraxor Dr A. ToMOw and the staffof the Yaeyama Branch of the Okinawa Fisheries Expaimental Station for their kind collaboration in colkcting the crab specimens. Thanks are also due to Professor M. Is:www, University of New Hampshire, U.S .A., for reading the manuscript of the preaeat paper. REFERENCES Hwssusoro, Y. (1979) Toxic crabs. In: Mmine Toxins mid Ot1Yer Blo~scttvr Marlne MetabolUes, p. 59. Tokyo: Japan Scientific Societies Press. Kwwwawrw, T. (1978) Assay method for paralytic shellfish pooson . Ia : Food XyaieneEmminatlon Mmueal, Vol. 2, p. 240. Tokyo: Japan Food Hygiene Assodation. Korws;t, Y., TwJnu, M., OSEm~tw, Y. and YwsuMOro, T. (1983) Identification of a calareous red alga as the Primary soura of paralytic shellfish toxins is coral reef crabs and gastropods . Bull. Jprt. Soc. scient. Firh. ~!, 283. KOYAMw, K., NoaucEn, T., Ucnw, Y. and Hwsw~toro, K. (1981) Occurrence of neosaxitootin and otherparalytic ahellf-ash poisons in toxic crabs belonging to the family Xanthidae . Ball. Jpn. Soc. scient. F7sh . 47, %S . KOYAMw, K., NoaucFn, T., Uw, A. and Hws~oro, K. (1983) Resistibility of toxic and nontoxic crabs against paralytk shellfish poison and tetrodotoxin . Bull. Jpn. Soc. acknt. FYsh. 49, 485. NAOnsHUU, Y., MwauYwMw, J., Noaucxt, T. and Hwsttutoro, K. (198 Separation of paralytic shellfish poisom and tetrodotoxin by ion-pairing liquid chromatography . Bull. Jpn. Soc. acknt. FYsk . (in press). Noaucxi, T., Uzu, A., KoYwatw, K., Mwawwntw, J., Nwa~swMw, Y. and HwsFttworo, K. (1983) Oaurreace of tetrodotoxin as the m~ior toxin in a xanthid aab AtergattrJlorldus. Ball. Jpn. Soc. acieat. FYrle. 49, 1887 . Noaucsi, T., Uzu, A., Dw~ao (KoYwsaw), K., SHIDA, Y. and Hwsx~MOro, K. (1984) A tetrodotoxin-like substance as a minor toxin in the xanthid cnb AtergatisJlorldus. Ta:ricnrt 22, 423. NooucFU, T., Dwioo, K., Aawtuww, O. and HwstiIMOro, K. (198 Release of paralytic shellfish poison from the exoskeleton of a xanthid crab Zastrnus aureus. In: l'nvceadings ojthe Thud International Corfferrna on Taxé lNnoJlaadlate Blooms . New York : EhKVier (in press). Sut.t.wwN, J. J. and Iwwoxw, W. T. (1983) High pressure liquid chromatographic determination of toxins associated with paralytic shellfish poisoning . J. Ass. ojf. analyt. CNevn. f6, 297.
0041-olova6s5.ao+ .ao
1986.
Perpmon Journals Ltd.
LOCAL DIFFERENCES IN TOXIN COMPOSITION OF A XANTHID CRAB ATERGATIS FLORIDUS INHABITING ISHIGAKI ISLAND, OKINAWA TAMAO NOGUCHI, OSAMU ARAKAWA, KINUE DAIGO Snd KANEHISA HASHIMOTOw Laboratory of Marine Biochemistry, Faculty of Agriculture; University of Tokyo, Bunkyo, Tokyo 113, Japan (Acceptedjorpublirntion 4 Mmrh 1986) T. Nooucltl, O. ARAICAWA, K. D.~loo and K. HASHIIrtOTO . Loral differences in toxin composition of a xanthid crab Atergatisjloridus inhabiting Ishigaki Island, Okinawa. Toxicon 24, 705-711, 1986 .-Spaimens of a xanthid crab AtergatisJloridus were collected from two adjacent aresß in Kabira Hay, Iahigaki Island, Okinawa and compared with respect to toxicity and toxin composition . 'Reef specimens', which were collected from the reefs of Kabira Bay, showed an average toxicity score of 380 MU/g as paralytic ahellfulh poison, aad a toxin composition consisting mainly of saJdtoxin and a neosaxitoxin-related substance. On the other hand, 'Kojima specimens', which were collected from a small island in the bay, showed an average toxicity of 38 MU/g as tetrodotoxin, and a toxin composition consisting mainly of tetrodotoxin and related silbatance(s) .
INTRODUCTION
Atergatisfloridus is known to contain high levels of paralytic toxins (HASHIMOTO, 1979),
which are composed mainly of paralytic shellfish poison (PSP) in Okinawa specimens (KOYAMA et al., 1981) and composed of tetrodotoxin (TTX) in specimens inhabiting the Pacific coasts of Japan Proper (NOßUCHI et al., 1983). Recently we have found that this crab, as well as another xanthid crab Zosimus aeneus, releases appreciable amounts of toxin when wiped with gauze (NOGUCHI et al., 198 . This suggests that the toxin could be a biological defense agent or an essential substance for these crabs to survive. This is supported by evidence that its resistance to PSP or TTX is several thousands times higher than that of non-toxic species of crabs (KOYAMA et al., 1983). The mechanism by which crabs, including A. floridus, Z. smear and Platypodia granulosa, become toxic has not yet been fully elucidated. In this connection, KOTAKI et al. (1983) consider a calcareous alga Jania sp. to be the origin of crab toxin. The present paper deals with a comparison of the toxin compositions of A. floridus specimens coliected from two environmentally close areas on Ishigaki Island, as part of studies on the mechanism by which crabs become toxic. MATERIALS AND METHODs
CYnb specimens In May 1984 11 specimens (S males and 6 females; mean body weight t S.E ., 12 t 2 g) of A. Jloridus were collected on reefs near the entrance to Kabira Hay, Ishigaki Island, Okinawa Irefature, and S I specimens from rI'o whom reprint requests should be addressed. 70S
70 6
TAMAO NOGUCHI n al.
a small island, Kojima, which is separated by about 100 m from one of the reefs by astrait . The specimens were immediately frotta, transported to our laboratory and kept frozen at -20°C until use. A portion of each of the 11 `reef specimens' and 10 of the 'Kojima spedmrns' (S males and 5 females; 19 t 3 g) was assayed for toxicity, and the remaining portions, along with the rest of the `Kojima specimens', were used for toxin purification, as described below. Amy of toxicity The appendages were removed from each crab and toxicity determined by the official assay method for PSP (KAWAHATA, 1978). The toxicity of 'reef spaimrns' was calculated as PSP, that of `Kojima specimens' as TTX. Purjjication ojAtergatis toxins The method applied was essentially the same as that previously reported (Noaucxt et aL, 1983). Briefly, appendages and carapaces of the crab specimens were combined, and extracted with S0~ ethanol acidified to pH 2 with HCI. The extract was defatted with dichloromethane, and the toxin was purified by a procedure which consisted of activated charcoal treatment, followed by several types of chromatography using a Bio-0el P-2 column (6 x SS cm) and a Bio-Rex 70 (H` form) column (0 .8 x 94 cm) (Fig . 1). In Bio-Rex70 chromatography toxins were doted by a two-aRep linear gradirnt, the Eirat step 0-O.OS M acetic acid and the second 0.05 -1 .S M acetic add (Fig . 2) . The toxic fraction (6000 MU) from `reefspedmena' was separated into two peaks, which were designated Sr-I and -II. Fractions indicated by the arrow: were combined and analyzed as described below. The toxic fraction (6000 MU) from `Kojima specimens' was separated into two aasymmetrical peaks. Fractions as indicated by the arrows were combined sad designated Gk-I, -II and -III . Fradiona 31-S8 from 'reef apedmens' were, though apparently non-toxic, combined and concentrated to aff~d a toxic solution, which was designated Gr. Similarly, fractions 154-174 from `Kojima spaimena' were combined and concentrated, resulting in fraction Sk . The 7 fractions thus provided (Gr, Sr-I, -II, Gk-I, -II, -III and Sk) were analyzed for toxin composition a: described blow . laectropJtorrsis Electrophoresis was performed on S x 18 cm cellulose acetate strips (Chanetron, Milano, Italy) in 0.08 M Tris-HCl buffer (pH 8.7) at 0.8 mA/cm width for 30 min. After the run, the strlpa were sprayed with IY~ H,O followed by heating at 110°C for S min. The toxins were visualized under u.v . (365 nm). The strips were then sprayed with 10~ KOH, heated for 10 min and observed similarly. Thin-layer ehrontatogiaphy
Thin-layer chromatography was performed on silica gel LHP-K plates (whatmaa, Clifton, U.S .A .) with a solvent system of pyridine-ethyl acetate-acetic acid-water (15:5 :3 :4). After devdopmrnt the toxins were detecrted in the same manner as in datrophoreais . perjorrnareoe lfquld chrontatog+aPky Reverse phase high performance liquid dtronuttography (HPLC) was carried out on so AM-314 (YMC, Kyoto, Japan) cohmm (0.6 x 30 cm), using heptatkaulfonic add as an ion-pairing reagent, with 19s McOH-O .OS M potassium phosphate (pH 7 .0) for Gr and Ok-I- IiI and with 20s1s McOH-0.03 M potassium phosphate (pH 7.0) for Sr-I, -II and Sk (Nwa+stt~wtw et aL, 1986). The periodate reagent used to detect PSP was prepared according to themethod of Stn.uvwN et al. (1983). The fluorogenic reaction was performed at 63°C for 0.7 min. The ratio of column duale to periodate reagent volume is the mixing tee was 1:1 . The fhorophors formed were monitored at 388 nm with 344 am excitation . TTX was dilated by repLdng the paiodate reagent with 3 N NaOH, batting the reaction mixture at 100°C for 0.4 min and monitoring the fluorescence at SOS am with 38l nm excitation . Hish
RESULTS
Results of the lethal potency assays of the A. floridus specimens were as follows. Nine out of the 11 `reef specimens' were toxic. The lethal potency ranged from 10 to 1300 MU/g (mean t S.E ., 380 f 136 MU/g). All of the 10 `Kojima specimens' assayed were toxic, the lethal potency ranging from 11 to % MU/g (38 t 8 MU/g). The lethal potency did not depend upon body weight nor sex. Yields of toxin at each step of purification are shown in Fig. 1 . In Bio-Rex 70 cohimn chromatography of the `reef specimens' most of the lethal potency (4700 MU as PSP, 9~4a7a) was eluted in fractions Sr-I and -II, while only a small portion (310 MU as PSP,
Toxin Composition of a Xanthid Crab Material
707
(R = 80 gs R = 380 q)
HomogQnized Extracted with BOi EtOH (pH 2 .0) Extract Defatted with CH 2 C1 2 Aqueous layer ~(R = 12,000 Mo* ; R = 6,000 MU**) Activated charcoal treatment Eluted with l~ Ac08 in 20i EtOH Elutee
(R = 8,000 MU* : R = 6,600 MO**)
Hio-Gel P-2 Eluted with 0 .1 M AcOH Elutee (R = 6,000 MU*= R = 6,000 MU**) Hio-Rex 70
(H + )
Eluted with 0-0 .05 M AcOH, 0 .05-1 .5 M AcOH
0-0 .05 M Ac08 eluatn
0 .05-1 .5 M ACOH eluate
( "GT7C fraction")
( "STX fraction")
(R = 310 MU* ; R = 4,900 MO**)
(R = 4,700 MO*t R ~ 140 MU*)
Fic . 1 . PaocenuASS rnoa pu~cwnoN oF Ateraatis roxuvs . R = 'reef apecimens ; K = 'Kojima apaimena' . " Calculated as PSP . "" Calculated as TTX .
6%) was in fraction Gr. The elution profile is shown in Fig. 2. In strong contrast, the 'Kojima specimens' gave a lethal potency ratio of Sk to Gk (I + II + III) of 3% (140 MU as PSP):97% (4900 MU as TTX). The 7 fractions (Gr, Sr-I, -II, Gk-I, -II, -III and Sk) thus obtained were submitted to the following analyses . Electrophoresis of fraction Gr (Fig. 3) gave five spots, three of which corresponded in mobility to GTX,_, . One yellow spot with R. 0.69 was hardly detected before spraying with 10% KOH, suggesting that it was similar to TTX. Fraction Sr-I was composed of three unknown components, whereas Sr-II was composed of STX and two unknown components, whose mobilities agreed well with those of two unknowns from Sr-I . Fraction Gk-I showed two yellow spots, one of which (R= 0.59) was toxic . These spots were visualized after 10% KOH was sprayed, suggesting their similarity to TTX and not to PSP. Fraction Gk-II also showed two yellow spots which were visualized by spraying with 10% KOH. One of them (R= 0.63) was judged to be TTX from the R value and yellow fluorescence . Fraction Gk-III showed the presence of TTX. It also gave two faint spots corresponding to C3TX2,~ and, in addition, several unidentified yellow spots which
708
TAMAO NOGUCHI et al.
FIa. 2. BIO-REX 7O COLlJh4d CHAOMATOOAAPHY OF THE TOXIC FRACTION FROM `REEFSPECIMENS' AND `COJIMA SPECIMENS' . The toxic fraction was obtained from the extract of 'reef specimens' or 'Kojima specimens' of Ater~atis cab by Bio-0el P-2 treatment (ii= . 1), and subjected to Bio-Rex 70 column (0.8 x 94 cm) chromato~aphy, usins a two-step linear ~adient of 0-0.05 M and 0.05 -1 .5 M AcOH . The rate of flow was I ml/min. Thetoxic fraction (6000 MU) from 'reefspaimens' was separated into two peals, which were desisnated Snl and -II. Fractons indicated by the arrows was combined and analysed . The toxk fraction (6000 MU) from 'Koj{ma specimens' was separated into two assymmarical pealu. Fractions w indicted by the arrows was combined and designated Gk-I, -II and -III. Fractions 31- 58 from `reef specimens' were combined and concentrated to afford a toxic solution, which was de:;gnated Gr. Similarly, fractions 154-174 fmm `Kojima specimens' were combined and concentrated, resulting in fraction Sk. A total of acwen fractions (Gr, Sr-I, -II, Gk-I, -II, -III and Sk) were analysed by ekctrophoraic and other techniques . " " `Reef specimens' ; p.-Q `Kojima specimens' .
were visualized with 109o KOH . Fraction Sk was composed of the same components as Sr-I and -II from the `reef specimens', except for an additional component with an R of 0.33. The results with thin-layer chromatography of the 7 fractions were essentially consistent with the electrophoretic results (Fig. 4). . Upon high performance liquid chromatography, fraction Gr gave four peaks with the periodate reagent (for PSP), as shown in Fig. 5A. Two large peaks and the small slowest one were identified as GTX,_,, respectively, from their retention times. On the other hand, a small peak with the same retention time as TTX (10.9 min) also appeared when 3 N NaOH (for TTX) was used instead of the periodate reagent (data not shown) . Fraction
Toxin Composition of a Xanthid Crab
Reference standards
6TX4.1 .3 .2 neoSTX.STX ~
Fr . ~r 'Reef 0 Fr . $j-1 saeclmens' Fr . ,~-II
'KO11n:o
speclss;ns'
0
~
e
~
709
Q
e
0 il 0 ®
e
Fr, §j;-I Fr . Ç~k-II
Fr . if ~-m Fr . $Iç (
~
~
~
.
I
.
~
,
.
I
Fta. 3. Et .ecraorttoat~rs of Atergatis ~rowN trawcnoNS, wtAtvo wrrx wtrrxe:vrt~ ~rowrrs. Electrophoresis was performed on S x 18 cm cellulose soetate strips (Chemetron) in 0.08 M Tris-HCl buffer (pH 8.~ at 0.8 mA/cm width for 30 min. After the run the Grips were spayed with 1 ß~ H,O,, followed by heating at 110°C for S min. The toxins were visualized under u.v . (365 ). The strips were en sprayed with l0ß` KOH, heated for 10 min and observed~m~1arlY. Blue fluorecena ; greenish yellow fluorescence, both visualized with lß~ H,O,. " Yellow fluorescence, visualised with lß4 H,O, and 10ßi KOH. Sr-I showed a main peak with a shoulder on its preceding edge followed by a smaller peak (Fig . SC). The retention times of the peaks did not correspond to STX or neoSTX . The shoulder, however, was identified as neoSTX from its retention time (8 .9 min) . Fraction Sr-II showed a large peak, which was identified as STX by its retention time (11 .9 min), with a shoulder on its preceding edge (Fig . SC).
Reference standards
0 0 ~ O"
~- 67X3.2 "4, t STX.n~STX
r"c s o " Gr
Fr, fiL 'Reef I Fr, ~-I speclss;ns" ~ Fr, ~-II Fr . ¢g-I 'Ko11nq Fr . ~k-II snect:sens' Fr, I1lç-m Fr, Slç
r N sis~" o " m I 0
.
.
~
~
I , 0.5
~ "
Py :AcOEt :AcOH :H20 (15:5 :3 :4)
I ~ 1,0
Fta. 4. Twx-urea cxao~wTOanwrFtr of Ateraatis TowN . Thin-layer chromatography was Performed on silica gd LHP-K plats (Whatman) using a solvent system of pyrWine-ethyl acetate-acedc acid-water (15 :5 :3:4). After devdopmeat toxins were detected as in elatmphoresb (see legend in Fig. 3) .
710
TAMAO NOCiUCHl d al.
A B
_ rr . ~-a
0
nwB~S
0
10 ~o Il~t~ntSoa sir (~ial
o
m a~bntion tir (~Sa1
0
oa
10 1~t~nsim tir l~Inl
FIC . S . RICH PERFORMANCE LIQUID CHROMATOORAPI{Y OF SOME FRACTIONS OF Atergatis ToxINS. Reverse-phase high performance liquid chromatography wag arried out on an AM-314 (YMC) column (0 .6 x 30 cm), using heptanesulfonic add as an ion-pairing reagent, with 1~ McOH-0 .03 M potassium phosphate (pH 7.0) for Gr and Gk-I-III, and with 20% McOH-0.05 M potassium phosphau (pH 7.0) for Sr-I, -II and Sk. The periodate reagent (SULLIVAN et d., 1983) wag used to detect PSP. The fluorogenic reaction wsa performed at 65°C for 0.7 min. The fluorophors formed wen monitoredat 38B nm with 344 nm exdtation. TTXwag detected by replacing the periodau reagent with 3 N NaOH, heatingthe reaction mixtureat 100°C for 0.4 min and monitoring the fluorescence at SOS nm with 381 nm exdtation.
As shown in Fig. SB, fraction Gk-I gave two peaks, each with a trailing shoulder, when 3 N NaOH was used for detection. Both of these peaks remain unidentified, but the shoulder of the main peak was identified as TTX from its retention time (10.9 min) . Fraction Gk-II showed a dominant peak, whose retention time (10.9 min) agreed well with that of TTX (Fig. SB). Fraction GK-III showed several peaks, one of which (retention time 11 .1 min), in spite of a little lag of retention time, was judged to be TTX from cochromatography (data not shown) . In addition, small amounts of GT7C,.., were detected when the periodate reagent was used (Fig. SA). In fraction Sk neoST7{ and STX were detected, along with several unknown components (data not shown) . DISCUSSION
`Reef specimens' were found to contain ST?C and a neoSTX-related substance as major components, along with GT7C's and a small amount of a TTX-related substance as minor components, by electrophoresis, thin-layer chromatography and high performance liquid chromatography . We previously obtained (NOOUCHI et al., 1984) essentially the same results for `Ishigaki specimens' of A. floridus. On the other hand, it was also shown that `Kojima specimens' possessed TTX and unidentified TTX-related substances) as major
Toxin Composition of a Xanthid Crab
711
toxins, along with some PSP components, such as neoSTX; STX and GTX's, as minor toxins . This toxin composition is similar to that of A . floridus inhabiting the Pacific coasts of Japan proper (NOGUCHI et al., 1983). Thus the toxin compositions of the specimens collected from two adjacent areas almost simultaneously were extremely different from each other . This supports the view that the toxin of A. fTorzdus is derived from an exogenous origin, such as its foods or symbiotic microorganisms, rather than being of endogenous origin . The present results, along with other toxicity data, suggest that TTX might be an essential constituent for this crab and that the crab could be additionally toxic due to the harboring of a PSP-producing organisms) which lives indigenously on coral reefs. In this connection, KOTAKI et al. (1983) detected PSP in a calcareous alga Jania sp. inhabiting tropical and subtropical seas and presume its exclusive involvement in the intoxication of xanthid crabs. The intoxication of those crabs, however, can not be unambiguously accounted for by this alga alone, because the PSP content of this alga generally is very low and because A. floridus also often contains TTX as the major toxin, as demonstrated here and as reported previously (NocucHl et al., 1983). Acknowledgemarts-The authors express sincere thanks to Diraxor Dr A. ToMOw and the staffof the Yaeyama Branch of the Okinawa Fisheries Expaimental Station for their kind collaboration in colkcting the crab specimens. Thanks are also due to Professor M. Is:www, University of New Hampshire, U.S .A., for reading the manuscript of the preaeat paper. REFERENCES Hwssusoro, Y. (1979) Toxic crabs. In: Mmine Toxins mid Ot1Yer Blo~scttvr Marlne MetabolUes, p. 59. Tokyo: Japan Scientific Societies Press. Kwwwawrw, T. (1978) Assay method for paralytic shellfish pooson . Ia : Food XyaieneEmminatlon Mmueal, Vol. 2, p. 240. Tokyo: Japan Food Hygiene Assodation. Korws;t, Y., TwJnu, M., OSEm~tw, Y. and YwsuMOro, T. (1983) Identification of a calareous red alga as the Primary soura of paralytic shellfish toxins is coral reef crabs and gastropods . Bull. Jprt. Soc. scient. Firh. ~!, 283. KOYAMw, K., NoaucEn, T., Ucnw, Y. and Hwsw~toro, K. (1981) Occurrence of neosaxitootin and otherparalytic ahellf-ash poisons in toxic crabs belonging to the family Xanthidae . Ball. Jpn. Soc. scient. F7sh . 47, %S . KOYAMw, K., NoaucFn, T., Uw, A. and Hws~oro, K. (1983) Resistibility of toxic and nontoxic crabs against paralytk shellfish poison and tetrodotoxin . Bull. Jpn. Soc. acknt. FYsh. 49, 485. NAOnsHUU, Y., MwauYwMw, J., Noaucxt, T. and Hwsttutoro, K. (198 Separation of paralytic shellfish poisom and tetrodotoxin by ion-pairing liquid chromatography . Bull. Jpn. Soc. acknt. FYsk . (in press). Noaucxi, T., Uzu, A., KoYwatw, K., Mwawwntw, J., Nwa~swMw, Y. and HwsFttworo, K. (1983) Oaurreace of tetrodotoxin as the m~ior toxin in a xanthid aab AtergattrJlorldus. Ball. Jpn. Soc. acieat. FYrle. 49, 1887 . Noaucsi, T., Uzu, A., Dw~ao (KoYwsaw), K., SHIDA, Y. and Hwsx~MOro, K. (1984) A tetrodotoxin-like substance as a minor toxin in the xanthid cnb AtergatisJlorldus. Ta:ricnrt 22, 423. NooucFU, T., Dwioo, K., Aawtuww, O. and HwstiIMOro, K. (198 Release of paralytic shellfish poison from the exoskeleton of a xanthid crab Zastrnus aureus. In: l'nvceadings ojthe Thud International Corfferrna on Taxé lNnoJlaadlate Blooms . New York : EhKVier (in press). Sut.t.wwN, J. J. and Iwwoxw, W. T. (1983) High pressure liquid chromatographic determination of toxins associated with paralytic shellfish poisoning . J. Ass. ojf. analyt. CNevn. f6, 297.