The Effect of Temperature on the Survival in vitro of Rabbit Spermatozoa Obtained From the Vas Deferens

To the physiologist the spermatozoon presents many features of interest. As a type of cell it is particularly differentiated. Its functional activities are apparently confined to locomotion and fertilisation. The mechanism of locomotion can be observed in each isolated cell. Spermatozoa can be readily removed from the body without injury to the cellular structure and as these experiments show the individual cells can retain their functional integrity over long periods (up to 7 days). It would appear therefore that the spermatozoon affords suitable material for the investigation of in vitro problems. While considerable research has been carried out on marine spermatozoa, technical difficulties have prevented much use being made of mammalian material, and apart from the actual function of fertilisation little is known regarding the general physiological properties of this important mammalian cell. The majority of experiments have been confined to observations on the duration of motility in various artificial media and toxic solutions. The accurate estimation of duration of motility is attended with considerable difficulty owing to the cessation of motility being variably distributed in the heterogeneous population of a sperm-suspension, and also, since motility may be reversibly inhibited, loss of motility cannot always be taken as a criterion of loss of functional integrity or fertilising capacity. Observations on the duration of motility therefore lack definition and if used to assess fertilising capacity may be entirely misleading. On the other hand, the method of artificial insemination, adopted in these experiments, provides an absolute proof of functional integrity of the cell and strictly quantitative results are obtained in terms of the number of offspring born. The method has, however, a considerable drawback in that many technical difficulties are involved and experimentation is necessarily slow where only a limited number of animals can be kept.

Apart from the theoretical interest in the subject, the experiments were carried out with certain practical aims. Artificial insemination is a well known practice in horse breeding but the usefulness of the method is limited by the short duration of fertilising capacity of the semen outside the body. That the investigation may ultimately prove to be of practical importance may be gathered from the fact that spermatozoa of the rabbit have been sent from Cambridge to Edinburgh through the post with successful impregnation (Walton, 1926). Iwanow (1926) also reports successful fertilisations with spermatozoa of rabbit and guinea-pig seven and eight days after removal from the male. In these experiments the entire testis with epididymis was removed aseptically and kept at a temperature of 1-2⁰ C. Before insemination the spermatozoa were diluted with physiological salt solutions.

The effect of temperature was selected for primary investigation, first because of its importance in biological experimentation, and secondly because it appeared from existing literature to have a marked effect on the survival of motility.

In addition to the experiments described here a parallel series has been carried out by Mr Hammond with spermatozoa obtained from the vagina of the doe after coitus. This method closely resembles that which could be used in practice on the domesticated animals and the whole subject has been approached more from the practical standpoint. The methods differ in certain respects and are therefore published separately, but the general conclusions with regard to the effect of temperature on the spermatozoon are the same and each series is confirmatory of the other (see Hammond, 1929).

The rabbit was used because of its low cost and suitability for artificial insemination (see Hammond and Asdell, 1926 and Walton, 1927). The females were genetically heterogeneous, having been purchased from various sources, but before use each was tested for normal fertility. Since in the course of the experiments many inseminations were negative, the does were periodically tested by normal matings with two or more fertile males to ensure that they were not persistently sterile. These normal matings give a fair indication of the fertility of the stock and indicate that nutrition and general management were adequate. Out of 148 such matings 85-8 per cent, were fertile and the average size of the litter was 7-0. There was little or no seasonal variation, but to equalise any such influence, experiments at different temperatures were distributed as far as possible throughout the year. In order to equalise also the effect of individual variation in fertility, the does for each experiment were selected at random from those available. Each doe was in an equivalent physiological condition, that is to say, on “heat” shortly after the termination of a pregnant or pseudopregnant period. The rabbit does not ovulate spontaneously, so that in order to initiate ovulation the does were mated at least twice with vasectomised males just before insemination.

The males used to provide the spermatozoa were mainly surplus stock from inbred strains of known genetic characteristics and were at least 7 to 9 months old. Some variation in fertilising capacity probably existed, but no definite associations could be traced between high or low fertility and age, breed or seasonal variation.

One of the chief difficulties which beset the investigation of the mammalian spermatozoon is that of obtaining suitable, material. Spermatozoa in large numbers can be obtained in the ejaculate of larger animals, but the collection of the semen offers many difficulties and the spermatozoa are considerably diluted with the complex secretions of the accessory sexual glands. Bacterial contamination is almost unavoidable. In order to obtain a dense, uniform suspension of cells it is necessary to kill (or castrate) the male and remove the contents of the epididymis and vas deferens. Slaughterhouse material should be avoided since post mortem changes occur with great rapidity. The technique adopted in these experiments can be readily explained by reference to Fig. 1 and the explanation which follows.

A male was killed by a blow behind the ears, the abdominal cavity opened and the testes with vas deferens attached removed and laid on a sterile glass plate. Into the free end of the vas was inserted and tied a small capillary glass tube, previously filled with medicinal paraffin (Paraffinum Liquidum B.P.). Over the free end of the capillary was placed a small glass collecting tube, also filled with paraffin. By squeezing the epididymis between finger and thumb and gently drawing a blunt pair of forceps down the vas a small drop of sperm¹ was run down into the bottom of the tube. The tube was then withdrawn and replaced by another, and the process repeated. In this way from 6 to 10 tubes each containing a drop of sperm were obtained from each male. Exposure of the sperm to the air was thus avoided and the sperm in the tube protected from evaporation or rapid gaseous exchange. The tubes were closed with a plug of cotton wool and a drop of paraffin wax. (This did not provide an air-tight joint since it was found that on cooling the wax broke away from the side of the tube, but it probably reduced gaseous exchange.)

The tubes were then stood upright in a small holder and transferred to the thermostat. The apparatus used is illustrated at B (Fig. i). The air chamber in which the holder was placed communicated by an open glass tube with the space under the outer metal cover of the thermos flask. In this way the pressure inside the chamber was equal to that of the atmosphere and allowance made for expansion or contraction within the chamber during adjustment to the temperature of the surrounding water. The water in the flask was adjusted to the required temperature some time before use and the whole flask placed inside an incubator, in an insulated box in a warm room, or in an ice chest according to the temperature required. Thermocouple readings indicated that the sperm in the tubes reached within 0·5° C. of the temperature of the water in the flask in 15-20 minutes. From time to time or when samples were withdrawn the temperature of the water was adjusted by adding a little hot or cold water. The rate at which the temperature of the water changed depended upon circumstances, but by attending to the flask at least twice in 24 hours the variation could be kept within ± 0·5° C. of the required temperature. The temperature variation of the sperm would be slightly less. The temperatures investigated were 45°, 40°, 37°, 30°, 20°, 15°, 10°, 5° and 0° C.

At stated intervals the sperm tubes were removed. The spermatozoa were washed out with about 1 cm. 0·15 Molar NaCl, using the pipette illustrated at C (Fig. 1), and immediately injected into the vagina of the doe. Usually two tubes were removed at each sampling and a doe inseminated with each tube. The technique of insemination has been previously described (Hammond and Asdell, 1926 and Walton, 1927). The instrument is illustrated at D (Fig. 1).

At birth, a month later, the resulting young were removed, killed, and sexed. A few days later the doe would be used in further experiments.

The results are shown in a series of tables (I-XI), one table for each point on the temperature sqale at which determinations of survival were made. In the tables each experiment is represented by one or more horizontal lines. The columns on the left record the number of the experiment and the date on which it was started. The headings of the columns to the right give the times after removal from the male at which inseminations were made. The entries under these headings register the number and sex of the offspring. The number of males is written to the left and separated from the number of females by a colon. The entry “o” indicates a negative result, and “S” that the doe was subsequently proved by normal matings to be permanently sterile and hence could not be regarded as an index of the Fertility of the sperm¹.

An example will make clear the general arrangement of the tables. Table II records determinations made at 40° C. Experiment 106 was started on April 21st, 1928. At 0 hrs., i.e. immediately after removal of the sperm, two does were inseminated, one produced no young but was subsequently proved to be sterile. The other doe produced 4 males and 3 females. Fertile inseminations were made at 2 and 4 hrs. Two negative inseminations were obtained at 6 hrs.

The combined results of several experiments are summarised in the form of a frequency distribution of litter size. It is clear that as the interval between collection and insemination increases fertilising capacity falls off. There is an increase in the proportion of negative results and simultaneously a decrease in the size of the litter. (This is most clearly seen in Tables VIII and IX.) Each of these changes might be taken separately as a criterion of the loss of fertilising capacity, but, as will be explained in the discussion which follows, both are expressions of one and the same thing, namely, a decrease in the probability of fertilisation and it is advantageous to combine the two and take as the criterion of loss of fertilising capacity the ratio of the total number of offspring to the total number of inseminations. This ratio is termed the mean fertility and is given immediately below the frequency distribution. Below this again is the summary of sex-ratios.

The results from all the tables are collected and shown graphically in Fig. 2. The mean fertilities are plotted on a logarithmic scale of time for each temperature determination. From the mean fertilities an “end point” or maximal survival has been estimated and from the “end points” the curves in Fig. 3 have been drawn. It is clear from these curves that temperature has a marked effect on the continuance of fertilising capacity. At body temperature (40° C.) survival is short (about 12 hrs.), but as the temperature falls there is an increasing prolongation of survival until at about 15° C. a maximum of approximately 7 days is reached. Below the maximum, survival falls off again. However, before discussing these results in detail, it is necessary to examine various technical problems and the method by which the curves have been obtained.

The use of medicinal paraffin was arrived at more or less empirically in an endeavour to determine whether the spermatozoa removed from the vas deferens immediately after death would be motile in the undiluted sperm, even when direct exposure to the air was avoided. This proved to be the case, but it does not disprove the hypothesis that the spermatozoa are immotile within the vas deferens and only activate on exposure to gaseous exchange, since the paraffin, although preventing rapid gaseous exchange, is nevertheless permeable to gases and can absorb considerable amounts of carbon dioxide. For a discussion of the problem of the motility of the spermatozoon within the vas the work of Redenz (1926) may be consulted. It is sufficient for our purpose that the method proved exceedingly useful for handling small quantities of sperm and it was found that sperm obtained by this method retained its motility longer than by any other method tried (e.g. hanging drop, dilution with isotonic solutions, etc.). A somewhat similar method was adopted with success by Amantea and Krzyszkowski (1921) in their study of survival of motility. Apparently the spermatozoon is sensitive to dilution, to evaporation, or to rapid gaseous exchange. The rate at which gaseous exchange takes place through the paraffin in the tubes has not been determined, and constitutes one of the uncertainties of the technique.

Another point which arises from the technique is the possible destruction of fertilising capacity at higher temperatures due to the activity of micro-organisms. Although the glass tubes were cleaned in concentrated sulphuric acid and bichromate, rinsed thoroughly in tap water, boiled in several changes of distilled water and finally allowed to dry under cover in a desiccator, no special precautions were adopted to sterilise either the tubes or the paraffin immediately before use. In order to test contamination a number of agar platings were made from the sperm tubes after keeping them for 24 hours at 30° C. The results showed that out of 22 tubes, 16 (73 per cent.) remained absolutely sterile; 5 (23 per cent.) were slightly contaminated (10 colonies or less), but this contamination, as shown by control tubes without sperm, was due to the use of the paraffin or to contamination on the outside of the tubes, and did not affect the sperm. The organisms were mainly moulds. Only 1 (5 per cent.) of the sperm tubes showed serious contamination. Although presumably the sperm would provide a ready nidus for the growth of bacteria, tubes were kept for many days without any sign of bacterial infection when examined microscopically. One can assume then that bacterial decomposition was not a factor in determining loss of fertilising capacity. In the parallel series of experiments with semen recovered from the vagina of the doe, bacterial contamination was, of course, very considerable; nevertheless the survival of the spermatozoa was not markedly reduced, and the reduction may have been due to other differences in technique than bacterial infection.

When the work was started it was not anticipated that so many variables would affect the results. Much of the technique was empirical and adopted simply because in the first few trials it gave successful results. During the course of the experiments it became clear that many improvements, more especially in the standardisation of the technique, might have been adopted, but rather than disturb the constancy of the methods by introducing improvements an effort was made to keep conditions uniform throughout. A glance through the tables and fertility distributions will reveal the extreme variability of the determinations. For example, in Table III, Experiment 75 should be compared with Experiment 78, or in Table VII, Experiment 77 with Experiment 79. These might be taken as evidence of a specific difference in the fertilising capacity of the spermatozoa of the different males, but the number of inseminations which can be made from one male is small and the results themselves so variable that such differences would be difficult to establish with a high degree of probability. However, in a few cases, sperm was removed from the male which was obviously abnormal and which gave completely negative results. Such cases have been eliminated from the records. Without attempting to estimate the relative importance of each variable factor it may be profitable to discuss how they affect the results. Perhaps the most important factor and the one which underlies the interpretation of the results is the variability of the individual spermatozoa within the suspension. We have as yet no quantitative measure of this variability, but examination of a sperm-suspension or a sample of semen, more especially towards the mid-period of active life, shows that it is markedly heterogeneous. Some spermatozoa are fully active, others completely immotile, and there is every gradation between these extremes. Since locomotion is essential for the ascent of the female tract it is reasonable to assume that variability in activity will imply, also, variability in potential fertilising capacity. In order to ascend the tract and fertilise an egg the spermatozoon must possess at the time of insemination at least some residual energy so that the potential fertilising capacity of a sperm-suspension at any time will be expressed by the proportion of spermatozoa which possess more than this minimal energy. In a fresh suspension containing a large number of spermatozoa this proportion will be numerically in excess of the total number required to fertilise all the eggs, but as time progresses the number will fall off until it becomes a limiting factor. From this point onwards successively smaller litters will result and finally sterility, both results depending upon the probability of fertilisation and the number of spermatozoa employed (see Walton, 1926). Provided then that a large but constant number of spermatozoa was used at each insemination under absolutely uniform conditions, fertilising capacity would remain constant over a certain period and then decline somewhat rapidly towards the end. In practice, however, determinations will also be subject to variables other than the heterogeneity of the sperm-suspension. In the experiments under consideration, for example, the number of spermatozoa inseminated was not constant. Every effort was made to keep the volume of the drop of sperm in each tube the same, but this proved difficult and the following estimations show the extent of the variation. The estimates are subject to a standard deviation of about 7 per cent., and were taken at various times throughout the course of the experiments.

Number of spermatozoa in millions, 26·8, 27·6, 17·0, 45·4, 17·2, 29·9, 32·9.

Mean 28·1.

This variation may affect not only the probability of fertilisation by reason of numerical value but also the survival of the spermatozoa within the tube.

The female also contributes considerable variability in the number of eggs shed and possibly also in the conditions of the genital tract favouring or impeding the ascent of the spermatozoon.

The combined effect of all the variables will be to increase the variability of the “end point.” Consequently one finds that instead of the mean fertility remaining constant over a certain period and then declining rapidly, it falls off almost from the start and only gradually approximates to zero. With a greatly increased number of determinations it would be possible to eliminate the variables statistically, but it is doubtful whether the information to be derived would repay the expenditure of time which might otherwise be devoted, to direct experimentation along other lines with improved technique.

In order to obtain from the existing data a numerical value to describe survival as a function of temperature the mean fertilities have been plotted and the “end point” taken where the values approximate to zero, consideration being given to the weight of each mean. This figure gives approximately the maximal survival at each temperature and will therefore be more sensitive to differences than would be the mean (a satisfactory method of calculating the latter has not been devised). The values so obtained are shown in Table X in the second column. Now survival will be inversely proportional to the velocity of the reaction destroying the spermatozoon, so that the relative velocity of destruction can be calculated and is shown in the third column. From these figures the respective curves have been drawn and are shown in Fig. 3.

At 45° C., a temperature just above that of the body, the spermatozoon is almost instantaneously destroyed. One case of fertility is recorded after 15 minutes exposure, but the lag in raising the spermatozoa to this temperature is as prolonged as the exposure itself and it is doubtful if the spermatozoa ever actually reached it. The rapidity with which the spermatozoa are destroyed indicates that there is a critical temperature just above that of the body.

At 40° C. (body temperature) maximal survival is about 13 hours. Hammond and Asdell (1926) determined the maximal survival within the female tract to be about 30 hours, but in order to make the data comparable, it is necessary to add to the first the 10 hours which the spermatozoa spend in the female tracts before ovulation takes place, making 23 hours altogether. The difference, 7 hours, is not very great and may be due to the fact that in the present experiments the spermatozoa were already partly depleted of energy and would take longer to ascend the tract than if the ascent were made by fresh spermatozoa, as in Hammond and Asdell’s experiments.

At 37° C., a temperature a little below that of the scrotum, maximal survival outside the body is not appreciably greater. Hammond and Asdell in the paper mentioned above found that the spermatozoon within the epididymis retained fertilising capacity up to about 40 days. No comparison can therefore be drawn between the survival of the spermatozoon in the epididymis and that outside the body at the same temperature. Other factors than temperature difference must be responsible for the comparative longevity of the spermatozoon in the epididymis.

As the temperature is lowered from that of the body there is an increasing prolongation of survival until a maximum of about 7 days at 15° C. is reached. In other terms, the velocity with which the spermatozoa are destroyed increases with rise of temperature. The curve is only approximately exponential, as is seen from the semi-logarithmic curve in Fig. 1, and it is therefore doubtful whether the velocity of destruction is accelerated by temperature in a similar way to the velocity of many biological reactions which follow approximately the van’t Hoff or Arrhenius equations. The Q₁₀ and μ, the constant of the Arrhenius equation, have been calculated for each temperature interval over this range, and are shown in columns 4 and 5 of Table X. Although of the order of expectancy there is a decided decline in value with rise in temperature. The variability of the data and the approximate value of the “end point” do not allow a more exact comparison.

As to the nature of the destructive reaction very little direct evidence is available. As we have previously mentioned, the spermatozoa are motile below the surface of the paraffin, and the motility of the spermatozoon is reduced at lower temperatures. It might therefore be postulated that destruction was due to the utilisation of energy by the metabolic activity of the spermatozoa. It is, however, not possible to establish a direct association between the activity of the spermatozoa within the tubes and loss of fertilising capacity, because the conditions within the tubes are not uniform. If instead of running the drop of sperm down to the bottom of the tube it is interposed about half-way as illustrated at E (Fig. 1) the activity of the spermatozoa can be observed at the meniscus at each end of the drop. (The centre of the drop is too opaque to see individual activity.) Now at 15° C. it was found that the spermatozoa towards the open end of the tube retained some degree of activity throughout the time of survival of fertilising capacity. On the other hand, the spermatozoa at the closed end of the tube were immobilised in about 36 hours. When later the closed end of the tube was broken off allowing gaseous exchange to take place activity was regained. One cannot then say whether fertilisation was affected by the spermatozoa which were active throughout or by those in which activity had been reversibly inhibited or by those somewhere in the middle of the drop where the degree of inhibition was at an optimum. The success attained by the paraffin technique in comparison with others tried suggests that restriction of gaseous exchange may favour the survival of the spermatozoon. These are points which require further elucidation, and until then it would be premature to put forward any definite theory with regard to the nature of the destructive reaction.

In comparison with these results the data obtained by Gray (1928) on the respiratory activity of echinoderm spermatozoa may be consulted.

Below the optimum, lowering the temperature increases the velocity of destruction. The sharp peak at the optimum denotes a sudden change in the reaction, possibly at a critical temperature. The change is not associated with any obvious change in metabolic activity. The motility of the spermatozoon is reduced at lowered temperatures, but complete inhibition of a fresh suspension does not take place until about + 5⁰ C. It appears therefore that the destructive action of temperatures below 15° C. is due to some indirect effect not necessarily associated with metabolic activity. This effect is the subject of research at present in progress.

According to the genetic theory of sex-determination equal numbers of male and female determining spermatozoa are produced. Assuming that the chances for each are the same, the sex-ratio at conception would be unity. It has, however, been advanced that the sex-ratio at conception is not invariably unity and that the deviation may be due to a differential mortality of the spermatozoa. Since the experiments offered an opportunity of testing any differential mortality in vitro a record of the sex of the offspring was kept.

For each temperature the total survival time has been divided into four periods as nearly equal as the grouping would permit (see Tables I-IX). Inseminations made immediately after removal of the sperm serve as controls. The results are summarised in Table XI. The control inseminations indicate that the normal sexratio does not deviate significantly from unity. Each period has therefore been tested for significant deviations from a male percentage of 50. It is seen that there is a slight but continuous increase in the male percentage. Taking each quarter separately the deviations do not exceed three times the probable error, but the last two quarters taken together show a formally significant deviation and the significance is increased when the second quarter is also added, the deviation being in this case 3·7 times the probable error. It must be remembered, however, that deviations of this order are not infrequently found in experimental material associated with such factors as seasonal variation, differential pre-natal mortality, etc., and the slight correlation between survival in vitro and sex shown above cannot be regarded as significant of differential mortality of the spermatozoa until corrected for partial correlation with these and possibly other unknown factors. In the parallel series of experiments (Hammond) the sex-ratio shows an almost equal deviation in the opposite direction, the percentage of males tending to decrease, so that taken together the data lend no support to the hypothesis of a differential mortality of the sex-determining spermatozoa in vitro.