Thricon. Vol . 30 . No. 11 . pp.1449-1456,1992. Printed in Oreert Britain.
0041-0101/92 35.00 + .040 1992 Perpmon Ren Ltd
C
SEA ANEMONE RADIANTHUS MACRODACTYLUS - A NEW SOURCE OF PALMTOXIN VLADIMIR M . MAHNIR, EMMA P. KOZLOVSKAYA
and
ANATOLY I . KALINOVSKY
Pacific Institute of Bioorganic Chemistry, Russian Academy of Sciences, Vladivostok, 690022, Russia (Received 17 February 1992 ; accepted 5 May 1992) V . M . MAHNIR, E . P. KOZLOVSKAYA
and A.
I . KALINOVsKY .
Sea anemone
Radianthus macrodactylus - a new source of palytoxin. Toxicon 30, 14491456, 1992 .-A very potent non-protein toxin was isolated from the sea anemone Radianthus macrodactylus with the use of chromatography on polytetrafluoroethylene, CM-Sephadex C-25 and by cation and anion exchange HPLC. The toxin was identified as palytoxin by u.v.-, i.r.- and 500 MHz tH NMR spectroscopy . Its LD50 was 0.74 f 0.29,ug/kg by i.v. injection into mice . So far, palytoxin has been associated with zoanthids only. The toxin caused the loss of haemoglobin from erythrocytes but only in about 2 hr after the beginning of incubation, which is characteristic for palytoxin from zoanthids. Sea anemone palytoxin was divided into major and minor components by HPLC . The latter proved to be a product of degradation of palytoxin. INTRODUCTION
(phylum Coelenterata) are a rich source of biologically active compounds such as quaternary ammonium bases (tetramine, anthopleurine), biogenic amines (histamine, serotonine, dopamine), polypeptides (neurotoxins and cardiotoxins, proteinase inhibitors) and proteins (cytotoxins, enzymes) (BEREss, 1982). Polypeptide neurotoxins interacting selectively with sodium channels in excitable membranes have been a subject of major interest for investigators during the last decade (KEM, 1988). For the past several years our group has been studying homologous neurotoxins produced by the sea anemone Radianthus macrodactylus . Five of them were characterized and sequenced (ZYKovA et al., 1985; ZYKOVA and KOZLOVSKAYA, 1989). They are polypeptides with mol. wts of about 5000 which slow down the inactivation process of sodium channels (SOROKINA et al., 1984) . While studying minor toxins of the sea anemone we discovered a new and very potent non-protein toxin. The paper describes the isolation, characterization and identification of the toxin. SEA ANEMONES
MATERIALS AND METHODS Toxin purification procedure In general this is the same procedure as for polypeptide neurotoxins from this sea anemone (MAHNM et al., 1990) . Radianthus macrodactylus (wet weight 16 kg) was collected on reefs in the Seychelles and on the same day ethanol was added to it (about a double volume as compared to the volume of wet animals) . Extraction was allowed to proceed for a week at room temperature and the crude extract was then separated by decantation . 1449
1450
V. M. MAHNIR et al.
The animals were homogenized and then extracted by 60% aqueous ethanol for a week at room temperature . The extract was decanted, centrifuged, combined with the first extract and evaporated under reduced pressure at 40°C . The resulting aqueous extract was applied to a column (10x 80 cm) of Polychrom-1 (polytetrafluoroethylene, Soyuzreactiv, U.S .S.R .), equilibrated with water. Forty litres of water were passed through the column, after that it was eluted by a stepwise gradient of aqueous ethanol (10, 15, 20, 25, 30 and 70%). The toxic fraction eluted by 20-25% ethanol was evaporated under reduced pressure at 40°C and then lyophilized to give 12 .2 g of dry powder. The toxic material was applied to a column (2.8 x 50cm) of CM-Sephadex C-25 equilibrated with 0.01 M ammonium acetate, pH 5. The toxic fraction eluting from the column with 0.01 M ammonium acetate (pH 5.0) was applied to a column (4 x 30 cm) of Polychrom-I for concentration and desalting. The column was washed with water, toxic material was eluted by 60% ethanol, evaporated and lyophilized to give 1 .1 g of dry powder . The next purification step was HPLC on an Ultropac TSK CM-3SW column (2 .15 x 15 cm) with elution by 0.02 M ammonium acetate, pH 5.0. The most toxic fraction was collected, desalted on Polychrom-1, as described above, and lyophilized to give 175 mg of powder. The final purification was achieved by HPLC on an Ultropac TSK DEAF-3SW column (2 .15 x 15 cm). Elution with 0.02 M Tris-HC1 (pH 7.4) was performed followed by NSCI gradient (0-0 .5 M) in 0.04 M Tris-HCI, pH 7.4. Pure toxin eluted in the first chromatographic fraction was desalted by a Polychrom-1 column and lyophilized to give 11 mg of lyophilized powder. Monitoring of eluted material in all cases was performed at 280 nm. Analytical procedures Absorption spectra were recorded on a Cary 219 spectrophotometer (Varian) . Far u.v. CD spectra were obtained on a J5WA spectropolarimeter (Jasco, Japan). The i.r . spectrum was recorded in KBr on a Specord 75 IR spectrometer (Carl Zeiss Jena, Germany) . 'H NMR spectra were recorded on Bruker WM-250 (250 MHz) and Bruker AM-500 (500 MHz) spectrometers at 308K using Dz0 and DMSO-d6, respectively, as solvent. Four groups of ten 26-27 g male white etrainless mice were used for LDm determination by i.v . (lateral tail vein) injection of 0.1 ml of the toxin solution in saline. Deaths were scored 24 hr after injection. The LD yp value was calculated from a plot of the percentage of dead animals (mean±S .E.M .) vs. the dose injected. Mouse erythrocytes were prepared according to HABHtMANN et al. (1981) . One millilitre of 0.5% (v/v) erythrocytes in phosphate-buffered saline (123 mM NaCl/ 16.4 mM NaHP04/3 .7 mM NaH2PO4, pH 7.4) was incubated with the toxin (0.1 or 0.5 pg/ml) at 37°C with periodic shaking. After various time intervals the tubes were centrifuged for 4 min at 1200 x a and the absorbance of the supernatant was measured at 578 nm. As a control 0.5% erythrocytes were used without the toxin. Absorbance at 100% haemolysis was determined by incubating 1 ml of 0.5% erythrocytes with an excess of haemolysin from R. macrodactylus (BaszHEnovsrrv et al., 1988) at 37°C for 10 min. RESULTS
The toxin was strongly adsorbed by hydrophobic resin Polychrom-1 (polytetrafluoroethylene) and weakly interacted with cation or anion exchangers at pH 5.0 and 7.4, respectively . This allowed us to use procedures for the toxin purification similar to those previously described for polypeptide neurotoxins (MAHNm et al., 1990). The toxin gave a very weak absorbance when the method of Lowlty et al. (1951) was used, but could be detected with ninhydrin. Aqueous solution of the toxin produced a foam on agitation. The pure toxin was a white (slightly yellowish tint), amorphous, hydroscopic solid, quite soluble in water and water-ethanolic mixtures, but insoluble in chloroform . Its concentrated solutions are of yellow colour . Non-protein nature of the toxin was judged by negative results of N-terminal amino acid analysis (dansyl version) and acid hydrolytic amino acid analyses . A dark brown residue was formed when the toxin was treated with 6 N HCl containing 0.1 % of phenol at 110°C in a sealed and evacuated ampoule. The injection of the toxin into mice led to convulsions, drowsiness followed by collapse and death. The i.v. LD30 of the toxin was determined to be 0.74 f 0.29 kg/kg. When toxin doses close to the LD, value were injected, the first symptoms of intoxication appeared no sooner than in 1-2 hr, and the animals died no sooner than in 8-10 hr from the injection. The toxin caused loss of haemoglobin from mouse erythrocytes, but it began only after 2 hr from the beginning of the incubation (Fig. 1).
145 1
Palytoxin from a Sea Anemone
w r ó E m
x
FIG . 1 . TDdE DEPENDENCE OF A LOSS OF HAEMOGLOBIN BY WE ERYI7I>LOCY7ES AFFECTED BY THE TOICIN FRom R. macrodactylus.
The toxin concentration was 0 .1 pg/ml (-) and 0.5 pg/ml (---- ), temperature 37°C.
The toxin was divided into major (no less than 95% of material eluted) and minor components by cation or anion exchange HPLC (Fig. 2a). The latter was about a quarter as toxic as the material of the main peak . In spite of all our attempts to purify the toxin from the minor component by HPLC on an Ultropac TSK DEAE-3SW column, it appeared every time when rechromatography was performed. On the contrary, rechromatography of the minor component did not reveal its transformation into the major one.
E
c o
N
C l9
O
to Q
0
4
8 12 0
4
Min
8
12
18
OF THE TOXIN FRom R. macrodactylw ON AN ULTROPAC TSK DEAE-3SW COLUMN (7 .5 x 150 mm). Isocmtic elution of 0.02 M Tris-HCI (pH 7.4) was employed at a flow rate of I ml/min . (b) RP-HPLC of TILE TOXIN FRom R. macrodactyhu ON AN ULTRASPHERE OCrYL COLUMN (10 x 250 mm) . Isocratic elution of 33% acetonitrile containing 0 .05% trifluoroacetic acid was employed at a flow FIG . 2 . (a)
HPLC
rate of 4 ml/min.
1452
V . M . MAHNIR et al. 0 .50
á
0.25
X (nm) FIG . 3 .
AesonpnoN SPECTRA of THE ToxIN FRom R . macrodactylus IN 0.02 M Tris-HO (pH 7.4) BEFORE (-) AND AFTER (---) RP_HPLC . The toxin concentration was 0.6 mg/nil and the optical pathway length was 1 mm.
The absorbance spectrum of the major component showed maxima at 233 and 263 nm and also a small absorbance in the region of 304-360 nm (Fig. 3). The absorbance spectrum of the minor component was the same as that of the major one. Far u.v. CD spectra of both the components were completely located in a positive area and showed bands at 192 and 236 nm (Fig. 4). Reversed phase HPLC showed the presence of the minor component in the toxin as well, and also revealed some asymmetry of the major peak . A portion of the toxin was applied to an Ultrasphere Octyl column using acetonitrile in aqueous trifluoroacetic acid (pH 2.13) as an eluent (Fig. 2b). The u.v. spectra of all the chromatographic fractions were identical to each other (Fig. 3) but differed in A233/A263 ratio (4.21 and 1 .66, respectively)
X (nm) FIG .
4. FAR u.v. CIRCULAR DICHROISM SPECTRA of THE TOXIN FRom R. macrodactylus IN 0.02 M Tris-HC1 (pH 7.4) aEPoRE (-) AND AFTER (_ _ _) RP_HPLC . The toxin concentration was
0 .6
mg/ml and the optical pathway length was I mm .
Palytoxin from a Sea Anemone
I
l 40
I 30
I
I
cm-1
20
I
1453
N
FIG . 5 . THE INFRARED SPECTRUM of THE TOXIN FROM R .
I
I
I
10
x100
macrodactylus.
from the spectrum of the toxin before the chromatography . In CD spectra of these fractions the ellipticity peak at 292 run disappeared almost completely (Fig. 4). RP-HPLC led to about a five-fold decrease of toxicity in the fractions from the major peak and to more than a 150-fold decrease in the minor component. The data obtained suggest that the A2. chromophore is both labile in acidic medium and important for toxicity . We have failed to restore both the A263 peak and the toxicity by incubation of the toxin at pH 7.4 for 48 hr at room temperature. The i.r. spectrum of the toxin (Fig. 5) showed a band at 1670 cm-1 assigning to amide carbonyl . The 250 MHz IH NMR spectra of the major and minor components in DZO proved to be indistinguishable from each other (data not shown) . The 500 MHz IH NMR spectrum of the minor component in DMSO-d, is shown in Figs 6 and 7.
I 8
7
I 6
FIG. 6. THE 500 MHz IH NMR SPECTRUM OF THE MINOR COMPONENT OF THE TOXIN FROM R. macrodactylus IN DIMETHYL SULPHOXID-4 Numbers refer to the carbon position of protons in the palytoxin molecule (MOORE, 1985) .
145 4
V.
1 .4
FIG.
M. MAHNIR
et al.
1 .0
1-
0 .6 PPM
O .í~1 .4 REGIONS of THE 500 MHz 'H NMR specTium of COMPONENT OF THE TOXIN IN DIMETxYL suLPHDxm-d6. Numbers refer to the carbon position of protons in palytoxin molecule (MooRE, 7 . Tm: 4 .9-6 .4
AND
T11F haNoi
1985) .
DISCUSSION Our spectroscopic data show that the substance isolated from the sea anemone R. macrodactylus is indistinguishable from palytoxin as described earlier (MOORE, 1985) . Palytoxin (M 2678.5) was first obtained in a pure form by MOORE and SCHEUER (1971) from zoanthid Palythoa toxica, and it is one of the most potent toxins so far identified (MOORE, 1985). Other toxic Palythoa have been identified as sources of palytoxin (for review see IBRAIInK and SHm, 1987). UEMURA et al. (1985) demonstrated that there are several palytoxin variants with slightly different structures. It should be emphasized that sources other than zoanthids have not been revealed so far. MooRE et al. (1982) suggested that palytoxin is produced by a bacterium (Vibrio sp.) that is symbiotically associated with zoanthids . This point of view has not been fully confirmed, as a stable bacterial culture producing biologically active palytoxin in vitro has not yet been obtained. The u.v. absorbance spectrum of the toxin from R. macrodactylus showed maxima at 233 and 263 run, which are characteristic for palytoxin isolated from various zoanthid species. The first maximum was shown to be due to two conjugated diene systems, whereas the second one was assigned to an N-(3'-hydroxypropyl)traps-3-aminoacrylamide . chromophore was labile to acid and unit of the palytoxin molecule. The palytoxin i.2 base. Its destruction led to the toxin inactivation (MooRE, 1985) . Like palytoxin, the toxin from R. macrodactylus proved to be labile in acidic medium . Both its u.v. absorbance at 263 nm and toxicity were reduced when trifluoroacetic acid was used as a buffer (pH 2.13). The i.r. spectra of the toxin from R. macrodactylus (Fig. 4) and palytoxin from zoanthids (BEREss et al., 1983) are very similar though they are not fully identical . This might be explained by different degrees of purity of the toxin preparations, or by the fact that we
Palytoxin from a Sea Anemone
1455
used a sample containing toxin, partly inactivated in the course of RP-HPLC. It should be noted that the i.r. spectrum of the toxin from R. macr6dactyhts showed the 1670 cm -' band characteristic for i.r. spectrums of palytoxin. The 250 MHz tH NMR spectra of the major or minor components of the toxin from R. macrodactylus in D20 are indistinguishable from the 270 MHz rH NMR spectrum of palytoxin from Palythoa tuberculosa published by HIRATA et al. (1981) . The 500 MHz rH NMR spectrum of the minor component in DMSO-db (Figs 6 and 7) was practically identical to that of palytoxin from zoanthid Palythoa toxica published previously (MooRE, 1985). The only discrepancy in our toxin spectrum (Fig. 7) was the presence of an additional broad doublet at 5.57 ppm (double bond protons) . Thus, our data are insufficient to evaluate the changes of palytoxin structure which are responsible for the minor component formation. These changes do not seem very substantial. Palytoxin displays a variety of biological activities (see review by IBRAHIm and SHIER, 1987). It causes a loss of haemoglobin from erythrocytes but needs a relatively long incubation time for this to be observed (HABERMANN et al., 1981 ; HABERMANN, 1989). This delay may be considered as a characteristic feature of palytoxin haemolysis . The kinetic curves, presented in Fig. 1, show that the effect of the toxin from R. macrodactylus on erythrocytes is characterized by a significant latent period . This is absolutely analogous to the effect of zoanthid palytoxin. Palytoxin toxicity varies for different species of animals and depends on route of administration (IBRAHIM and SHIER, 1987; WILES et al., 1974). The i.v. LD,, of palytoxin in mice was determined by Wnm et al. (1974) to be 0 .45lAg/kg. This is somewhat less than the value (0.74 f 0.29 ug/kg) obtained by us for palytoxin from the sea anemone. This discrepancy probably indicates that sea anemone palytoxin may not be identical with previously characterized zoanthid palytoxin. Further investigation should be performed to define chemical or conformational differences between zoanthid palytoxin and palytoxin from a sea anemone as far as between the major and minor forms of the latter . The analysis of palytoxin from R. macrodactylus by HPLC revealed a permanent presence of a small amount of the minor component, though it was carefully removed in a series of rechromatographies . This suggests that the minor component results from a degradation of palytoxin. The asymmetry of the main palytoxin peak revealed by RP-HPLC may be due to the presence of isotoxins, similar to those isolated from zoanthid P. tuberculosa (Uemura et al., 1985). Another explanation may also be suggested. Carbon-13 NMR studies (MooRE, 1985) showed that interaction of palytoxin with reversed phase resins led to artifacts which most likely resulted from an alteration in an equilibrium between a- and ß-pyranose and furanose forms of six-membered hemiketal C(51)-C(55). We probably observed just the same artifacts when we applied sea anemone palytoxin to a reversed phase HPLC column . Acknowledgeneents--We thank Dr T. BALAMOVA (Shemyakin Institute of Bioorganic Chemistry, Moscow) for recording the 500 MHz IH NMR spectrum . REFERENCES BEsess, L. (1982) Biologically active compounds from coelenterates. Pure appl. Chem . 54, 1981-1994. BEarss, L., ZwtcK, J., Kor.KENmocK, H. J., KAut., P. N. and WAnmRmANN, O. (1983) A method for the isolation of the Caribbean palytoxin (C-PTX) from the coelenterate (zooanthid) Palythoa caribaeorwn . Toxicon 21, 285-290.
1456
V.
M. MAHNIR
et al.
BREzHESrovsKiy, P. D ., MONASfYRNAYA, M . M ., KOZLOVSKAYA, E. P. and ELYAKOV, G . B . (1988) Effect of haemolysin from the sea anemone Radianthus macrodactylus on erythrocyte membrane. Dokl. Akad. Nauk
SSSR 299, 748-750 .
HABERmANN, E . (1989) Palytoxin acts through Na',K*-ATPase . Toxicon 27, 1171-1187 . HABEttuANN, E., AHNERT-MLc m, G ., CHHATWAL, G. S . and BEREss, L. (1981) Delayed haemolytic action of palytoxin . General characteristics. Biochhn . biophys . Acts 649, 481-486 . HutATA, Y., UEwuRA, D . and USDA, K. (1981) Structure of palytoxin. In : Proc. 1st Int . Conf. Chem . Biotechnol. Biol. Active Nat. Products (Varna, Bulgaria), Main plenary lectures, pp. 79-86 . Sofia : Bulgarian Academy of Sciences . IBRAmm, A .R . and SmEtt, W . T . (1987) Palytoxin : mechanism of action of a potent marine toxin . J.
Toxicol. Toxin Rev . 6,159-187.
of
Nematocysts, pp . 375-405 . New KEN, W. R . (1988) Sea anemone toxins: structure and action . In : The Biology York : Academic Press. LOWRY, O. H., ROmBROUOH, N . J ., FARR, A . L. and RANDALL, R . J . (1951) Protein measurement with the folin phenol reagent . J. biol. Chem. 193, 265-275 . MAHNIR, V . M ., KOZLOVSKAYA, E . P. and ELYAKOV, G. B . (1990) Modification of carboxyl groups in sea anemone toxin RTX-III from Radkathus macrodactylus. Toxicon 29,1255-1263 . MoostE, R. E. (1985) The structure of palytoxin . In : Fortschr . Chem . Organ. Naturstoffe 48, pp . 81-202 (HERz, W ., GRL%BAcH, H ., KmBY, G . W . and TAmm, CH., Eds) . New York : Springer. MooRE, R. E. and SCHEUER, P . J . (1971) Palytoxin : a new marine toxin from a Coelenterate. Science 172, 495-498 .
MooRE, R . E ., HELt7RtcH, P . and PATTEmN, G . M. L. (1982) The deadly seaweed of Hana . Oceanus 25, 54-63 . SOROKINA, Z . A ., CmzHmAxov, I . V., ELYAKOV, G . B ., KOZLOVSKAYA, E . P . and VOZHZHOVA, E. 1. (1984) A study of sodium channel inactivation mechanism by a toxin from sea anemone Radianthus macrodactylus and by various chemical agents. Phiziol. J. (U.S.S.R .) 30, 571-579 . UEmuRA, D ., HIItATA, Y ., IwASxrrA, T. and NAOKI, H . (1985) Studies on palytoxins. Tetrahedron 41, 1007-1017. WILES, J. S ., VjcK, J . A . and CHRifmNsEN, M. K. (1974) Toxicological evaluation of palytoxin in several animal species . Toxicon 12, 427-433 . ZYKOVA, T . A. and KozLov=AYA, E . P . (1989) Amino acid sequence of neurotoxin I from the sea anemone
Radianthus macrodactylus . Bioorg . KUm. (U.S.S.R.) 15, 1301-1306.
ZYKOVA, T . A ., ViNoxuRov, L. M., KOZLOVSKAYA, E . P. and ELYAxov, G . B . (1985) Amino acid sequence neurotoxin III from the sea anemone Radianthus macrodactylus. Bioorg. Khbn . (U.S.S.R.) 11, 302-310 .
of
0041-0101/92 35.00 + .040 1992 Perpmon Ren Ltd
C
SEA ANEMONE RADIANTHUS MACRODACTYLUS - A NEW SOURCE OF PALMTOXIN VLADIMIR M . MAHNIR, EMMA P. KOZLOVSKAYA
and
ANATOLY I . KALINOVSKY
Pacific Institute of Bioorganic Chemistry, Russian Academy of Sciences, Vladivostok, 690022, Russia (Received 17 February 1992 ; accepted 5 May 1992) V . M . MAHNIR, E . P. KOZLOVSKAYA
and A.
I . KALINOVsKY .
Sea anemone
Radianthus macrodactylus - a new source of palytoxin. Toxicon 30, 14491456, 1992 .-A very potent non-protein toxin was isolated from the sea anemone Radianthus macrodactylus with the use of chromatography on polytetrafluoroethylene, CM-Sephadex C-25 and by cation and anion exchange HPLC. The toxin was identified as palytoxin by u.v.-, i.r.- and 500 MHz tH NMR spectroscopy . Its LD50 was 0.74 f 0.29,ug/kg by i.v. injection into mice . So far, palytoxin has been associated with zoanthids only. The toxin caused the loss of haemoglobin from erythrocytes but only in about 2 hr after the beginning of incubation, which is characteristic for palytoxin from zoanthids. Sea anemone palytoxin was divided into major and minor components by HPLC . The latter proved to be a product of degradation of palytoxin. INTRODUCTION
(phylum Coelenterata) are a rich source of biologically active compounds such as quaternary ammonium bases (tetramine, anthopleurine), biogenic amines (histamine, serotonine, dopamine), polypeptides (neurotoxins and cardiotoxins, proteinase inhibitors) and proteins (cytotoxins, enzymes) (BEREss, 1982). Polypeptide neurotoxins interacting selectively with sodium channels in excitable membranes have been a subject of major interest for investigators during the last decade (KEM, 1988). For the past several years our group has been studying homologous neurotoxins produced by the sea anemone Radianthus macrodactylus . Five of them were characterized and sequenced (ZYKovA et al., 1985; ZYKOVA and KOZLOVSKAYA, 1989). They are polypeptides with mol. wts of about 5000 which slow down the inactivation process of sodium channels (SOROKINA et al., 1984) . While studying minor toxins of the sea anemone we discovered a new and very potent non-protein toxin. The paper describes the isolation, characterization and identification of the toxin. SEA ANEMONES
MATERIALS AND METHODS Toxin purification procedure In general this is the same procedure as for polypeptide neurotoxins from this sea anemone (MAHNM et al., 1990) . Radianthus macrodactylus (wet weight 16 kg) was collected on reefs in the Seychelles and on the same day ethanol was added to it (about a double volume as compared to the volume of wet animals) . Extraction was allowed to proceed for a week at room temperature and the crude extract was then separated by decantation . 1449
1450
V. M. MAHNIR et al.
The animals were homogenized and then extracted by 60% aqueous ethanol for a week at room temperature . The extract was decanted, centrifuged, combined with the first extract and evaporated under reduced pressure at 40°C . The resulting aqueous extract was applied to a column (10x 80 cm) of Polychrom-1 (polytetrafluoroethylene, Soyuzreactiv, U.S .S.R .), equilibrated with water. Forty litres of water were passed through the column, after that it was eluted by a stepwise gradient of aqueous ethanol (10, 15, 20, 25, 30 and 70%). The toxic fraction eluted by 20-25% ethanol was evaporated under reduced pressure at 40°C and then lyophilized to give 12 .2 g of dry powder. The toxic material was applied to a column (2.8 x 50cm) of CM-Sephadex C-25 equilibrated with 0.01 M ammonium acetate, pH 5. The toxic fraction eluting from the column with 0.01 M ammonium acetate (pH 5.0) was applied to a column (4 x 30 cm) of Polychrom-I for concentration and desalting. The column was washed with water, toxic material was eluted by 60% ethanol, evaporated and lyophilized to give 1 .1 g of dry powder . The next purification step was HPLC on an Ultropac TSK CM-3SW column (2 .15 x 15 cm) with elution by 0.02 M ammonium acetate, pH 5.0. The most toxic fraction was collected, desalted on Polychrom-1, as described above, and lyophilized to give 175 mg of powder. The final purification was achieved by HPLC on an Ultropac TSK DEAF-3SW column (2 .15 x 15 cm). Elution with 0.02 M Tris-HC1 (pH 7.4) was performed followed by NSCI gradient (0-0 .5 M) in 0.04 M Tris-HCI, pH 7.4. Pure toxin eluted in the first chromatographic fraction was desalted by a Polychrom-1 column and lyophilized to give 11 mg of lyophilized powder. Monitoring of eluted material in all cases was performed at 280 nm. Analytical procedures Absorption spectra were recorded on a Cary 219 spectrophotometer (Varian) . Far u.v. CD spectra were obtained on a J5WA spectropolarimeter (Jasco, Japan). The i.r . spectrum was recorded in KBr on a Specord 75 IR spectrometer (Carl Zeiss Jena, Germany) . 'H NMR spectra were recorded on Bruker WM-250 (250 MHz) and Bruker AM-500 (500 MHz) spectrometers at 308K using Dz0 and DMSO-d6, respectively, as solvent. Four groups of ten 26-27 g male white etrainless mice were used for LDm determination by i.v . (lateral tail vein) injection of 0.1 ml of the toxin solution in saline. Deaths were scored 24 hr after injection. The LD yp value was calculated from a plot of the percentage of dead animals (mean±S .E.M .) vs. the dose injected. Mouse erythrocytes were prepared according to HABHtMANN et al. (1981) . One millilitre of 0.5% (v/v) erythrocytes in phosphate-buffered saline (123 mM NaCl/ 16.4 mM NaHP04/3 .7 mM NaH2PO4, pH 7.4) was incubated with the toxin (0.1 or 0.5 pg/ml) at 37°C with periodic shaking. After various time intervals the tubes were centrifuged for 4 min at 1200 x a and the absorbance of the supernatant was measured at 578 nm. As a control 0.5% erythrocytes were used without the toxin. Absorbance at 100% haemolysis was determined by incubating 1 ml of 0.5% erythrocytes with an excess of haemolysin from R. macrodactylus (BaszHEnovsrrv et al., 1988) at 37°C for 10 min. RESULTS
The toxin was strongly adsorbed by hydrophobic resin Polychrom-1 (polytetrafluoroethylene) and weakly interacted with cation or anion exchangers at pH 5.0 and 7.4, respectively . This allowed us to use procedures for the toxin purification similar to those previously described for polypeptide neurotoxins (MAHNm et al., 1990). The toxin gave a very weak absorbance when the method of Lowlty et al. (1951) was used, but could be detected with ninhydrin. Aqueous solution of the toxin produced a foam on agitation. The pure toxin was a white (slightly yellowish tint), amorphous, hydroscopic solid, quite soluble in water and water-ethanolic mixtures, but insoluble in chloroform . Its concentrated solutions are of yellow colour . Non-protein nature of the toxin was judged by negative results of N-terminal amino acid analysis (dansyl version) and acid hydrolytic amino acid analyses . A dark brown residue was formed when the toxin was treated with 6 N HCl containing 0.1 % of phenol at 110°C in a sealed and evacuated ampoule. The injection of the toxin into mice led to convulsions, drowsiness followed by collapse and death. The i.v. LD30 of the toxin was determined to be 0.74 f 0.29 kg/kg. When toxin doses close to the LD, value were injected, the first symptoms of intoxication appeared no sooner than in 1-2 hr, and the animals died no sooner than in 8-10 hr from the injection. The toxin caused loss of haemoglobin from mouse erythrocytes, but it began only after 2 hr from the beginning of the incubation (Fig. 1).
145 1
Palytoxin from a Sea Anemone
w r ó E m
x
FIG . 1 . TDdE DEPENDENCE OF A LOSS OF HAEMOGLOBIN BY WE ERYI7I>LOCY7ES AFFECTED BY THE TOICIN FRom R. macrodactylus.
The toxin concentration was 0 .1 pg/ml (-) and 0.5 pg/ml (---- ), temperature 37°C.
The toxin was divided into major (no less than 95% of material eluted) and minor components by cation or anion exchange HPLC (Fig. 2a). The latter was about a quarter as toxic as the material of the main peak . In spite of all our attempts to purify the toxin from the minor component by HPLC on an Ultropac TSK DEAE-3SW column, it appeared every time when rechromatography was performed. On the contrary, rechromatography of the minor component did not reveal its transformation into the major one.
E
c o
N
C l9
O
to Q
0
4
8 12 0
4
Min
8
12
18
OF THE TOXIN FRom R. macrodactylw ON AN ULTROPAC TSK DEAE-3SW COLUMN (7 .5 x 150 mm). Isocmtic elution of 0.02 M Tris-HCI (pH 7.4) was employed at a flow rate of I ml/min . (b) RP-HPLC of TILE TOXIN FRom R. macrodactyhu ON AN ULTRASPHERE OCrYL COLUMN (10 x 250 mm) . Isocratic elution of 33% acetonitrile containing 0 .05% trifluoroacetic acid was employed at a flow FIG . 2 . (a)
HPLC
rate of 4 ml/min.
1452
V . M . MAHNIR et al. 0 .50
á
0.25
X (nm) FIG . 3 .
AesonpnoN SPECTRA of THE ToxIN FRom R . macrodactylus IN 0.02 M Tris-HO (pH 7.4) BEFORE (-) AND AFTER (---) RP_HPLC . The toxin concentration was 0.6 mg/nil and the optical pathway length was 1 mm.
The absorbance spectrum of the major component showed maxima at 233 and 263 nm and also a small absorbance in the region of 304-360 nm (Fig. 3). The absorbance spectrum of the minor component was the same as that of the major one. Far u.v. CD spectra of both the components were completely located in a positive area and showed bands at 192 and 236 nm (Fig. 4). Reversed phase HPLC showed the presence of the minor component in the toxin as well, and also revealed some asymmetry of the major peak . A portion of the toxin was applied to an Ultrasphere Octyl column using acetonitrile in aqueous trifluoroacetic acid (pH 2.13) as an eluent (Fig. 2b). The u.v. spectra of all the chromatographic fractions were identical to each other (Fig. 3) but differed in A233/A263 ratio (4.21 and 1 .66, respectively)
X (nm) FIG .
4. FAR u.v. CIRCULAR DICHROISM SPECTRA of THE TOXIN FRom R. macrodactylus IN 0.02 M Tris-HC1 (pH 7.4) aEPoRE (-) AND AFTER (_ _ _) RP_HPLC . The toxin concentration was
0 .6
mg/ml and the optical pathway length was I mm .
Palytoxin from a Sea Anemone
I
l 40
I 30
I
I
cm-1
20
I
1453
N
FIG . 5 . THE INFRARED SPECTRUM of THE TOXIN FROM R .
I
I
I
10
x100
macrodactylus.
from the spectrum of the toxin before the chromatography . In CD spectra of these fractions the ellipticity peak at 292 run disappeared almost completely (Fig. 4). RP-HPLC led to about a five-fold decrease of toxicity in the fractions from the major peak and to more than a 150-fold decrease in the minor component. The data obtained suggest that the A2. chromophore is both labile in acidic medium and important for toxicity . We have failed to restore both the A263 peak and the toxicity by incubation of the toxin at pH 7.4 for 48 hr at room temperature. The i.r. spectrum of the toxin (Fig. 5) showed a band at 1670 cm-1 assigning to amide carbonyl . The 250 MHz IH NMR spectra of the major and minor components in DZO proved to be indistinguishable from each other (data not shown) . The 500 MHz IH NMR spectrum of the minor component in DMSO-d, is shown in Figs 6 and 7.
I 8
7
I 6
FIG. 6. THE 500 MHz IH NMR SPECTRUM OF THE MINOR COMPONENT OF THE TOXIN FROM R. macrodactylus IN DIMETHYL SULPHOXID-4 Numbers refer to the carbon position of protons in the palytoxin molecule (MOORE, 1985) .
145 4
V.
1 .4
FIG.
M. MAHNIR
et al.
1 .0
1-
0 .6 PPM
O .í~1 .4 REGIONS of THE 500 MHz 'H NMR specTium of COMPONENT OF THE TOXIN IN DIMETxYL suLPHDxm-d6. Numbers refer to the carbon position of protons in palytoxin molecule (MooRE, 7 . Tm: 4 .9-6 .4
AND
T11F haNoi
1985) .
DISCUSSION Our spectroscopic data show that the substance isolated from the sea anemone R. macrodactylus is indistinguishable from palytoxin as described earlier (MOORE, 1985) . Palytoxin (M 2678.5) was first obtained in a pure form by MOORE and SCHEUER (1971) from zoanthid Palythoa toxica, and it is one of the most potent toxins so far identified (MOORE, 1985). Other toxic Palythoa have been identified as sources of palytoxin (for review see IBRAIInK and SHm, 1987). UEMURA et al. (1985) demonstrated that there are several palytoxin variants with slightly different structures. It should be emphasized that sources other than zoanthids have not been revealed so far. MooRE et al. (1982) suggested that palytoxin is produced by a bacterium (Vibrio sp.) that is symbiotically associated with zoanthids . This point of view has not been fully confirmed, as a stable bacterial culture producing biologically active palytoxin in vitro has not yet been obtained. The u.v. absorbance spectrum of the toxin from R. macrodactylus showed maxima at 233 and 263 run, which are characteristic for palytoxin isolated from various zoanthid species. The first maximum was shown to be due to two conjugated diene systems, whereas the second one was assigned to an N-(3'-hydroxypropyl)traps-3-aminoacrylamide . chromophore was labile to acid and unit of the palytoxin molecule. The palytoxin i.2 base. Its destruction led to the toxin inactivation (MooRE, 1985) . Like palytoxin, the toxin from R. macrodactylus proved to be labile in acidic medium . Both its u.v. absorbance at 263 nm and toxicity were reduced when trifluoroacetic acid was used as a buffer (pH 2.13). The i.r. spectra of the toxin from R. macrodactylus (Fig. 4) and palytoxin from zoanthids (BEREss et al., 1983) are very similar though they are not fully identical . This might be explained by different degrees of purity of the toxin preparations, or by the fact that we
Palytoxin from a Sea Anemone
1455
used a sample containing toxin, partly inactivated in the course of RP-HPLC. It should be noted that the i.r. spectrum of the toxin from R. macr6dactyhts showed the 1670 cm -' band characteristic for i.r. spectrums of palytoxin. The 250 MHz tH NMR spectra of the major or minor components of the toxin from R. macrodactylus in D20 are indistinguishable from the 270 MHz rH NMR spectrum of palytoxin from Palythoa tuberculosa published by HIRATA et al. (1981) . The 500 MHz rH NMR spectrum of the minor component in DMSO-db (Figs 6 and 7) was practically identical to that of palytoxin from zoanthid Palythoa toxica published previously (MooRE, 1985). The only discrepancy in our toxin spectrum (Fig. 7) was the presence of an additional broad doublet at 5.57 ppm (double bond protons) . Thus, our data are insufficient to evaluate the changes of palytoxin structure which are responsible for the minor component formation. These changes do not seem very substantial. Palytoxin displays a variety of biological activities (see review by IBRAHIm and SHIER, 1987). It causes a loss of haemoglobin from erythrocytes but needs a relatively long incubation time for this to be observed (HABERMANN et al., 1981 ; HABERMANN, 1989). This delay may be considered as a characteristic feature of palytoxin haemolysis . The kinetic curves, presented in Fig. 1, show that the effect of the toxin from R. macrodactylus on erythrocytes is characterized by a significant latent period . This is absolutely analogous to the effect of zoanthid palytoxin. Palytoxin toxicity varies for different species of animals and depends on route of administration (IBRAHIM and SHIER, 1987; WILES et al., 1974). The i.v. LD,, of palytoxin in mice was determined by Wnm et al. (1974) to be 0 .45lAg/kg. This is somewhat less than the value (0.74 f 0.29 ug/kg) obtained by us for palytoxin from the sea anemone. This discrepancy probably indicates that sea anemone palytoxin may not be identical with previously characterized zoanthid palytoxin. Further investigation should be performed to define chemical or conformational differences between zoanthid palytoxin and palytoxin from a sea anemone as far as between the major and minor forms of the latter . The analysis of palytoxin from R. macrodactylus by HPLC revealed a permanent presence of a small amount of the minor component, though it was carefully removed in a series of rechromatographies . This suggests that the minor component results from a degradation of palytoxin. The asymmetry of the main palytoxin peak revealed by RP-HPLC may be due to the presence of isotoxins, similar to those isolated from zoanthid P. tuberculosa (Uemura et al., 1985). Another explanation may also be suggested. Carbon-13 NMR studies (MooRE, 1985) showed that interaction of palytoxin with reversed phase resins led to artifacts which most likely resulted from an alteration in an equilibrium between a- and ß-pyranose and furanose forms of six-membered hemiketal C(51)-C(55). We probably observed just the same artifacts when we applied sea anemone palytoxin to a reversed phase HPLC column . Acknowledgeneents--We thank Dr T. BALAMOVA (Shemyakin Institute of Bioorganic Chemistry, Moscow) for recording the 500 MHz IH NMR spectrum . REFERENCES BEsess, L. (1982) Biologically active compounds from coelenterates. Pure appl. Chem . 54, 1981-1994. BEarss, L., ZwtcK, J., Kor.KENmocK, H. J., KAut., P. N. and WAnmRmANN, O. (1983) A method for the isolation of the Caribbean palytoxin (C-PTX) from the coelenterate (zooanthid) Palythoa caribaeorwn . Toxicon 21, 285-290.
1456
V.
M. MAHNIR
et al.
BREzHESrovsKiy, P. D ., MONASfYRNAYA, M . M ., KOZLOVSKAYA, E. P. and ELYAKOV, G . B . (1988) Effect of haemolysin from the sea anemone Radianthus macrodactylus on erythrocyte membrane. Dokl. Akad. Nauk
SSSR 299, 748-750 .
HABERmANN, E . (1989) Palytoxin acts through Na',K*-ATPase . Toxicon 27, 1171-1187 . HABEttuANN, E., AHNERT-MLc m, G ., CHHATWAL, G. S . and BEREss, L. (1981) Delayed haemolytic action of palytoxin . General characteristics. Biochhn . biophys . Acts 649, 481-486 . HutATA, Y., UEwuRA, D . and USDA, K. (1981) Structure of palytoxin. In : Proc. 1st Int . Conf. Chem . Biotechnol. Biol. Active Nat. Products (Varna, Bulgaria), Main plenary lectures, pp. 79-86 . Sofia : Bulgarian Academy of Sciences . IBRAmm, A .R . and SmEtt, W . T . (1987) Palytoxin : mechanism of action of a potent marine toxin . J.
Toxicol. Toxin Rev . 6,159-187.
of
Nematocysts, pp . 375-405 . New KEN, W. R . (1988) Sea anemone toxins: structure and action . In : The Biology York : Academic Press. LOWRY, O. H., ROmBROUOH, N . J ., FARR, A . L. and RANDALL, R . J . (1951) Protein measurement with the folin phenol reagent . J. biol. Chem. 193, 265-275 . MAHNIR, V . M ., KOZLOVSKAYA, E . P. and ELYAKOV, G. B . (1990) Modification of carboxyl groups in sea anemone toxin RTX-III from Radkathus macrodactylus. Toxicon 29,1255-1263 . MoostE, R. E. (1985) The structure of palytoxin . In : Fortschr . Chem . Organ. Naturstoffe 48, pp . 81-202 (HERz, W ., GRL%BAcH, H ., KmBY, G . W . and TAmm, CH., Eds) . New York : Springer. MooRE, R. E. and SCHEUER, P . J . (1971) Palytoxin : a new marine toxin from a Coelenterate. Science 172, 495-498 .
MooRE, R . E ., HELt7RtcH, P . and PATTEmN, G . M. L. (1982) The deadly seaweed of Hana . Oceanus 25, 54-63 . SOROKINA, Z . A ., CmzHmAxov, I . V., ELYAKOV, G . B ., KOZLOVSKAYA, E . P . and VOZHZHOVA, E. 1. (1984) A study of sodium channel inactivation mechanism by a toxin from sea anemone Radianthus macrodactylus and by various chemical agents. Phiziol. J. (U.S.S.R .) 30, 571-579 . UEmuRA, D ., HIItATA, Y ., IwASxrrA, T. and NAOKI, H . (1985) Studies on palytoxins. Tetrahedron 41, 1007-1017. WILES, J. S ., VjcK, J . A . and CHRifmNsEN, M. K. (1974) Toxicological evaluation of palytoxin in several animal species . Toxicon 12, 427-433 . ZYKOVA, T . A. and KozLov=AYA, E . P . (1989) Amino acid sequence of neurotoxin I from the sea anemone
Radianthus macrodactylus . Bioorg . KUm. (U.S.S.R.) 15, 1301-1306.
ZYKOVA, T . A ., ViNoxuRov, L. M., KOZLOVSKAYA, E . P. and ELYAxov, G . B . (1985) Amino acid sequence neurotoxin III from the sea anemone Radianthus macrodactylus. Bioorg. Khbn . (U.S.S.R.) 11, 302-310 .
of
