Isolation and Characterization of the Toxic Component of Enhydrina schistosa ( Common Sea Snake ) Venom *

The toxic component from the venom of the common sea snake (Enhydrina schisfosa) was isolated by passage of the venom through a carboxymethyl cellulose column. The preparation, which was shown to be homogeneous by zonal electrophoresis, ultracentrifugation, isoelectric focusing, and gel filtration, had amedian lethal dose 50 % (LD& toward mice of 0.044 pg per g of body weight. The molecular weight of the toxic protein, as determined by sedimentation velocity and diffusion experiments, was found to be 7200. The minimum molecular weight calculated from amino acid analysis was 7000. The isoelectric point, determined by isoelectric focusing, was 9.2. Although hyaluronidase, alkaline phosphatase, phosphodiesterase, phospholipase A, acetylcholinesterase, DNase, leucine amino peptidase, and clotting activity could all be detected in unfractionated venom; no enzymatic activity could be found in the purified toxin. The compound Nbromosuccinimide completely destroyed the toxicity of the preparation, which, under the conditions employed, strongly suggests that tryptophan is necessary for toxicity.

It is generally believed that the over-all action of snake venoms is caused by the combined effect of a number of different proteins (both enzymatic and nonenzymatic) found in the venoms (l-3). Recent studies have shown that venom from snakes of the family Hydrophiidae (Sea Snake) are much simpler in composition than the venoms of the venomous land snakes (Families: Crotalidae, Viperidae, and Elapidae (4)). However, to date, very few investigations have been undertaken on sea snake venoms. Many of these snakes spend their entire lives in the ocean, which makes capturing enough specimens to do a detailed study of their venom most difficult.
In addition, the amount of venom which can be extracted from a venomous sea snake is only 3 to 4 mg as compared to up to 1 g from land snakes.
In 1961, Carey and Wright (5) studied the venom of Enhydrina this snake is at the neuromuscular junction. Chan and Geh (6) found that this same venom produced muscle weakness and respiratory paralysis, while Cheymol et al. (7) attributed toxicity of this venom to an almost irreversible blocking of specific receptors of the postsynaptic membrane.
The only chemical investigations reported to date on toxins from sea snakes have come from the sea snake subfamily Lat.icaudinae. Tamiya and Arai (8) isolated two tosins from the venom of Laticauda semijasciata.
Each toxin contained 62 amino acids and had a molecular weight of about 6900. Recently, Sato et al. (9) isolated the toxins from the venom Latica.uda laticaudata and Laticauda colubrina. Although the amino acid composition of these two toxins varied slightly from that of the toxins of L. semifasciata, the molecular weight of each was about 7000. This paper will describe a procedure for the isolation of the toxin from the Common Sea Snake (E. schistosu) commonly found in the Gulf of Thailand, Strait of Malacca, Vietnam Coastal waters, and the Bay of Bengal.
This toxin will be shown to be homogeneous by a number of criteria, and report on some of its chemical and physical properties. Sephadex G-10 and G-75 were obtained from Pharmacia.
The Ampholine carrier ampholytes and electrofocusing column used in the electrofocusing experiments were purchased from LKB Instruments (Stockholm, Sweden).
Lecithin was purchased from Mann and purified as previously described (10). Other substrates used in the enzymatic assays were purchased from Sigma, Calbiochem, Fisher, and Mann.
Isolation Procedure-Previously washed and charged Cmcellulose was equilibrated with the initial buffer (0.01 M sodium phosphate, pH 7.5) to be used in the separation.
The flow rate was adjusted to 15 ml per hour and the column was loaded with 0.500 g of E. schistosa venom which had been dissolved in 1.0 ml of 0.01 M sodium phosphate, p1-I 7.5. Fractions were eluted by means of increasing salt concentrations added in a stepwise manner.
The column effluent was continuously monitored with p-toluenesulfonyl-L-arginine methyl ester, acetyl-L-tyrosine TUBE NUMBER ethyl ester, and N-benzoyl-L-tyrosine ethyl ester, were carried out as described previously (11). Acid and alkaline phos- Phosphodiesterase activity was measured (spectrophotometrically) by following the hydrolysis of the substrate calcium bis-(p-nitrophenyl)phosphate (14). Hyaluronidase activity was I.200 followed by measuring the decrease in turbidity of a hyaluronic acid-protein complex (16). Phospholipase -4 was measured V-C titrimetrically as previously described (10). Analytical Ultracentrifugation-L4 Spinco model E ultra-0.800centrifuge with an AN-D rotor, an Epon double sector, and an 2 Epon double sector capillary synthetic boundary cell was used g to measure the sedimentation and diffusion coefficients. Diffusion measurements were made at 9,945 rpm with 0.10 M glycine, pH 9.2, as solvent.
Photographs were taken of the Schlieren i m 0.400 patterns at 4-min intervals for a total of 60 min. Sedimentation velocity measurements were made at 59,780 rpm in the same g -X buffer, where photographs were also taken at 4-min intervals for 60 min. Physical measurements were made with a Nikon profile projector, model 6c microcomparator.
Isoelectric Focusing and Polyacetate Electrophoresis-Isoelectric focusing was performed in a 115-ml column as described previously (11 sample of toxin was first treated with performic acid as described by Schram, Moore, and Bigwood (18).
Toxicity Tests-Toxicity was determined by injecting 0.25 ml of toxin dissolved in 0.85% NaCl solution into the tail vein of 20-g Swiss white mice. Ten to fifteen mice were injected at each of 15 dosage levels for each toxicity determined.
The number of mice which died within 24 hours was recorded, and the median lethal dose 50% (LD,,) determined by the method of Lit&field and Wilcoxon (19). 3. Isoelectric focusing profile of toxin. Details of experiment given in text.

Purification
of Toxin-A typical fractionation started with 500 mg of lyophilized E. schistosa venom. After fractionation with a NaCl salt gradient on a Cm-cellulose column at pH 7.5, the fraction displaying highest toxicity was further purified by passage through second Cm-cellulose column at pH 9.7. A typical fractionation is illustrated in Figs. 1 and 2. The fractionation was independently repeated five times with the same patterns and yields. A summary of the purification is presented in Table I, where it can be seen that E. schistosa consists of five toxic components, the most toxic of which was about 2 times more toxic than the starting material.
The 124 mg of this toxin which was obtained represents about 25% of the original material. Criteria of Purity-Following electrophoresis at pH 8.8 on polyacetate strips; 12 components could be detected in unfractionated venom, three components after passage through the first Cm-cellulose column, and only a single component after passage through the second Cm-cellulose column.
To further establish homogeneity of the toxin, additional electrophoresis experiments were carried out at pH values of 7.5, 6.0, and 5.0. In each experiment, only a single protein band could be detected.
Isolectric focusing of unfractionated venom revealed seven components.
As can be seen from Fig. 3, when the purified toxin was subjected to isolectric focusing under the same conditions, only a single component could be found. Fig. 4 presents the results obtained when the purified toxin was passed through a column of Sephadex G-50. Rechromatography of the peak tube gave a similar molecular sieve pattern with band broadening of a comparable degree. Thus, it can be assumed that the preparation is homogeneous in respect to molecular weight.
Ultracentrifugation of the toxin was performed at four different concentrations, from two separate toxin preparations. In each of the eight experiments, only a single component could be detected.
Physical Parameters-At four different concentrations of toxin, 1, 0.72, 0.52, and 0.38%, sedimentation coefficients of 1.31 S, 1.34 S, 1.37 S, and 1.38 S were obtained, respectively. Extrapolation to zero protein concentration gave a siO,, value of 1.4 S. At the same concentrations of protein, diffusion coefficients (D) of 11.2 X lo+ cm2 per see, 12.2 X lo+ cm2 per set, 13.1 x lo-' cm2 per see, and 13.8 X lo+ cm2 per set were obtained.
Extrapolation to zero protein concentrations gave a %., value of 15.5 X 10V7 cm2 per sec.
By a combination of sedimentation and diffusion coefficients, a molecular weight was determined by means of the Svedberg  Each value represents the average of three independent determinations.
The most probable value represents the average for each amino acid except serine and threonine which are extrapolated to zero hydrolysis time.
Cysteine was determined as cysteic acid after oxidation with performic acid (18). Tryptophan was estimated by the spectrophotometric method of Edelhoch (21)) and by titration with N-bromosuccinimide (22). Each method suggested 1 mole of tryptophan per mole of enzyme. The minimum molecular weight of the toxin as calculated from the amino acid analysis was 6981.
The isoelectric point of the toxin is at or near pH 9.2. This value is taken from isoelectric focusing experiments where the toxin forms a band at pH 9.17 (Fig. 3).
Nonenzymatic Nature of Toxin-The enzymatic activities of both purified toxin and unfractionated common sea snake venom were tested on 16 substrates.
Enzymatic activity toward each of these substrates has been shown to be present in venoms of at least some venomous land snakes. Although clotting activity, hyaluronidase, alkaline phosphatase, phosphodiesterase, deoxyribonuclease, acetylcholinesterase, and leucine aminopeptidase, activities could be detected in unfractionated venom, none of these activities could be detected in the purified toxin.
Both the original venom and the purified toxin are devoid of the following enzymatic activities: ribonuclease, acid phosphatase, amino acid esterase (with N-benzoyl-L-arginine ethyl ester, N-benzoyl-L-tyrosine ethyl ester, p-toluenesulfonyl-n-arginine methyl ester, and acetyl-L-tyrosine ethyl ester as substrates), and proteases (with casein and hemoglobin as substrates). TAULE  has been achieved. The preparation was shown to be homogeneous by a number of criteria.
The toxin appeared as a single, sharp, symmetrical peak when subjected to isoelectric focusing, and to Sephadex gel filtration. Only one band could be detected with zonal electrophoresis on polyacetate at various pH values. The preparation also appeared to settle as a single component when the preparation was centrifuged at 59,780 rpm.
The amino acid analysis of the toxin showed that the toxin consists of 62 amino acids with a molecular weight of about 7000. It would thus appear that this toxin is very similar to the toxins isolated from L. semifasciata captured in Japan by Tamiya and Arai (8) who found 61 amino acids in their toxins with molecular weights of 6900; and to the toxins isolated from L. laticaudata and L. colubrina by Sato et al. (9), who also found 62 amino acids with a molecular lveight of 7000. The toxins isolated from L. semifasciata captured in the Philippines are also found to be similar to that of E. schistosa venom (23). The amino acid composition of this preparation is also similar to the above preparations; the major differences being in a higher threonine content, less proline, and the absence of phenylalanine. This preparation also contained 1 mole of alanine, while the above toxins were devoid of this amino acid. It should also be pointed out, that each of the above studies reported more than one toxic component which agrees with the findings in this investigation.
The toxin may act in a nonenzymatic fashion, as none of the substrates normally acted on by snake venoms were hydrolyzed by the toxin.
This finding is in agreement with Carey and Wright (24), who found that the t.oxic factor of this snake could be separated from phospholipase A by dialysis, and of Ghan and Geh (6) who found that there was no acetylcholine esterase activity in the toxin of this venom.
The inhibition of the toxin by N-bromosuccinimide under the conditions employed, strongly suggests that tryptophan is necessary for the toxic action of this venom. Future studies, which have as their goal the mechanism by which tryptophan is involved, should help us to understand the mode of action of Common Sea Snake venom.