ABSTRACT
Several studies have been made on the bacteriostatic and fungicidal properties of oxine (8-hydroxyquinoline) and other chelating agents, and it has been suggested that they might act, in part at least, by depriving cells of essential trace metals (Albert, 1951; Martell & Calvin, 1952).
In an attempt to determine the trace element requirements of mammalian spermatozoa, White (1955) investigated the effect of several chelating agents on motility. A number were found to be toxic to ram, bull, rabbit and human spermatozoa. In fact some (e.g. sodium diethyl dithiocarbamate, 1-nitroso-2-naphthol and ethyl potassium xanthate) were much more spermicidal than oxine, which has long been used as a chemical contraceptive in the form of Chinasol. Tests with bull semen showed that cobalt reduced the toxicity of sodium diethyl-dithiocarbamate, nitrosonaphthol and o-phenanthroline. Other mixtures of heavy metals and chelating agents were, however, more toxic than the chelating agents alone.
Further studies on the interaction between metals and chelating agents in bull, ram, rabbit and human spermatozoa are presented in this paper.
MATERIALS AND METHODS
Semen and diluents
The method of collecting semen was the same as described previously (White, 1954). An isotonic diluent of pH 7·0 was used in all experiments and had the following composition: 0·032 M-NaH2PO4H2O, 0·048M-NajHPO4.I2H2O, 0·036 M-NaCl, 0·004 M-KC1,0·022 M-fructose or glucose. Glucose was only used in the experiments with purified diluents (Tables 3 and 5). The metals referred to in Table 3 are copper, cobalt, manganese, iron and zinc, which were all added as the A.R. salts. The first three were used at a final ionic concentration of 0·05 mg./100 ml. and the last two at a concentration of 0·20 mg./100 ml. The chelating agents were B.D.H. Laboratory reagents.
Motility scoring
The semen was diluted 1 in 40 in small tubes for motility observations which were made at 37 ° C. Motility was scored at hourly intervals over a 4 hr. period by the system of Emmens (1947). Full motility was rated as 4 and complete immotility as zero. Quarter grades were frequently used, and in order to eliminate fractions, the total 4 hr. score for each ejaculate has in all cases been multiplied by 4. The highest possible score, based on five observations, is 80.
Statistical analysis
Where responses have been virtually of the all or none type, statistical analysis has not been undertaken. In other cases, experiments have been evaluated by a t-test or by the analysis of variance (Fisher, 1955), with isolation of sums of squares attributable to differences between ejaculates and treatments. The total figure for the period has been used as unit observation and the treatment-ejaculate interaction mean square as the error term.
Purification of diluents
Diluents were purified by the absorption method recommended by Donald, Passey & Swaby (1952). Five g. Al2O3 (British Drug Houses Ltd. chromatographic absorption analysis material) was added to 500 ml. of diluent in a conical flask and heated with frequent swirling for 1 hr. The flask was then left to cool in a refrigerator over night and filtered through an acid-treated, sintered glass filter connected by glass joints to a Pyrex filter flask.
Test of purity of diluents
The following test for heavy metals is based on the method of Stout & Amon (1939). Ten millilitres of glass-distilled water, 5 ml. of redistilled chloroform, 10 drops of 0·1% dithiozone in redistilled chloroform and 5 drops of ammonia (redistilled, approximately 6 N) were shaken in a small Pyrex separating funnel. The chloroform layer was run off and 5 ml. more added and shaken. The chloro-form (which was found to separate much more rapidly in cold weather on exposure to an infra-red lamp) was now colourless or a slight blue. Ten millilitres of the test solution were then added to the funnel which was shaken and the colour of the chloroform layer noted. The funnel was rinsed with 0·5 N-HCI between determinations to remove traces of metal adhering to the wall (see Piper, 1944). Standard zinc sulphate solutions made up with glass-distilled water gave the following colours: 1·00 μg. Zn/ml., intense red; 0·10 μg. Zn/ml., purple pink; 0·01 μg. Zn/ml., no colour or light blue.
Using this method the following approximate estimates of heavy metal content were obtained: tap water, 1·00 μg./ml.; diluents in either glass-distilled or ordinary distilled water, 0·10 μ.g./ml.; ordinary distilled water, 0·02/μg./ml.; glass-distilled water, purified ordinary distilled water, purified diluents, 0·01 μg. /ml.
RESULTS
Previous tests (White, 1955) showed that cobalt, copper, zinc and cadmium could influence the toxicity of some chelating agents to bull spermatozoa. This observation has been checked for the semen of this species and extended to other spermatozoa. The result of a comparative study on bull, ram, rabbit and human semen is set out in Table 1. The chelating agents used were nitrosonaphthol, o-phenanthroline, sodium diethyldithiocarbamate, cupferron (N-nitrosophenylhydroxylamine), ethyl potassium xanthate and hydroxyquinoline, all at 0·1 mM., with metals at 0·2 mM. concentration. All chelating agents have been previously shown to be toxic to mammalian spermatozoa (White, 1955). An overall analysis of variance has not been undertaken since the variances of the groups in Table 1 were obviously heterogeneous. This is due to the extremely spermicidal nature of some of the metal-chelate mixtures, in which the effects of the metals are so clear cut anyway as to make statistical analysis unnecessary. The significance of conclusions concerning the reverse effect of cobalt have, however, been checked by comparing groups A and B by means of the Z-test.
Cobalt caused a highly significant decrease in the toxicity of nitrosonaphthol for ram (t =19·0, P>0·01), bull (t=11·9, P>0·01), rabbit (t = 9·1, P>0·01) and human spermatozoa (t = 13·6, P> 0·01). It had a similar effect on o-phenanthroline with the ram (t= 11·9, P>0·01),bull (t = 3·0, P>0·01) and rabbit(t = 3·5,P>0·05), and on sodium diethyldithiocarbamate in the case of bull spermatozoa (1 = 6·4, P> 0·01).
There can be little doubt that, for all species, nitrosonaphthol, ethyl potassium xanthate and hydroxyquinoline are more spermicidal in the presence of copper. Similar effects were seen with cupferron using bull, ram and rabbit spermatozoa and with sodium diethyl dithiocarbamate in the case of the ram.
The toxicity of sodium diethyldithiocarbamate, ethyl potassium xanthate and hydroxyquinoline was clearly greater for ram spermatozoa in the presence of zinc. In the case of the last chelating agent a similar effect was also seen with bull sperm.
Cadmium enhanced the activity of ethyl potassium xanthate for all species and of hydroxyquinoline for the bull, ram and rabbit. Similar effects were seen using sodium diethylthiocarbamate with ram and rabbit spermatozoa and nitrosonaphthol with the former species.
The next experiments were designed to test if the increased toxicity of the metal-chelate mixtures was due to an interaction between the two or merely to the additive effect of two spermicidal substances. Factorial experiments were done with two bull, ram and rabbit ejaculates using copper, zinc and cadmium respectively and the chelating agents which gave the maximum effect for the particular metal and species in Table 1. The results are set out in Table 2. Direct factorial analysis is not possible because of the very low variance of the highly spermicidal metal chelate mixture which in many cases gave zero motility scores. By inspection, however, it is obvious that the metals were not themselves toxic and that the greatly increased spermicidal activity of the metal-chelate mixture is due to an interaction between the two—in a statistical sense, at least.
Experiments were then undertaken to see if the toxicity of the chelating agents could be decreased by purification of the diluent. In initial tests it was found that the motility of ram and bull spermatozoa was very much reduced in the purified fructose diluent without the addition of chelating agents. This was probably caused by the thermal decomposition of the ketosugar to toxic products. It was certainly not due to the spermatozoa being deprived of essential trace metals, since the addition of such metals in concentrations normally found in ram seminal plasma (copper, cobalt, manganese 0·05 mg./100 ml.; iron, zinc 0·20 mg./100 ml.) did not improve motility to any extent in the purified diluent. If glucose, which is less readily decomposed by heat, was used instead of fructose then there was no depression of motility on purification of the diluent. These effects are seen in Table 3 which are the results of a factorial experiment involving four bull and four ram ejaculates. The analyses of variance (Table 4) reveal a highly significant diluent/purification interaction for the bull and ram. This is due to the toxicity of the purified fructose diluent and accounts for the significant main effects of the diluents and purification. The significant diluent/metal interaction with the ram is due to the slightly improved motility on the addition of the metals to the glucose diluent. The effect, however, is very small and probably unimportant. Table 5 shows the results of toxicity tests of chelating agents for bull, ram, rabbit and human spermatozoa in purified and unpurified glucose diluent. There is no evidence that the toxicity of the chelating agents is affected by purification.
Nickel combines more readily than cobalt with most chelating agents (Williams, 1953). If cobalt acts by competing with other heavy metals to form a non-toxic complex, then nickel might be expected to be equally effective, provided, of course, the complex formed is also non-toxic. This possibility and also the action of vitamin B12 were examined in the next experiment (Table 6) in which the effect was studied of 0·2 mM. nickel, 40μg./100 ml. vitamin B12 and 0·2 mM. cobalt on the toxicity of 0·1 mM. nitrosonaphthol and 0·1 mM. o-phenanthroline for bull and ram spermatozoa. On testing against group B, cobalt is seen to reduce significantly the toxicity of both o-phenanthroline (t = 18·8 and 8·5 respectively, P>0·01) and nitrosonaphthol (t= 15·9 and 18·9 respectively, P>0·01) for ram and bull sperma-tozoa. Nickel had a similar effect with o-phenanthroline (t = 10·0 and 4·9, P> 0·01) but the nickel-nitrosonaphthol complex is apparently more toxic than the free chelating agent. Vitamin was not effective in reducing the toxicity of either chelating agent.
DISCUSSION
The potentiation of the toxicity of certain chelating agents by copper, zinc and cadmium and its reduction by cobalt seems to be of fairly general occurrence with mammalian spermatozoa. There are, however, some differences in detail between the species. Similar effects have been noted in other biological systems involving chelating agents and metals. Thus oxine is more toxic to fungi (Mason, 1948; Anderson & Swaby, 1951) and certain bacteria in the presence of copper. In the latter instance it has also been possible to reduce toxicity by cobalt (Rubbo, Albert & Gibson, 1950; Albert, Gibson & Rubbo, 1953). MacLeod (1952), too, has found that the bacteriostatic properties of o-phenanthroline were increased by copper and decreased by cobalt, nickel and iron. Mixtures of the heavy metals and other chelating agents used here would seem to merit trial as contraceptives and possibly also as fungicidal and bacteriostatic agents. Invertebrate spermatozoa differ from mammalian spermatozoa in that chelating agents are not toxic to them in the concentrations used here nor is there any potentiation of toxicity by heavy metals. In fact, certain chelating agents may reduce the toxicity of copper and other metals to diluted sea-urchin spermatozoa (Rothschild, 1950;Tyler, 1953; Rothschild&Tyler, 1954).
In the complete absence of copper or iron, oxine has been shown to be completely innocuous to AspergiUis niger (Anderson & Swaby, 1951) and Gram-positive bacteria (Rubbo et al. 1950; Albert et al. 1953). It is tempting to suggest that the spermicidal action of the chelating agent used here might also be normally dependent upon combination with traces of metals such as copper. The reverse effect of cobalt could then be due to it competing with these metals and thus preventing the formation of spermicidal complexes. The fact that nickel, which is higher than cobalt in the Mellor-Maley series (see Albert, 1951), can replace cobalt in the case of o-phenanthroline is consistent with the hypothesis that both metals act in this way. On the other hand, there was no evidence in these experiments that the toxicity of the chelating agent is reduced by purification of the diluent. The possibility cannot be excluded, however, that the chelating agents themselves contained traces of metals. It is unlikely that the free metal content of most mammalian semen would be sufficient to mask the effect of purifying the diluent. The inability of vitamin B^ to replace cobalt in reducing the spermicidal activity of o-phenanthroline and nitrosonaphthol would suggest, as might be expected, that their toxicity is not concerned with the inactivation of this substance.
It should also be borne in mind that the oxidation products of at least some of the chelating agents, e.g. sodium diethyldithiocarbamate and potassium ethyl xanthate, might be the final toxic agents (Keilin & Hartree, 1940). If this is so, then copper and other metals might act by influencing the rate of their formation.
SUMMARY
Studies have been made over a 4 hr. period at room temperature on the effect of 0·1 mM. chelating agents and 0·2 MM. metals on the motility of bull, ram, rabbit and human spermatozoa.
Cobalt decreased the toxicity of (1) nitrosonaphthol for all four species, (2) o-phenanthroline for the first three and (3) sodium diethyldithiocarbamate for the bull.
Copper increased the toxicity of (1) nitrosonaphthol, ethyl potassium xan-thate and hydroxyquinoline for all species, (2) cupferron for the first three and (3) sodium diethyldithiocarbamate for the ram.
Zinc increased the toxicity of (1) sodium diethyldithiocarbamate, ethyl potassium xanthate and hydroxyquinoline for ram spermatozoa and (2) hydroxyquinoline for bull spermatozoa.
Cadmium increased the toxicity of (1) ethyl potassium xanthate for all species, (2) hydroxyquinoline for the first three, (3) sodium diethyldithiocarbamate for the ram and rabbit and (4) nitrosonaphthol for the ram.
Factorial experiments using bull, ram and rabbit spermatozoa indicated that the spermicidal activity of the metal-chelate mixtures were due to an interaction between the two and not merely additive effects of two spermicidal substances.
Purification of the diluent did not influence the toxicity of the chelating agents to bull, ram or rabbit spermatozoa.
Vitamin B12 (40μ.g. %) did not reduce the toxicity of o-phenanthroline or nitrosonaphthol for bull or ram spermatozoa.
Nickel was almost as effective as cobalt in reducing the toxicity of o-phenan-throline for ram and bull spermatozoa.
ACKNOWLEDGEMENTS
The author wishes to acknowledge his indebtedness to Prof. C. W. Emmens for his interest and advice, to Mr A. W. Blackshaw for collecting ram semen, to the Camden Park Estate for bull semen and to the Women’s Hospital, Sydney, for the supply of human semen.