Tazkn~, Vol . 30, No. 12, pp. 1539-1544, 1992 . Rinted in Cm~t Brilain .

0041-0101/92 SS.00 + .00 ® 1992 Rrpmon Rea Ltd

TOXIN COMPOSITION OF RESTING CYSTS OF ALEXANDRIUM TAMARENSE (DINOPHYCEAE) YASUKAT$U OSFIIMA, 3 CHRLSTOPHEIt J . BOLCH2

and GUSTAAF

M . HALLEGRAEFF 3

'Department of Applied Biological Chemistry, Faculty of Agriculture, Tohoku University, Tsutsumidori Amamiya, Aoba-Ku, Sendai 981, Japan ; 'CSIRO Division of Fisheries, Marine Laboratories, GPO Boz 1538, Hobart, Tasmania 7001, Australia; and 'Department of Plant Science, University of Tasmania, GPO Hox 252 C, Hobart, Tasmania 7001, Australia (Received 18 March 1992; accepted 4 Auguct 1992)

Y. OstmKw, C. J . BOLCH and G. M. HALLEGRAEFF . Tôxin composition of resting cysts of Alexandrium tamarense (Dinophyceae). Toxicon 30, 1539-1544, 1992.-Paralytic shellfish toxin composition in the resting cysts of the dinoflagellate Alexandrium tamarense was investigated by means of high performance liquid chromatography . A comparison was made between cysts collected from ship ballast tank sediments, natural population of motile vegetative cells collected from the area where ballast water was taken, as well as cultured vegetative cells established from the cysts and the natural plankton bloom. Total toxin concentration of the cysts (595 mmole/cell) was six-fold higher than that of the natural population of vegetative cells. They contained the same ten toxic components but in different relative abundances. The higher proportion of 11-a-hydroxysulfate epimers in the cysts suggests that the biosynthesis of toxins is halted at an early stage in cyst formation.

INTRODUCTION

cysts (hypnozygotes) of toxin dinoflagellates such as Alexandrium ( _ Gonyaulax, Protogonyaulax) tamarense (Lebour) Balech can be a direct source of contamination of shellfish with paralytic shellfish toxins, especially in deeper waters and in seasons when there are few vegetative dinoflagellate cells in the water column . There exist considerable discrepancies in the literature regarding the relative toxicity of resting cysts and motile vegetative cells of the dinoflagellates . Some workers claim that cysts may be five to ten times more toxic than the corresponding motile stages (DALE et al., 1978 ; OsHnKw et al., 1982; LIRDWITAYAPRA3IT et al., 1990), while others consider cysts to be of equal (WI-nTe and Lswls, 1982) or even lesser toxicity (C~>?LLA et al., 1990). Potentially misleading comparisons may have resulted from examining natural cysts and cultured vegetative cells, as it has been repeatedly demonstrated that cultured Alexandrium cells are typically less toxic than wild cells from the same region (WHITE, 1986) and that toxicity levels of motile cells can vary according to culture conditions (BOYER et al., 1987; BOCZAR et al., 1986; OGwTw et al., 1987 ; Oslm~rw et al., 1990). Moreover, in most studies total toxicity was estimated using the mouse bioassay, and variations in toxicity caused by changes in the individual toxins thus have gone undetected . THE RasTIxc

1539

1540

Y. OSHIMA et al.

An opportunity to examine the toxin profiles of Alexartdrittm tarnarense resting cysts by sensitive high performance liquid chromatography (Os>:u1KA et al., 1989) arose during a study of the transfer of toxic dinoflagellate cysts via ships' ballast water (HALLEGRASFF and BOLCH, 1991). A woodchip cargo vessel which filled its ballast tanks during dinoflagellate bloom in the port of Muroran, Hokkaido, Japan, was found to contain some 300 million cysts upon arrival in the port of Eden, Australia. In the present work, we compare the toxin concentration and composition of these ballast water dinoflagellate cysts with those of natural and cultured motile cells collected from the Japanese port of origin in Funks Bay. Comparisons were also made with the toxin composition of clonal dinoflagellate cultures established through germination of the ballast water cysts. MATERIALS AND METHODS Dinoflggellate cysts Sediment containing Alexartdrium dinoflagellate cysts was collected from the bottom of the ballast tanks of a woodchip cargo vessel arriving in the port of Eden, Australia, on 20 July 1989 . This ship had filled its tanks (25,000 tonnes volume) at the port of Muroran, Japan, on I~ July 1989 and sailed 4900 miles to Australia without stopover in other port . Cysts for toxin analysis were stored in the dark at 4°C until examination in November 1989 . Sediment samples were mixed with filtered seawater and sonicated for 40 sec (Labsonic Homogenizer, Hraun, intermediate probe, 40 Vii to dislodge detritus particles. The sample was then screened through a 901mm sieve and cysts concentrated on to a 20 pm sieve. Individual cysts were isolated by micropipette under a Zeisa Axioplan microscope and washed several times in sterile culture medium . Two separate cyst samples were analyzed, one containing 140 and one containing 100 cells, respectively. Notwal vegetative cells Natural vegetative cells of A. tamarense were wllected 3 km off the coast of Mori in Funks Bay on 17 July 1989. Plankton was collected from 20 m depth by vertical tows with a plankton net of 20 pm mesh size and this material was gently screened through 90, and 30 pm nylon meshes . The organisms trapped on 30 tcm mesh were resuspended in filtered sea water and subjected to microscopic cell counting and toxin analysis . DinoJlgqellate cultures Individual cysts were placed into 55 mm polycarbonate petri dishes containing 10 ml of GSe medium (Buctcauax et al., 1989) and incubated at 15°C at a light intensity of 80 pEm- Sec-' (12 hr :12 hr light : dark cycle) . From these cysts motile vegetative cells germinated about 6 months later and clonal (non-sexual) dinoflagellate cultures were established for toxin analysis . Clonal vultures were also isolated from a natural A. tmnarense bloom in Funks Bay on 17 July 1989 and grown in f/2 medium (Guu .t.~ttn and Rrrt~e, 1962) under the same light and temperature conditions as above. Toxin onalysis Known concentrations of purified cysts or motile cells were placed in a 1 .5 ml or 10 ml plastic tube and seawattr or culture medium was removed by aspiration after centrifugation at 1000a for (0 min. The cells were re-suspended in 100-500Pl of o-5 N acetic acid and homogenized by sonication (three successive 30 sec treatments) using Labsonic Homogenizer equipped with a fine probe. The weight of the contents of the tube was measured prior to sonication to estimate the total volume. The supernatant obtained after centrifugation at 3000a for l0 min was passed through an ultrafiltration membrane (Ultrafree C3GC, Millipore) and 10 pl each of filtrate was subjected to analysis. HPLC using ion-pair chromatography and post column derivatization was carried out as described previously (OsFmsw et al., 1989), with alight modifications. A silica~l base reversed-phase column (Develosil G8-5, 0.46 x 15 cm, Nomura Chemical, Seto, Japan) and the following three mobile phases (flow rate 0.8 ml/min) were used for separation of the different toxin groups: (a) 2mM 1-heptanesulfonic acid in 10 mM ammonium phosphate buffer (pH 7.1) for the gonysutoxin gtuup, (b) 2 mM 1-heptaneaulfonic acid in 30 mM ammonium phosphate buffer (pH 7. i) : acetonitrile (100 : 5) for the saxitoxin group, and (c) 1 mM tetrabutylammonium phosphate solution adjusted to pH 6.0 with acetic acid for C1~4 toxins . The eluate from the column was continuously mixed with 7 mM periodic acid in 50 mM sodium phosphate buffer (pH 9.0) at 0.4 ml/min, heated at 65°C by passing through Teflon tubing (0.5 mm id, 10 m long), and then mixed with 0.5 N acetic acid at

Toxins in Atexmidriwn Resting Cysts

1541

0.4 ml/min just before entering the monitor. The fluoromonitor was set at 330 and 390 nm for excitation and emission wavelengths, respectively . Either a Hitachi L-6000 HPLC equipped wiW F-1050 fluoromonitor or an ETP KORTEC K34M equipped with Shimadzu RF-530 fluoromonitor was used . As external standard, mixtures of individual toxins which had been calibrated separately by the nitrogen measurements with combustion analysis were used . The following abbreviations of toxins are used hereafter: STX, saxitoxin; neoSTX, neosaxitoxin ; GTX1~rTX4, gonyautoxins 1~; dcGTX2 and dcGTX3, decarbamoylgonysutoxins 2 and 3. Total toxicity in the samples was calculated from the individual toxin concentrations based on the following specific toxicities of each toxin (in mouse units per Eanole) and expressed in either mouse unit (MU) or picogram STX equivalent (pg STX eq.) : C1-15, C2-240, GTX1-2470, GTX2-890, GTX3-1580, GTX4-1800, dcGTX2380, dcGTX3-940, neoSTX-2300, STX-2480 . RESULTS AND DISCUSSION

All available evidence indicates that the dinoflagellate cyst material examined in this work derived from a bloom of Alexandrium tamarense which occurred in Funks Bay, Japan, in June-July 1989 . The port of Muroran is located on the west side of Funks Bay where A. tamarense and associated toxicity of cultured scallops occurred in springsummer of almost every year since first recognized in 1978 (N1SFmIwMA et al., 1979). According to the plankton monitoring by the Hokkaido Hakodate Fisheries Station (HAYASHI, personal communication), in 1989 the population density of A . tamarense started to increase in the beginning of June and reached a maximum in mid-July . The dinoflagellate was observed all over the bay and a maximum density of 40,000 cells litre was observed at a station 5 km off Otoshibe on 11 July 1989 . At the time of our plankton sampling, the cell density of A. tamarense at the surface was around 5000 cells/litre, which included both vegetative cells and planozygotea . No hypnozygotes were observed in the plankton hauls. Although the sampling point was 40 km from the port of Muroran, the general distribution of the dinoflagellate around Funks Bay indicated that the organisms belonged to one and the same population. Since the minimum clearance from the bottom of the ship at Muroran wharf to the ballast water intake port was 3 m, it is unlikely that the old cysts resuspended from harbor sediments were taken into the tank. Microscopic observation of the ballast sediments showed large concentrations of cylindrical hypnozygotes with rounded end (27 t 2.5 fan diameter, 48 f 5 pm long) and a small number of planozygote stages. Cysts incubated on 21 September 1989 failed to germinate until February 1990, suggestive of a mandatory dormancy period of approximately 6 months (ANDmesoN, 1980). These observations confirm that we are dealing with newly formed cysts produced in the water column, most likely from planozygotes when water was taken into the ballast tanks, rather than mattue cysts resuspended from harbor sediments. The cysts had all walls which were more resistant than those of vegetative cells, but less than 4% of intact cysts were observed microscopically after toxin extraction. The high sensitivity of our analytical system and high toxicity of the cysts enabled us to determine the complete toxin profile on as little as 100 cysts. The chromatograms of a typical cyst sample are shown in Fig. 1 and the results of all analyses are summarized in Fig. 2. Total toxin contents and toxicities based on HPLC analyses are given in Table 1 . Cysts contained the same toxins (STX, neoSTX, GTX1-GTX4, dcGTX2, dcGTX3, C1 and C2) as vegetative cells but there were discrepancies in the relative abundance of the toxins. Total toxin content in cysts (average 595 mmole/cell) was approximately six times higher than that of the natural motile dinoflagellate cells collected from Funks Bay. Mouse toxicity of the cysts (1 .2 x 10 -3 MU/all or 270 pg STX eq ./cell) was also much higher than that of the natural and cultured vegetative cells.

154 2

Y . OSHIMA et al.

~al 5-

10 . 15 . 20 . cbl 5= 10= 15=

cl,c2 neoSTX STX

~cSTX

neoSTX STX

C4 FiG. 1 . ClfleoattAroaxAr~.S of roxnvs IN Alexandrium cysts Frost BAr i . .er sEDD~xr (LEFr) wrrH THo~ of RF~+EaENCE srANDAxDS (aIGHT). (a), (b) and (c) indicate mobile phases used for analyses and referred in the text . An extract

equivalent to approximately ten cysts was applied for each analysis .

Total toxin contents in clonal cultures established from both natural bloom and ballast cysts were variable (29-64 and 11-38 mmole/cell, respectively), but significantly lower than that in the natural motile cells. Both cultured cells contained higher proportion of C2 (50

T

UU ~FHFFFp ~ ooooo~

vv

°~ e

UUHFFFFFy~~ oooooa ve

i e

aera CiV x~~(x ' ~FH FFF~~Ffpyl ooooa~ vv

i e

Clonai eultun Irom ey~t

px Cr~7KKxfC FFFFFF~y oooooa

vv

FIG. T. RELATIVE ABUNDANCE OF EACH TOXIN (mole%) IN CYSr3 AND VEGETATIVE CELLS OF

Alexandrium tmnarense .

Y e

Toxina in Akxandrnun Resting Cysts Test.s 1. Tox~x coxrerrts erro zoxtcrrffs

Cyst from ballast tanks" Natural vegetative cellst Clonai culture from bloom$ Clonai culture from cysts¢

oF Akxandrfum

1543

tmnaraue cvsrs exu vsoEreTtve

Toxin content (fmole/cell)

(l0 - a MU/cell)

594.5 92.3 f 5.7 39.1 t20.1 21 .4 t 11 .0

1168 141 f9.2 41 .8 f22.2 18.4 f 9.8

r .~ c

Toxicity (pg STX eq ./cell) 269 32.5 t 2.1 9.6 t 5.1 4.2 f 2.2

" Average of two samples. t Average and S.D. of three samples collected on the same day. $ Average and S.D. of dupGcatod analysis of three clones. § Average and S.D. of six clones obtained by germination of two cysts.

and 70%) than natural organisms (20%) . High content of this low toxic component resulted in much lower toxicity levels in culture than those of the dinoflagellate from nature. The discrepancy in toxin profiles between cyst and natural vegetative cell reflects the biosynthesis of toxins in different parts of the dinoflagellate life cycle. In cysts, contents of a-epimers of 11-hydrozysulfate (GTX1, GTX2 and C1) were nearly three times higher than those of corresponding ß-epimers (GTX4, GTX3 and C2). In contrast, vegetative cells contained mostly ß~pimers, as reported previously (HALL et al., 1990; OSHQYIA et al., 1990). The ß-epimers are thought to be the first biosynthesis products, which then slowly equilibrate to the chemically more stable a-epimers. Thus toxin profiles in the cysts indicate that toxin production was halted during an early stage ofcyst formation . It is also postulated that toxin biosynthesis may be accelerated or catabolic pathways may halt at an early stage of zygosis and the toxins thus could accumulate by continuing biosynthesis for some time. CEMBELLA et al. (1990) also reported a reverse proportion for GTX2/GTX3 but not for C1/C2, and a significant increase of STX contents in cysts. Comparing the toxicity levels with previous investigators, much lower values of 73.9 fmole/cell were reported for artificially induced A. catenella cysts (LIItDWITAYAPRASIT et al., 1990), 3 x 10-' MU/cell for natural cysts from Ofunato Bay (OSHIMA et al., 1982) and 5 pg STX eq./cell (CEMBELLA et al., 1990), 20-50 pg STX eq./cell ~H>TE and LEwls, 1982) for natural population of Eastern North American cysts. Values comparable to our study of 50-200 (SELVIN et al., 1984) and 200 pg STX eq./cell (DALE et al., 1978) have also been reported. The discrepancy may be due to genetic and physiological difference in local dinoflagellate populations as well as differences in the methodology used.

Acbwwkdganenta--The authors wish to express their sincere thanks to Mr Pe»e Hwxsox of the Australian l~uarantine and Inspection Service for collection of the ballast tank cyst material and Dr Yesuo Tom of the Hokkaido Fisheries Cooperative Association for cooperation with plankton sampling in Funks Bay. We also thank Dr Tent~o HeYt~n of the Hokkaido Hakodate Fisheries Experimental Station for the phytoplankton data in Funks Bay, Mr Htnaw6t KOBAYAa1n of the Japanese Shipowneis' Association for the information on ship operation, and Mr RwY Baowt+ of the Tasmania Department of Sea Fisheries, Hobart, Tasmania, Australia, for the use of HPLC equipment during the first author's visit to Hobart. This study was partly supported by Fishing Industry Research and Couneil Development Grant 1989/39 and the Monbusyo Intennational Scientific Program (Joint Research) of the Ministry of Education and Gtiilture, Japan.

1544

Y . OSHIMA et al. REFERENCES

AxneßsoN, D . M . (1980) Effect of temperature conditioning on development and germination of Gonyaulax tarrrarensis (Dinophycese) hypnorygotes . J. Phycol. 16, lfiCr172 . Br .wcrcarrxx, S . L, Hwr.r .eoxw~F, G. M . and BOLCFr, C . J. (1989) Vegetative reproduction and sexual Gfe cycle of the toxic dinoflagellate Gymnodinium catrnatum Graham from Tasmania, Australia . J. Phycol. 25, 577-590. Hoczwx, B. A ., Bmz.err, M . A ., Lrsrox, J ., Suu .rvwx, J . J . and TwYr.ore, F . J . R. (1986) Paralytic shellfish toxins in Protogonyaulax tamarensis and Protogonyaulax catenella in axenic culture. Plant Physiol. 88, 1285-1290 . Hovea, G . L ., Sur,LrvAN, J. J., Arvnrxsox, R. J ., Hwxarsorr, P. J. and Twvr.oa, F. J. R . (1987) Effects of nutrient limitation on toxin production and composition in the marine dinofiagellate Protogonyaulax tamarensis . Mar. Biol. 96, 123-128 . G~r.r.w, A. D ., Dr~ro~, C. and Tuxoeorr, J . (1990) Toxin composition of alternative life history stages of Alexandrium excavatum as determined by high-performance liquid chromatography. In : Toxic Marine Phytoplankton, pp . 333-338 (Gaw~r,r, E., SuxnsraöM, B ., Enrrx, L . and Arvnnesox, D ., Eds) . New York : Elsevier . Dwr.E, B., Yt:xzscrr, C . M . and Htrasr, J . W . (1978) Toxicity in resting cysts of the red-tide dinoflagellate Gonysulax excavatum from deeper water coastal sediments . Science 201, 1223-1225. Gtm,r.wxn, R. R. L . and RYrt~rt, J . H . (1962) Studies of marine planktonic diatoms . I . Cyclotella nana Hustedt and Detonula conjervacea (Cleve) Gran . Can. Microbiology 8, 229-239 . Hwr .r., S ., Srarcxwß~r~s, G., Moczrnr.owsxr, E., RwvmmawN, A. and Rrrcnwanr, P . B. ((990) The saxitoxins : sources, chemistry and pharmacology. In: Marine Toxins: Origin, Structure, and Molecular Pharmacology, pp . 295 (Hwta., S . and STxrcxwrtrz, G ., Eds). Washington DC : The American Chemical Society . HALLEORAFFF, G . M, and BOI.CÜ, C. J. (1991) Transport of toxic dinofiagellate cysts via ships' ballast water . Mar. Pollut. Bull. 22, 27-30 . Lrnnwrrwvwpxwsrr, T., Nrsino, S ., Moxrwxt, S . and OtwcFU, T. (1990) The biochemical processes during cyst formation in Akxandrium catenella. In : Toxic Marine Phytoplankton, pp. 294-299 (Gxwt~t w, E., StrxnsrxBM, B ., Enr ea, L . and AxnsasoN, D ., Eds) . New York: Elsevier . Nrsrmrwstw, Y ., Ucrnnw, T . and Sw~ro, N . (1979) Gonyaulaax catenella-like species causing paralytic shellfish poisoning in Funka Bay, Hokkaido, 1978. Mon . Rep . Hokkaido Fish . Res . Inst. 36, 65-73 . OGATA, T ., Isttnbwxu, T . and Konwstw, M. (1987) Effect of water temperature and light intensity on growth rate and toxicity change in Protogonyaulax tamarensis. Mar . Biol. 95, 217-220 . Osr~w, Y., SINGH, H . T ., Furcuvo, Y . and YASUMOTO, T. (1982) Identification and toxicity of the resting cysts of Prorogonyaulax found in Ofunato Bay . Nippon Suis . Gakk . 48, 1303-1305 . Ostm~w, Y., SUCiINO, K . and Ywsttxto~ro, T . (1989) Latest advances in HPLC analysis of paralytic shellfish toxins. In: Mycotoxins and Phycotoxins 88, pp . 319-326 (Nwrow, S ., Hwsrn~roro, K. and Uexo, Y ., Eds). Amsterdam : Elsevier. Ost~w, Y ., Suarxo, K ., Irwatrxw, H ., HrROrw, M . and YA3UMOT0, T. (1990) Comparative study of paralytic shellfish toxin profiles of dinoflagellates and contaminated shellfish . In: Toxic Marine Phytoplankton, pp . 391396 (Gxwr~r.r, E ., St~xBM, B ., Ent.ee, L . and AxneasoN, D ., Eds) . New York: Elsevier . Ser.vnv, R. C., Lewrs, C . M ., Yentrscrr, C . M . and Huasr, J . W . (1984) Seasonal persistence of resting cyst toxicity is the dinoflagellate Gonyaulax tamarenris var. excavata . Toxicon 22, 817-820. Wta~re, A. W. (198 High toxin content in the dinoflagellate Gonyaulax excavata in nature . Toxicon 24, 60510 . Wry, A . W. and Lewrs, C . M . (1982) Resting cysts of the toxic ned tide dinofiagellate Gonyaulax excavata in Bay of Fundy sediments . Can. J. Fish. Aqua . Sci. 39, 1185-1194 .