BIOLOGICAL EFFECTS OF SURFACTANTS, III HYDRA AS A HIGHLY SENSITIVE ASSAY ANIMAL

Effects on hydra of five homologous series of surfactants representing nonionic, amphoteric and anionic classes were examined. When concentrations which had no effects were increased tenfold the animal disintegrated within 24 h. Lethal concentrations always coincided with a surface tension of 49 ± 4 dynes/cm. This is probably the level which disrupts the cell membranes. Because the deleterious effects are very rapid and obvious, hydra is a useful organism for monitoring the potential toxicity of detergents in freshwater environments.


INTRODUCTION
A large proportion of the surfactants in commercial use (i.e. detergents, wetting agents, spreaders) end up in surface or ground water. There they may have profound biological effects. In plants, surfactants may inhibit growth (see Parr & Norman, 1965 for a review), drastically alter the cellular ultrastructure  and on occasion enhance growth (Stone, 1960;Ernst et al., 1971). These effects are probably mediated by the adsorption of the surfactants to the cell wall of the plants (Fujita & Koga, 1966).
Little is known about surfactant properties which are responsible for biological effects. Therefore, one of us (R.E.) has synthesised a series of surfactants varying systematically with respect to charge, structure, molecular weight and hydrophiliclipophilic balance. We have used these compounds to investigate the effects of surfactant characteristics on several biological systems.
Hydra is one of the organisms employed in these investigations. This fresh water coelenterate is a useful test organism for two reasons. First, it is ubiquitous throughout freshwater habitats in the United States and other parts of the world. Secondly, it has no protective outer layers of any consequence so that the epithelial cells of the ectoderm are in direct contact with the surrounding aqueous 175 Environ. Pollut. (17) (1978)-© Applied Science Publishers Ltd, England. 1978 Printed in Great Britain environment. Therefore, hydra provides a very sensitive system for studies of surfactant effects on living tissue.

Assay procedure
Ten control animals were maintained in 10 ml 'M' solution at 20° ± 1 °C. They were fed heavily (8-15 larvae) three times per week and washed 6---8 h after feeding. The animals were removed and the dishes thoroughly cleaned once a week. Under these conditions a single hydra produced 10-12 buds in three weeks, which is a typical growth rate. Experimental animals were treated in the same manner except that surfactant dissolved in 'M' solution at a specific concentration was added to the culture medium. Each time the animals were washed they were placed in culture solution containing fresh surfactant.
The number of buds produced between two feedings was measured by counting those that had detached from the adult since the last feeding. Subsequently the buds were removed and the animals fed. Each non-lethal surfactant concentration was assayed in duplicate and the cumulative number of buds produced by the ten animals in each dish over a period of three weeks determined. Lethal concentrations in most cases led to disintegration of the animals one day after exposure to the surfactant. In two cases the animals disintegrated within IO days after the start of treatment.

Synthesis and purification
A total of 24 surfactants belonging to three major groups were synthesised (see Appendix). These consisted of 14 fatty alcohol ethoxylates, 6 sulphobetaines and 4 sodium alkyl sulphates. All compounds were repurified following synthesis to remove reaction by-products or unreacted starting materials (see Appendix, Tables Al, A2, A3).

Surface tension measurements
Determinations of surface tension (ST) for each of the several concentrations of every surfactant were made using the du Nouy surface tension (or ring) method (Schwartz & Perry, 1949) with a Fisher Model 20 Tensiomat.

Effect of surfactants on budding rate and sun·ival
Ten of the twelve fatty alcohol ethoxylates (nonionics) were toxic to hydra at 2 x 10-2 mM and all of these surfactants had a lethal effect at 2 x 10-1 mM within a period of 24h (Table I).
At concentrations of 2 x 10-3 mM or below, the entire range of nonionics Surfactants within a class are identified by then-alkyl group (R). The fatty alcohol ethoxylates are also characterised by the number of ethylene oxide units (n). The general formulae for the three classes are: Fatty alcohol ethoxylate: R-0-{CH 2 CH 2 0)"-H; Sulphobetaines: Sodium alkyl sulphates; R-O-S0 2 0-Na + "Budding rates measured relative to controls: + + +: 100--120 buds (controls); + +: 90-· 100 buds; +: 70--90 buds; ±: 20 buds, lethal to adults within 10 days; O: lethal to adults within 24 h; -: not tested. evaluated had no effect on these animals and budding rates were the same as those "'Of the controls (Fig. 1, Table 1 ). The response of hydra to sulphobetaines showed a direct relation to their molecular weight. Toxicity increased in line with increasing chain length of the lipophile (Table 1). A cumulative suppression on budding rate was further observed with these detergents at concentrations between their lethal and non-toxic levels (Fig. 2). Octyl sulphobetaine at 2 mM and dodecyl sulphobetaine at 2 x 10-1 mM induced the budding rate to fall off after about two weeks of exposure. The  tetradecyl derivative caused budding rates to decline severely within the first week (Fig. 2). A molecular weight-dependent effect was also exhibited by the anionic sodium alkyl sulphates. However, in this case the relation was the opposite of that of the sulphobetaines: the toxicity decreased with increasing molecular weight (Table 1 ).
Of the animals subjected to a lethal concentration (2·0 mM), one member from each class was examined with phase microscopy. Within I 0-15 min of exposure the animals were highly contracted. The ectodermal epithelium was either distorted or already disrupted in the case of the anionic alkyl sulphates. By 24 hall animals in all three classes were reduced to cellular debris.

Effects of surfactants on the surface tension of the culture medium
The surfactant concentration employed made possible observation of effects above and below the critical micelle level. This is reflected in a decreasing surface tension with increasing molarity up to the critical micelle concentration (CMC)

Effects of non ionics
A group of twelve fatty alcohol ethoxylates varying in size of the alkyl group and the number of ethylene oxide units were used to examine the effects of hydrophiliclipophilic properties and balance.
Our results show that the hydrophilic-lipophilic balance had surprisingly little effect on toxicity as in all but two cases the lethal level was 2 x 10-2 mM. This concentration is at the threshold or above the CMC, which for alcohol ethoxylates lies between 2 x 10-2 and 2 x 10-1 mM (Becher, 1967).
The surface tension values drop with increasing detergent concentration and level off as micelles form. This occurs within the same range shown above (Table 2). In addition, detergents above the CMC have solubilisation effects on water insoluble organics such as membrane lipids (McBain, 1950).

Effect of ::witterionics (sulphobetaines)
The six sulphobetaines employed by us covered a range of 8~ 18 carbon atoms in the lipophilic chain. Toxicity to hydra increased with the increase in lipophilic properties of these detergents and correlates well with the logarithmically decreasing CMC of sulphobetaines which accompanies increasing alkyl chain length (Herrmann, 1966). It is also in line with the lower surface tension values of sulphobetaines having higher alkyl radicals (Table 2). Similarly, solubilisation efficiency and extraction of proteins from 3T6 mouse fibroblasts by sulphobetaines improves with increasing alkyl chain length (A. Gonenne & R. Ernst, pers. comm.).
With hydra, lethal concentrations are above the CMC. The sulphobetaines are known to be 'mild' in their effects on biological materials (Allen & Humphries, 1975), so that a considerable concentration of micelles may be necessary for any observable effect on these animals.

Effect of anionics
Alkyl radicals of sodium alkyl sulphates employed by us ranged from 10~16 carbon atoms. Here, toxicity decreased with increasing lipophilic chain length. This, however, was undoubtedly due to reduced water solubility and resulting loss of surfactant activity at the assay temperature.
The temperature at which ionic surfactants and water become homogeneous is known as the Krafft point (Krafft & Wiglow, 1895;Hutchinson &Shinoda, 1967). In this connection, solubilities of sodium alkyl sulphates decrease sharply below their Krafft points with increasing alkyl chain length. These are about 35 °C for sodium tetradecyl sulphate and 45°C for sodium hexadecyl sulphate (Osipow, 1962). The reduced solubility of these surfactants at the assay temperatures would account for their lower toxicity to hydra (Table 1) and their higher surface tensions (Table 2).
In the laboratory, hydra fed heavily every day will double their tissue mass in 3-4 days (David & Campbell, 1972). Instead of increasing in size, the excess tissue is removed by budding in the form of new individuals. Therefore, the budding rate is a direct measure of the growth rate of the tissue mass and this was the parameter measured in these studies.
Budding rates are known to be sensitive to environmental conditions. Under normal circumstances they are proportional to the daily feeding level (Bode, pers. observation; J. Otto & R. D. Campbell, pers. comm.). However, if the culture dishes are not kept clean, the budding rate will decrease despite heavy feeding due to the presence of accumulated detritus and slime exuded by the animals (Bode, pers. observation). Similarly, any surfactants that interfere with growth in hydra would be expected to decrease the budding rate. Surfactants clearly have disruptive effects on cell membranes. Observations on a member from each class subjected to 2·0 mM concentrations indicated that the cell membranes were lysed either rapidly in anionics or more slowly for the nonionics and amphoterics. Other information indicates that the sodium alkyl sulphates and the sulphobetaines are very effective in disrupting membranes. For example, 3 mM sodium dodecyl sulphate (SOS) has been used to dissolve the tentacles of hydra for purposes of isolating nematocysts (Bode & Flick, 1976). In general, SOS and sulphobetaines have been used to solubilise proteins from membranes (Steck & Fox, 1972;Allen & Humphries, 1975).
Thus, the surface tension value of 49 ± 4 dynes/cm may reflect the concentration where the lytic effect of surfactants on hydra membranes is lethal.
Most reports on the biological effects of surfactants are based on experiments involving concentrations which are higher than those found in ground and surface water. Because hydra is very sensitive to chemicals in the environment, due to the absence of a sufficiently protective outer layer, we have been able to employ detergent concentrations which are in line with pollution levels. Therefore our findings are more representative of the situation in the environment.
propanesulton precursor (monitored by infrared spectophotometric procedures for the disappearance of the typical sultone absorption at 1O·3 µ and 11·2 µ). The solutions were then diluted with 50 % aqueous ethanol to give 10 % solids, and 250 ml portions were extracted repeatedly with 100 ml petroleum ether, to remove fatty matter present as impurities. Extracted solutions were agitated with 100 g mixed bed ion exchange resin (Amberlite MB 1, Rohm and Haas Company), filtered and dried.
Sulphobetaines, although having strong anionic and cationic radicals, act as 'ionic neutral' zwitterion compounds and pass through cation and anion exchange resins (Konig, 1972), but ionic impurities are removed. After this the dried, crystalline solids (20-25 g) were stirred into boiling acetone, and the dispersion filtered upon cooling. The yield was recrystallised from acetone once more, the crystals dried at 60°C and percent nitrogen determined by the Kjeldahl method. Molecular weights were calculated on the basis of Kjeldahl nitrogen (Table A-1).   Silica gel G plates are spotted with 50-250 µg of the surfactant and are developed with chloroform-methanol 9: I v/v. Visualising is by spraying first with 0·8 % ninhydrin in ethanol, followed by heating for 10 min at 120 °C, indicating the presence of residual precursor amines, and over-spraying with Dragendorff s reagent to stain the sulphobetaines. Thin layer chromatograms of purified sulphobetaines contain only a single bright yellow spot near the origin.
Anionics. Sodium alkyl sulphate R-O-S0 2 0-Na + (II) R = n-decyl, n-dodecyl, n-tetradecyl, n-hexadecyl, or n-octadecyl, were prepared from fractionally distilled, higher alcohols by sulphation with chlorosulphonic acid, followed by neutralisation with NaOH. For purification, each surfactant was dissolved in 50 % aqueous alcohol to yield a solution containing IO% solids. Unreacted fatty alcohol was removed by repeated extraction of 250 ml of this solution with 100 ml hexane. The extracted solution was dried and the solids redissolved in boiling anhydrous ethanol, filtered hot on a Buchner filter, and the