Response of Mice to Gambierdiscus toxicus Toxin

BONNIE A. KELLEY, DAVID J. JOLLOW, EDWIN 1. FELTON, MICHAEL S. VOEGTLlNE, and THOMAS B. HIGERD

Introduction Toxin(s) extracted from cultured G. clarified by centrifugation, the super­ toxicus injected intraperitoneally (i.p.) natant dried, and the resulting solids Ciguatera seafood poisoning is a seri­ into laboratory mice is reported to evoke weighed and dissolved in absolute meth­ ous human illness brought on by ingest­ gross symptoms indistinguishable from anol. This suspension was filtered, ing certain coral reef-associated fish in those reported for partially purified fish designated as crude dinoflagellate ex­ tropical and subtropical regions. The extracts containing ciguatoxin (Hoffman tract, and stored at 4°C. toxin carried by these fish was first et aI., 1983; Sawyer et aI., 1984). Both Assays for toxicity were conducted on isolated by Scheuer et al. (1967). The toxins exhibit similar dose-response ICR female, ICR male, and C57BL/6 presence of the toxin has been con­ curves and both elicit in mice a strik­ female mice all weighing approximate­ firmed in red snapper, Lutjanus bohor, ing hypothermia, which is reversed by ly 20 g each. Animals were maintained moray eel, Gymnothorax javanicus, increasing ambient temperature (Sawyer on Wayne Laboratory Animal diets amberjack, Seriola aureovittata, and et al., 1984). It is important, therefore, (Lab-Blox) 1 and water, ad libitum. A others. Toxic fish probably acquire that the biological and chemical rela­ known quantity of the crude dinoflagel­ ciguatoxin through their diet, and that tionship between these toxins be clari­ late extract was resuspended in phos­ one source of toxin is probably the ben­ fied. In this investigation, we have phate-buffered saline (PBS) containing thic dinoflagellate, Gambierdiscus tox­ utilized the mouse bioassay to examine 5 percent Tween-80 and administered icus, found in association with certain further the biological activities of the (in 0.2 rnl aliquots unless otherwise macroalgae of coral reefs (Yasumoto et dinoflagellate toxin for comparison with stated) to mice intraperitoneally, intra­ al., ICJ77). Extracts oflaboratory grown ciguatoxin and to gain an estimation of venously (Lv.), and by gavage. Control cultures of G. toxicus, however, have the biological half-life of G. toxicus animals received an equal volume of the yielded a more polar toxin similar to toxin in mice. solvent vehicle. Lethality was recorded maitotoxin described from surgeonfish, at 48 hours for some studies, and time­ Acanthurus sp. (Yasumoto et aI., 1CJ79). Materials and Methods to-death after injection was recorded for The relationship of this dinoflagellate Cells of G. toxicus, Adachi and Fuku­ other studies. One mouse unit (MU) of toxin to ciguatoxin is unknown, though ya, were supplied by Rick York of the toxicity is defined as that quantity of one possible explanation is that the dino­ Hawaii Institute of Marine Biology. The crude extract capable of eliciting a fatal flagellate toxin becomes chemically isolate, clone T-39, was cultured in dose in half of the test animals by 48 converted as it is metabolized during 100-liter vats in F/2 medium supple­ hours. Body temperatures were re­ transfer through the marine food web. mented with a seaweed extract. Har­ corded using a YSI rectal probe (Model vested dinoflagellate cells were shipped 43TA, Yellow Springs Instrument Co., to South Carolina in aqueous methanol, Yellow Springs, OH) within 2 hours and upon arrival, cells were extracted after administration to obtain early in­ ABSIRACF-Response to toxins extracted for 7 days at room temperature in meth­ dications of toxicity. from cultured Gambierdiscus toxicus was anol:water (80:20). The suspension was For the retention time study, a stock evaluated in mice. Toxin preparations ad­ solution of dinoflagellate extract (13.3 ministered intraperitoneally or intravenously gave similar dose-response curves, whereas Bonnie A. Kelley is with the Department of MU/rnl) was prepared in PBS contain­ Biology, Pembroke State University, Pembroke, ing 5 percent Tween-80 (67 ~ extract/ oral administration elicited no response. NC 28372; David 1. Jollow is with the Department Time-to-death determination was dose of Pharmacology, Medical University of South rnI). Each ICR female mouse was ad­ dependent and was quantitated to the dose­ Carolina, Charleston, SC 29425; Edward T. Felton ministered 0.15 rnl (10 JAg) i.p. of the response based on lethality. The 48-hour and Michael S. Voegtline are with the Department stock solution or 0.15 rnl of a 1:3 dilu­ lethality dose curves for the dinoflagellate of Basic and Clinical Immunology and Micro­ toxin were comparable to those previously biology, Medical University of South Carolina, tion; a reduced volume was used due to publishedfor ciguatoxin extracted from fish, Charleston, SC 29425; and Thomas B Higerd is administration of multiple injections. whereas the time-to-death curves showed a also with the Department of Basic and Clinical Immunology and with the Charleston Laboratory, One control group received 10 JAg of tox- strong difference. The W so response of NMFS Southeast Fisheries Center, p.o. Box 12607, scheduled multiple injections suggested a 4­ Charleston, SC 29425. This paper is Publication to 8-hour half-life for toxin activity in the no. 775 from the Department ofBasic and Clinical 'Reference to trade names or commercial firms mouse model. Sex and strain ofthe mouse Immunology and Microbiology, Medical Univer­ does not imply endorsement by the National did not affect susceptibility. sity of South Carolina. Marine Fisheries Service, NOAA.

48(4), 1986 35 100 • ic extract as a single injection, and death) than mice injected i.p. The tox­ ICR, female another control group received 3.3 ~g icokinetics of the dinoflagellate and fish in a single injection. Each test animal It toxins are not known. is interesting 50 in the eight remaining groups received that the time to develop overt symptoms, three equally spaced injections over a such as hypothermia in mice (Sawyer et time interval ranging from 0 to 30 hours al., 1984) or clinical symptoms in man between individual injections. Lethality (Bagnis et aI., 1979), occurs over a was recorded after 48 hours of the last period of hours. In view of the high lipo­ 100 • injection. philicity of these toxins, which should ~ ICR, male enhance their equilibration across mem- ~ Results and Discussion branes and promote their accessibility ~ 50 to the site(s) of action, a much more g Sex and Strain Differences u rapid response, perhaps within minutes, ~ Determination of the dose-response would be expected. relationship to the crude dinoflagellate The lack of any response, including 100 i.p. to ICR female, extract administered temperature depression in mice admin­ C571BL. female ICR male, and C57BL/6 female mice istered the dinoflagellate extract by • suggested that sex and strain of the gavage, raises interesting questions re­ mouse did not markedly influence tox­ garding the toxin's rate of absorption, 50 icity (Fig. 1). The LDso's for ICR fe­ physical or biochemical inactivation, male, ICR male, and C57BL/6 female etc. Preliminary studies (not reported), mice were estimated to be 260 ~g/kg, in which the toxicity of the dinoflagel­ 8.04 0.1 0.2 0.4 1 2 393 ~glkg, and 192 ~glkg, respectively. late extract was not lost when incubated G. toxicus extract (mg/kg) ICR females were used in later tests. under acidic conditions or with various The i.p. injection of ciguatoxin ex­ mouse tissue extracts, suggest that sim­ Figure I.-Percent response (death) in tracts into laboratory mice has been a ple physical or enzymatic inactivation 48 hours displayed by three mouse reliable bioassay method for ciguatera­ was not responsible for abrogation of strains to i. p. administration of G. tox­ icus extract diluted in PBS. Data associated toxins and has been widely dinoflagellate toxicity when adminis­ points represent a minimum of 8 adopted by investigators of this field. tered orally. The observation that the animals. The variability that may be present in route of administration influenced the the assay due to the sex and strain of dose-dependence of the toxicity was first mice used has not been studied. The made by Baden2 and is in contrast to results of this limited study indicated the animal response with fish toxin Table 1.-The effect of administration route on lethal­ that no apparent difference should be (putative ciguatoxin) described by Ban­ ity of G. toxlcus extract. ner et al. (1960), in which an equiva­ expected in dose-response with respect Adminis- No. of Time to to sex and strain of the test animal. lent response was obtained whether tration ani- Percent death 2 mice were injected with toxic fish ex­ route Amount rnals' response (h) Route of Administration tracts or fed the extract equivalent to I.p. 50 ~g 6 100 4to 22

To determine the effect of the route twice the injected dose. I.v. 50 ~g 6 100 1 to 3 of administration on the toxicity of G. Gavage 100 ~g 4 a toxicus extract, animals received the Determination of Mean Death Time 'Female ICR mice. toxin in three ways: Intraperitoneally, 'Percentage of fatalities after 48 hours. intravenously, and orally (p.o.). Re­ A standardized mouse bioassay has sponse of mice to 10 MU of toxin ad­ been universally accepted to determine ministered by i.p. and i.v. routes was the toxicity ofsaxitoxin, a marine toxin uniformly fatal (Table 1). Average time­ ofdinoflagellate origin and the causitive to-death, however, was much shorter agent of paralytic shellfish poisoning. ship between the dose of toxin and time­ following i.v. injections. Administration The method described by Schantz et al. to-death in the mouse. For comparison by gavage of approximately 20 MU re­ (1957) involved determination of median with these published reports, we ob­ sulted in no test mice fatalities. death time for a given dilution ofa sus­ served the lethality of the dinoflagellate The potency of the dinoflagellate tox­ pect extract injected i. p. relative to a toxin as judged by time-to-dealth re­ in, as defined by percent fatalities of the saxitoxin standard. For ciguatoxin, sponse in relation to the 48-hour LDso treated mouse populations at 48 hours, Tachibana (1980) reported the relation- response (Fig. 2). Groups of test mice was not significantly different between were injected i.p. with dinoflagellate ex­ mice receiving the dinoflagellate extract tract equivalent to 2-33 MD. At these 2Baden, D. Department of Biochemistry, Univer­ i.p. or i.v. However, mice administered sity of Miami School of Medicine, Miami, FL doses, death occurred between 2 and 16 the toxin i.v. responded earlier (time-to- 33101. Personal commun., 1982. hours, and the mean time-to-death was

36 Marine Fisheries Review '6 80 ~ 14 fish has not been performed. It will be 212 : ofinterest to know ifciguatoxin, unlike ~1O! ~ 60 the dinoflagellate toxin, is retained in the - , c: ~ 8 i o 0. mouse model. ~ 6 : '"~ 40 Biological studies of the toxins asso­ : 4 \ " ~ C' ciated with ciguatera have been mini­ 2 \ " o0':----:'='O------:2l::-0-----:f:,------!40 "Q; 30 Cl. 20 mal. However, many of the biological Mouse units properties described can be very useful Figure 2.-Relationship between G. in defining the various toxins that have toxicus extract dose and time-to-death 0.5 1 2 4• 8 16 •32 been isolated. As a clear understanding ofICR female mice (solid line). Mice Interval between injections (hours) of the biological activities of these received doses expressed in equiva­ Figure 3.-Percent response (death) ciguatera-associated toxins is obtained, lents of mouse units (I MU = 5.2 j.lg within 48 hours following three equal­ extract). Error bars indicate one stan­ ly spaced i.p. injections of G. toxicus their relationship to each other and a dard error ofthe mean (n = 8). Curve extract. more rational approach to minimizing defined by the dotted line represents the impact of ciguatera may become the administration of fish ciguatoxin evident. as reported by Tachibana (1980). of the animal to clear, sequester, or Acknowledgments otherwise inactivate ciguatera-associ­ We thank Marilyn Orvin for her tech­ ated toxins is vitally important to our nical assistance, Rick York for supply­ dose dependent. A nonlinear equation understanding of the human illness, ing extracts ofthe dinoflagellate, and H. estimated the relationship between the ciguatera. Unfortunately, direct methods Hugh Fudenberg for valuable discus­ independent variable (dose) and the de­ of determining the biological fate of sions. This research was supported in pendent variable (time-to-death). Using these toxins is not readily available. To part by NOAA Grants NA80AA-D­ the Simplex method for determining the gain insight as to the "half-life" of the 00101 and NA84A-H-SK098. values of the parameters by least dinoflagellate toxin, portions of an squares, the following equation gave a otherwise lethal dose were administered Literature Cited reasonable approximation of the curve over several time intervals in order to Bagnis, R., T. Kuberski, and S. Laugier. 1979. Clinical observations on 3,009 cases of cigua­ in Figure 2: extrapolate the time point in which the tera (fish poisoning) in the South Pacific. Am. effect of individual doses were no longer 1. Trop Med. Hyg. 28:1067-1073. Dose (MU) = 0.8 (time-to-death additive (Fig. 3). A control group which Banner, A. H., P. 1. Scheuer, S. Sasaki, P. Hel­ frich, and C. B. Alender. 1960. Observations in hours)-l. received a single 10 I-Ig standard dose of on ciguatera-type toxin in fish. Ann. N.Y. Acad. the toxin extract elicited the expected 75 Sci. 90:770-7lr7. Coupled with differences in sensitivity percent lethality response. Another con­ Hoffman, P. A., H. R. Granade, and 1. P. McMil­ lan. 1983. The mouse ciguatoxin bioassay: A to orally administered toxins, the trol group received a single 3.3 I-Ig dose dose-response curve and symptamology ana­ marked difference in time-to-death re­ and exhibited no fatalities. Groups that lysis. Toxicon 21:363-369. Sawyer, P. R., D. 1. JoUow, P. 1. Scheuer, R. York, sponse to ciguatoxin reported by Tachi­ received three injections of 3.3 I-Ig each 1. P. McMillan, N. W. Withers, H. H. Fuden­ bana (1980) and that obtained in our at time intervals ranging from 0 to 4 berg, and T. B. Higerd. 1984. The effect of study with G. toxicus extracts further il­ hours between injection gave between ciguatera-associated toxins on body temperature in mice. In E. P. Ragelis (editor), Seafood tox­ lustrates that the two toxins may not be 40 and 60 percent response, while those ins, p. 321-329. Am. Chern. Soc. Symp. Ser. equivalent biologically. The dashed line animals in groups injected at intervals 262, Wash., D.C. in Figure 2, superimposed on the dino­ between 8 and 30 hours gave less than Schantz, E. 1., 1. D. Mold, D. W. Stanger, 1. Shavel, F. 1. Riel, 1. P. Bowden, I. M. Lynch, flagellate response curve, represents a 10 percent response. From these results, R.1. Wyler, B. Reigel, and H. Sommer. 1957. curve based on the formula reported by it appeared that an estimation of bio­ Paralytic shellfish poisoning. VI. A procedure Tachibana (1980) for ciguatoxin: logical half-life for this toxin in the for the isolation and purification ofthe poison from toxic clam and mussel tissues. 1. Am. mouse model lies somewhere between Chern. Soc. 79:5230-5236. Log dose (MU) = 2 log (1 + time-to­ 4 and 8 hours, and that this toxin may Scheuer, P. 1., W. Takahashi, 1. Tsutsumi, and T. Yoshida. 1967. Ciguatoxin: Isolation and chem­ death in hours)-l. not accumulate in the body. ical nature. Science 155:1267-1268. The persistence for months of the Tachibana, K. 1980. Structural studies on marine The differences between the fish-toxin­ neurological symptoms associated with toxins. Ph.D. Thesis, Univ. Hawaii, Honolulu, 157 p. derived curve and the dinoflagellate­ ciguatera poisoning in humans certain­ Yasumoto, T., I. Nakajima, R. Bagnis, and R. derived-curve are evident from this ly suggests that either ciguatoxin is re­ Adachi. 1977. Finding of a dinoflagellate as a figure. tained and remains active for long likely culprit ofciguatera. Bull. Jpn. Soc. Sci. Fish. 43:1021-1026. periods of time or that the damage ---=_-,--' I. Nakajima, Y. Oshima, and R. Estimation of caused by ciguatoxin is not quickly Bagnis. 1979. A new toxic dinoflagellate found Toxin Retention Time in association with ciguatera. In L. Taylor and repaired. As yet, a similar "half-life" H. H. Seliger (editors), Toxic dinoflagellate Our knowledge regarding the ability study using crude extracts of ciguatera blooms, p. 65-70. Elsevier Sci. Publ., N.Y.

48(4), 1986 37