Toxin Production by Isolated from the Florida Keys

JOHN A. BABINCHAK, DAVID J. JOLLOW, MICHAEL S. VOEGTLlNE, and THOMAS B. HIGERD

Introduction al., 1984) were extracted and partitioned compared with a highly toxic clonal G. between solvents ofdifferent polarities. toxicus isolated in Hawaii (Sawyer et al., Ciguatera is a tropical fish-borne The biological and chemical nature, 1984). disease in which the lipid-soluble neuro­ however, ofthe limited quantity of lipid­ toxin, ciguatoxin, is believed to be trans­ soluble toxin obtained by Materials and Methods ferred through the food chain and bio­ these nonstringent extraction, partition­ Benthic dinoflagellate samples were concentrated primarily in carnivorous ing, and purification techniques can collected from the Florida Keys at a site reef fish to levels toxic to humans. Since only be conjecture. Cultured G. toxicus previously involved in ecological studies ecological research in the South Pacific readily produce a toxin distinguished on benthic associated implicated Gambierdiscus toxicus as the from ciguatoxin on the basis of higher with ciguatera (Bomber, 1985). Samples probable cause of ciguatera (Yasumoto molecular weight and lower polarity. were collected in December 1983, and et al., IfJ77), this dinoflagellate had been This toxin has been tentatively identified in February, May, July, and December collected from the wild, grown uni­ as maitotoxin (Yasumoto et aI., lfJ79) , 1984, from Knight Key, which con­ algaliy, and extracted for toxins that have a toxin first isolated from the gut of sistently had a high G. toxicus popula­ been analyzed principally by mouse surgeonfish, Acanthurus sp. (Yasumoto tion. This site, station 3 of Bomber bioassay. et aI., 1fJ76). (1985), is located in Florida Bay, 0.8 kIn Evidence that G. toxicus produces the Unialgal cultures, initiated with up to east of seven-mile bridge (lat. 42°44' fish-extracted ciguatoxin is based on 30 cells, are used in most studies on tox­ 20"N, long.·81°07'I6"W). The area con­ reports in which either wild cells (Yasu­ in production by cultured G. toxicus. sists ofan algal reef depauperate in coral moto et aI., lfJ77, lfJ79; Bagnis et aI., The objectives of this study were to with strong tidal currents and up to 95 1980; Shimizu et aI., 1982) or cultured isolate and culture clonal G. toxicus col­ percent cover by Halimeda spp. cells (Yasumoto et aI., lfJ79; Tindall et lected concurrently from a single site in (Bomber, 1985). Pieces of macroalgae the Florida Keys, and compare the were collected in plastic jars with sea­ quantity and nature of the toxins pro­ water, shaken, and the seawater filtered duced. Toxicity, as measured in mice through a 250 ~m Nitex 1 sieve. The (LD ), was determined using whole filtrate was then refiltered through a 25 ABSTRACT-The toxicities of six clonal so Gambierdiscus toxicus cultures collected cells or nonfractionated extracted tox­ ~m Nitex sieve and the retentate ana­ concurrently from Knight Key, Fla., were ins, and the relative polarity of the ex­ lyzed microscopically for G. toxicus. compared with the toxicity ofthe Hawaiian tracted toxins was analyzed by high Samples with high populations of G. G. toxicus strain, 739. W 50 values obtained performance liquid chromatography toxicus were suspended in 1 liter poly­ using mouse bioassay demonstrated a hundredfold range in whole-cell toxicity. The (HPLC) (Higerd et al., 1986). The toxic carbonate bottles containing 700 rnl Hawaiian and two Floridian strains had characteristics ofthe Florida isolates are filtered (20 ~m) seawater for shipment comparable mouse toxicity (LD5oJ ofabout to the laboratory. 2.5 x 10 4 cells/kg. Two additional groups The medium used for isolation and ofFloridian strains had toxicities ofabout John A. Babinchak and Thomas B. Higerd are with growth of G. toxicus was Provasoli's ES 2 X 105 and >1 million cells/kg, respective­ the Charleston Laboratory, Southeast Fisheries ly. Fractionation of methanol extracts by Center, National Marine Fisheries Service, enriched seawater medium as modified high-performance liquid chromatography NOAA, P.O. Box 12607, Charleston, SC 29412. by 1. West (McLacWan, lfJ73). Seawater suggests that toxins produced by different David 1. Jollow is with the Department of Phar­ macology, Medical University of South Carolina, was collected either at the Florida Keys clones of G. toxicus are indistinguishable Charleston, SC 29425. And Michael S. Voegtline from each other but are more polar than fish and Thomas B. Higerd are with the Department toxin. Isolates ofOstreopsis heptagona, also of Basic and Clinical Immunology and Micro­ 'Mention of trade names or commercial firms does isolatedfrom Knight Key, had relatively low biology, Medical University of South Carolina, not imply endorsement by the National Marine 6 toxicities (W50 > 5 X 10 cells/kg). Charleston, SC 29425. Fisheries Service, NOAA.

48(4), 1986 53 sampling site, or at two sites from 7 to 20 /-1m Nitex filter and resuspending the Table 1.-Whole-eell toxicity of cultured G. tox/cus 10 miles off Charleston, S.c., and fil­ retentate in 30 percent aqueous meth­ clones. tered onsite through a 20 /-1m Nitex sieve anol (v/v). This suspension was evapor­ Cell dia. LDso Clone Source (jAm)' (cells/kg) into 10 to 20 liter polycarbonate carboys. ated to dryness under a stream of nitro­ T39 Tern Island, HI 76 2.5 X 10' (4)' Upon arrival at the laboratory, the water gen. Cd20 Knight Key, FL 79 3.5 x 10' (4) was refiltered through 0.45 /-Iffi cellulose Toxicities of whole cells (LDso) were Cd4 Knight Key, FL 81 4.5 x 10' (3) Cd8 Knight Key, FL 94 2.5 x 10' (2) nitrate membranes in a Nuclepore radial determined by suspending the dried Cd10 Knight Key, FL 81 3.0 x 10' (2) flow cell into sterile polycarbonate car­ sample in phosphate buffered saline Cd9 Knight Key, FL 86 >2.3 x 10' (2) Cd13 Knight Key, FL 77 boys and refrigerated until used to pre­ containing 5 percent Tween 80 and in­ >2.5 x 10' (2) pare media. The vitamin mixture (stored jecting 0.2 ml of appropriate cell con­ 'Cellular measurements were obtained by measuring live individuals at 400 x with bright-field microscopy (n =30). frozen) and enrichments were prepared centrations intraperitoneally into female 'Bioassay analyses (n). in concentrated stocks, filter-sterilized ICR mice weighing approximately 20 g and added aseptically to the seawater. (Kelly et al., 1986). In addition to deter­ The prepared medium was filter-steril­ mining whole cell toxicities, the seven ized through 0.2 /-1m cellulose nitrate G. toxicus clones were grown and har­ membranes and used immediately or vested to obtain total cell densities refrigerated. greater than 1 x 106 cells. These sam­ and toxin production. This precipitate Clonal cultures of G. toxicus and Os­ ples were extracted in 80 percent aque­ was prevented by eliminating the Tris treopsis sp. were established from single ous methanol (v/v) for 48 hours at room buffer from ES medium. cells isolated with a micropipet, washed temperature and bioassayed or frac­ Five 0. heptagona clones isolated three or four times in sterile seawater, tionated by HPLC using a 50-100 per­ from Knight Key, prepared and bio­ and inoculated into 16 x 125 rom culture cent methanol linear gradient. Each assayed using the same procedure de­ tubes containing 5 mI ES medium. Os­ fraction was later assayed for toxicity by scribed for G. toxicus, had relatively treopsis spp. occur with G. toxicus in the mouse bioassay (Higerd et al., low toxicity (LDso > 5 x 106 cells/kg. the Florida sampling site (Bomber, 1986). Mouse mortalities were observed at in­ 1985) and have been shown to be toxic jections of 5 x 106 cells/kg, but higher by Nakajima et a1. (1981). Growth was Results and Discussion dosages were not assayed so a definitive monitored with an inverted microscope, Six G. toxicus clones were isolated LDso could not be calculated. Al­ and ifapparent within 30 days, the cul­ concurrently from a dinoflagellate sam­ though 0. siamensis and 0. ovata were tures were transferred to 25 mI mini­ ple collected at Knight Key, 20 Decem­ shown to be toxic by Nakajima et al. Fernbach flasks or 20 x 150 rom ber 1983 (Bomber, 1985, station 3). The (1981), their toxicities were also> 5 X culture tubes containing 15 mI ES Hawaiian strain, 139, was hand-carried 106 cells/kg. Besada et a1. (1982) re­ medium. After successful growth for from Hawaii to the SEFC Charleston ported no toxicity in cultures 15-30 days, the cultures were placed in Laboratory. Five Ostreopsis clones isolated from the Caribbean Sea. 250 mI polycarbonate Erlenmeyer flasks represent a new species, 0. heptagona, The LDso values using whole cells of containing 100 mI ES medium. Stock with cellular lengths >100 /-1m (Norris G. toxicus are shown in Table 1. The cultures were maintained in 250 mI et al., 1985). G. toxicus clones were Hawaiian clone, 139, was similar in tox­ flasks and transferred to new medium identified microscopically by their icity to Cd20 and Cd4 from the Florida every 30 days. These cultures were cellular shape and the characteristic Keys and to 139 cultures grown in maintained at 27°C under an illumina­ "fishhook" apical pore slot (Adachi and Hawaii. Florida Key clones Cd8 and tion of 30-50 /-IE'M-2'S-1 and a 16:8 Fukuyo, 1979; Taylor, 1979). Further Cd10 were 10 times less toxic, while the hours light:dark cycle without aeration. taxonomic studies are in progress using toxicities of Florida Key clones Cd9 and For toxin bioassay, 100 mI cultures a chloral hydrate-hydriodic acid stain­ Cd13 were just evident with more than were inoculated into 2.8 liter Fernbach ing method (Schmidt et a1., 1978) and 2 x 106 cells/kg. As with 0. hep­ flasks containing 1.5 liters ES medium. scanning electron microscopy. tagona, mortalities were observed, but These cultures were grown for 30-45 Maximal cellular yields obtained for higher dosages were not assayed so a days under incubation conditions de­ G. toxicus cultures grown for toxin bio­ definitive LDso could not be calculated. scribed above. Ostreopsis clones were assays were 1,000 cells/mI. Larger LDso values for nonfractionated meth­ cultured for 10-21 days under the same yields of2,000-4,000 cells/mI have been anol extracts of two Cd20 cultures were conditions. Cell counts were determined reported for cultures of smaller G. tox­ 1 x lOS cells/kg, or about 70 percent in Palmer-Maloney chambers. To com­ icus (35-55 /-1m) (Bagnis et a1., 1980; less toxic, than whole cells. pare cell diameters with other reported Carlson et a1., 1984). On occasion, This is the first report ofcomparative G. toxicus isolates, cells were measured stored ES medium prepared with specif­ toxicities using clonal cultures of G. tox­ during their late log growth phase at ic lots of natural seawater collected both icus isolated concurrently from a single 400x magnification with bright-field from the Florida Keys and South Caro­ site. Two unialgal cultures isolated from microscopy. Cells were harvested by lina coastal waters produced a precipi­ the Florida Keys have been reported in filtering the culture medium through a tate that proved detrimental to growth a previous study (Bergmann and Alam,

54 Marine Fisheries Review Table 2.-Relative HPLC elution times icus (Bergmann and A1am, 1981) since G. toxicus (Sick et aI., 1986). Such a for toxic fractions of extracted G. toxicus. maitotoxin, the primary and most potent technique may be capable of quan­ G. toxicus toxin, is produced late in the titating and comparing the chemical Source of extracted material 'R t growth phase (Yasumoto, et aI., 1979). nature of toxins from clonal G. toxicus G. toxicus (Hawaii) T-39 2.2 G. toxicus (Hawaii) T-39 2.1 Insufficient data in other studies and the cultures and determining the effects that G. toxicus (Florida) Cd-4 2.8 lack of a standardized mouse bioassay cultural parameters have on toxin pro­ G. toxicus (Florida) Cd-4 2.6 G. toxicus (Florida) Cd-10 2.6 made it difficult to compare quantita­ fIles, similar to a study reported for PSP G. toxicus (Florida) Cd-20 2.7 tively the levels of G. toxicus toxicity toxins (Boyer et aI., 1985). Fish (St. Thomas, U.S.V.I.) 4.0 reported in this study with previous in­ 'Ratio of toxic activity elution time vestigations. However, approximate tox­ relative to eiution time for phenol stan­ Acknowledgments dard, where phenol = 1. icities reported for cultured G. toxicus range from 6 X 103 cells/kg (Yasumoto The authors gratefully acknowledge et al., 1979) to 1 X 106 cells/kg (Tin­ the cooperation of Dean Norris and Jeff dall et aI., 1984). Toxicity quantitation Bomber, Florida Institute of Technol­ presently is limited by the nonspecific ogy, Melbourne, Fla., in collecting ben­ nature and lack of precision and ac­ thic dinoflagellates and confirming the 1981) and these cultures had toxicities curacy of the mouse bioassay. identification of Ostreopsis clones. Our in mice of 3.1 x lOs cells/kg and 1.2 x Toxins extracted with aqueous meth­ thanks also to Rick York, Hawaii Insti­ lOs cells/kg. anol were 70 percent less toxic than tute of Marine Biology, Honolulu, for To compare polarities of dinoflagel­ whole cells as measured by mouse bio­ the culture of G. toxicus, strain T39. late toxins, the ratio between the toxic assay, indicating that the toxic moieties fraction elution times from a HPLC col­ of extracted nonfractionated toxins and Literature Cited umn for extracted dinoflagellates and a whole cells may differ or that extraction Adachi, R., and Y. Fukuyo. 1979. The thecal struc­ phenol standard was calculated. Results efficiency requires improvement. Tin­ ture of a marine toxic dinoflagellate Gambier­ were expressed as a relative retention dall et al. (1984) reported ciguatoxin, discus toxicus gen. et sp. nov. collected in a ciguatera-endemic area. Bull. lpn. Soc. Sci. time (R ) , with phenol equivalent to one ciguatoxin derivative, and maito­ t Fish_ 45:67-71. 1.00 (Table 2). Only a single toxic com­ toxin present in fractionated extracts of Bagnis, R., S. Chanteau, E. Chungue, 1. M. ponent was detected in each of the four cultured unialgal G. toxicus. Their Hurtel, T. Yasumoto, and A. Inoue. 1980. G. cultures and they exhibited limited data, however, could not exclude Origins of : A new toxicus dinoflagellate, Gambierdiscus toxicus Adachi similar Rt values. In contrast, the dino­ the possibility of multiple toxins derived and Fukuyo, definitely involved as a causal flagellate toxins were far more polar from carryover of maitotoxin during the agent. Toxicon 18:199-208. Bergmann,1. S., and M. Alam. 1981. On the tox­ than the fish toxin, which eluted much extraction and separation procedures as icity of the ciguatera producing dinoflagellate, later in the linear methanol gradient. observed by Yasumoto et al. (1979) or Gambierdiscus toxicus Adachi and Fukuyo, Since toxic fractions were detected by interchangeable toxin forms like that isolated from thhe Florida Keys. 1. Environ. Sci. Health, Part A: Environ. Sci. Eng. 16(5): mouse bioassay, this technique may have reported for purified ciguatoxin (Nukina 493-500. missed toxic fractions since LDso et aI., 1984). The HPLC system used Besada, E. G., L. A. Loeblich, and A. R. Loeb­ values ~ 1.0 X 106 cell/kg were not in the current study permitted distinct lich III. 1982. Observations on tropical ben­ thic dinoflagellates from ciguatera-endemic assayed. Cells extracted from the re­ separation of several toxins, but because areas: Coolia, Gambierdiscus, and Ostreopsis. maining three G. toxicus samples did of limited amounts ofextracted material, Bull. Mar. Sci. 32:723-735. Bomber, 1. W. 1985. Ecological studies of ben­ not provide enough toxic activity to be toxins present in minor quantities may thic dinoflagellates associated with ciguatera detected with this procedure. The have gone undetected. from the Florida Keys. M.S. Thesis, Fla. Inst. chromatographic details of this HPLC There is a need to identify and quanti­ Technol., Melbourne, 104 p. Boyer, G. L., 1. S. Sullivan, R. 1. Anderson, P. technique can be found in an accom­ tate toxins present in cultured benthic 1. Harrison, and F. 1. R. Taylor. 1985. Toxin panying conference paper (Higerd et al., dinoflagellates associated with cigua­ production in three isolates of Protogonyaulax 1986). tera. An extremely valuable analytical sp. In 0. M. Anderson, A. W. White, and 0. 1. Baden (editors), Toxic dinoflagellates, p. This study demonstrated a hundred­ method for determing total toxin profIles 281-286. Elsevier Sci. Publ., NY. fold variation in toxicity of clones iso­ has been developed for paralytic shell­ Carlson, R. D., G. Morey-Gaines, 0. R. Tindall, and R. W. Dickey. 1984. Ecology of toxic lated concurrently, which would suggest fish poisoning (PSP) toxins (Sullivan dinoflagellates from the Caribbean Sea: Effects that toxin production may vary in and Wekell, 1984; Sullivan et al., 1985). of macroalgal extracts on growth in culture. In natural populations of G. toxicus. Uni­ This HPLC analysis uses derivatives of E. P. Ragelis (editor), Seafood toxins, p. 271­ 287. Am. Chern. Soc. Symp. Ser. 262, Wash., algal cultures, initially containing PSP toxins to fluorometrically detect D.C. several competing clones, might pro­ toxic fractions separated in an HPLC Higerd, T. B., 1. A. Babinchak, P. 1. Scheuer, and duce variable toxin profIles until a single column at concentrations four times D. 1. lollow. 1986. Resolution of ciguatera­ associated toxins using high-perfonnance liquid clone became dominant and the other more sensitive than mouse bioassays. A chromatography (HPLC). Mar. Fish. Rev. clones were lost through transfer dilu­ technique similar to the HPLC pro­ 48(4):23-29. Kelley, B. A., D. 1. lollow, E. T. Felton, M. S. tion. Cultural parameters can also in­ cedure for PSP toxins has been devel­ Voegtline, and T. B. Higerd. 1986. Response fluence the relative toxicity of G. tox- oped for detecting toxin extracted from of mice to Gambierdiscus toxicus toxin. Mar.

48(4), 1986 55 Fish. Rev. 48(4):35-38. and 1. W. Hastings. 1978. Comparative study Wash., D.C. McLachlan, 1. 1m. Growth media-marine. In 1. of luminescent and nonluminescent strains of Taylor, F. 1. R. 1979. A description of the benthic R. Stein (editor), Handbook of phycological Gonyaulax excauata (Pyrrhophyta). 1. Phycol. dinoflagellate associated with maitotoxin and methods: Culture methods and growth measure­ 14:5-9. ciguatoxin, including observations on Hawaiian ments, p. 37-45. Camb. Univ. Press, NY. Shimizu, Y, H. Shimizu, P. 1. Scheuer, Y material. In D. L. Taylor and H. H. Seliger Nakajima, I., Y Oshima, and T. Yasumoto. 198!. Hokama, M. Oyama, and 1. T. Miyahara. 1982. (editors), Toxic dinoflagellate blooms, p. 71-76. Toxicity of benthic dinoflagellates in Okinawa. Gambierdiscus toxicus, a ciguatera-causing Elsevier Sci., Pub!., NY. Bull. Jpn. Soc. Sci. Fish. 47:1029-1033. dinoflagellate from Hawaii. Bull. Jpn. Soc. Sci. Tindall, D. R., R. W. Dickey, R. D. Carlson, and Norris, D. R., 1. W. Bomber, and E. Balech. 1985. Fish. 48:811-813. G. Morey-Gaines. 1984. Ciguatoxigenic dino­ Benthic dinoflagellates associated with cigua­ Sick, L. V., D. C. Hansen, 1. A. Babinchak, and flagellates from the Caribbean Sea. In E. P. tera from the Florida Keys. I. Ostreopsis hep­ T. B. Higerd. 1986. An HPLC-fluorescence Ragelis (editor), Seafood toxins, p. 225-240. tagona sp., nov. In D. M. Anderson, A. W. method for identifying a toxic fraction extracted Am. Chern. Soc. Symp. Ser. 262, Wash., D.C. White, and D. 1. Baden (editors), Toxic dino­ from the marine dinoflagellate Gambierdiscus Yasumoto, T., R. Bagnis, and 1. P. Vemoux. 1976. flagellates, p. 39-44. Elsevier Sci. Pub!., N.Y. toxicus. Mar. Fish. Rev. 48(4):29-35. Toxicity of the surgeonfishes - II. Properties Nukina, M., L. M. Koyanayi, and P. 1. Scheuer. Sullivan, 1. 1., 1. Jonas-Davies, and L. L. Ken­ of the principal water-soluble toxin. Bull. Jpn. 1984. Two interchangeable forms ofciguatoxin. tala. 1985. The determination of PSP toxins by Soc. Sci. Fish. 42:359-365. Toxicon 22:169-176. HPLC and autoanalyzer. In D. M. Anderson, _-'---:-'---:--:-' I. Nakajima, R. Bagnis, and R. Sawyer, P. R., D. 1. Jollow, P. 1. Scheuer, R. York, A. W. White, and D. 1. Baden (editors), Toxic Adachi. 1977. Finding of a dinoflagellate as a 1. P. McMillan, N. W. Withers, H. H. Fuden­ dinoflagellates, p. 275-280. Elsevier Sci. Pub!., likely culprit ofciguatera. Bull. Jpn. Soc. Sci. berg, and T. B. Higerd. 1984. Effect of cigua­ NY. Fish. 43:1021-1026. tera-associated toxins on body temperature in ~---,--~_, and M. M. Wekell. 1984. Deter­ ____ , I. Nakajima, Y. Oshima, and R. mice. In E. P. Ragelis (editor), Seafood tox­ mination of paralytic shellfish poisoning tox­ Bagnis. 1979. A new toxic dinoflagellate found ins, p. 321-329. Am. Chern. Soc. Symp. Ser. ins by high pressure liquid chromatography. In in association with ciguatera. In D. L. Taylor 262, Wash., D.C. E. P. Ragelis (editor), Seafood toxins, p. and H. H. Seliger (editors), Toxic dinoflagel­ Schmidt, R. 1., V. D. Gooch, A. R. Loeblich III, 197-205. Am. Chern. Soc. Symp. Ser. 262, late blooms. p. 65-70. Elsevier Sci. Pub!., NY.

56 Marine Fisheries Review