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

The experiments were performed in conjunction with Dr A. Walton and the results should be compared with a parallel series of experiments reported by him (this Journal, 7, 201).

The vitality of the mammalian sperm outside the body is important from the point of view of artificial insemination. The latter has been up to now used mainly as a cure for sterility rather than as a method in normal breeding for extending the use of the best sires. A knowledge of the factors affecting the vitality of the sperm outside the body is essential to the success of this method of breeding, which may in the future play an important part in the improvement of livestock, for with the development of rapid aeroplane transport long distances can be covered in a short time and the transportation of semen would be far less expensive than the transport of breeding animals.

In addition to this practical aspect of the question the problem is that of the factors affecting the life of the mammalian cell detached from the body and as such forms a convenient study of the optimum conditions for life.

The effect of temperature was selected for study first because it was judged from preliminary experiments to be one of the most important factors affecting the vitality of the sperm outside the body. The rabbit was used as the experimental animal because of cost, the experience already obtained with it, and the ease with which the time of ovulation could be arranged (10 hours after a sterile coitus).

Since in a previous publication (Hammond and Asdell) it was shown that the vitality of the sperm in the male tract of the rabbit was very much greater than that in the female tract−38 days as compared with 30 hours—the present experiment was planned in two parts : I, in which the spermatozoa were obtained from the vagina of the doe after mating, and II, in which the sperm was collected under liquid paraffin direct from the epididymis of the buck. The present paper deals with I; for the results obtained from II the reader is referred to Walton (this Journal, 7, 201). It may be stated here that there does not seem to be an essential difference between the length of the life of the sperm obtained by these two methods, other than that which might be due to differences in the technique involved. For the practical application of this method of reproduction to animal breeding it is necessary to use the first method, but for the scientific investigations of the factors involved the second method is preferable.

The technique involved in this method was simple—such as could be used by the farmer himself on domestic animals, except that as the vagina of the rabbit was so small the doe was killed and the semen collected from it after death, instead of by hand from the vagina as in the cow and mare. It was necessary to kill the doe to avoid all danger of the sample being contaminated with urine, for while with the cow or mare the urethra can be passed by the hand before the tube is inserted to collect this is not possible in the rabbit. Various methods were tried—rubber bags and sponges inserted in the vagina before service and withdrawn afterwards, which were attended with difficulty in getting the mating made at all ; and sucking the fluid from the vagina after service with an inseminator, which as often as not was attended by urination. The method of collection and insemination eventually used was the same as that described in a previous paper (Hammond and Asdell).

The same tubes for containing the semen were used throughout the experiment and the amount of semen collected each time was as far as possible about the same, approximately 10 does being inseminated from each sample; when not sufficient semen was available, fewer does were used, so that approximately the same bulk of semen was used for each doe. The experiments were run throughout the year, although, as is well known, the results from normal breeding are not so good in the autumn and winter as they are in the spring, but as far as possible the different temperatures were spread over the seasons. In evaluating each experiment, however, the time of year at which it is made plays a part and for this reason the results are given in detail in Appendix A.

The glass tubes (with paraffined corks) in which the semen was kept held about 10 c.c. of fluid, and they were usually half to three-quarters full. In keeping at o° C. the corks were covered with grease paper (to prevent contamination with water when the tube was opened) and the tubes were buried in crushed ice in a wide necked thermos flask, while the flask was placed in an ice chest and surrounded with ice. The experiments at 5°, 10°, and 15° C. were, as far as possible, performed when the normal air temperature was at these levels; the thermos was placed in the opened ice chest and the temperature of this was reduced to the required degree by ice or water, or in winter increased by warming the room in which it was standing. When the temperature was constant the semen was collected and the tube containing it was left in the ice chest to cool down to the temperature required and then inserted into the thermos at that temperature and the ice chest closed. High-low recording thermometers were used, one standing in the ice chest and another in the room ; by opening the doors and getting draughts or by heating the room the temperature of the room itself could be regulated usually within a few degrees of that required, while that of the ice chest varied still less and it was presumed that the thermos temperature itself varied very little. The most the temperature ever varied in the ice chest was 2°C. each way; usually it was 1° each way, up during the day and down during the night. At 35°C. the tubes were inserted within a corked boiling tube which was immersed in a water-bath kept at this temperature by a thermostat. On standing in tubes the semen separates out into a white flocculent precipitate below, containing most of the sperms, and a clear serum above. On insemination a portion of each was drawn off without shaking the tube.

Before inseminating the does were mated two or three times with sterile vasectomised bucks.

The details of all the experiments made are given in Appendix A. Each sample is denoted by a separate line starting with the date, while the figures show the size of the litter obtained from each doe inseminated. The S denotes that the result, which was negative, has had to be neglected owing to the doe being sterile from adherent tubes. Not until towards the end of the series were does tested for fertility after each insemination ; in the earlier experiments they were usually put through three experiments before being tested by normal matings and the results of those that proved to be sterile (which was confirmed on autopsy) were marked as 5 back to the time the last litter was produced. These results have been neglected in making up the tables of results; the cause of the sterility will be referred to again below. The mark ϕ signifies that the animal made a nest of fur at the 16–20th day (indicating that a period of pseudopregnancy had occurred), showing that the lack of fertility was not due to failure to ovulate. As will be seen from Appendix A, the method of testing each sample over a 10-hour period at intervals of 2 hours and repeating the process on other samples at similar times tended to eliminate the chance effects of differences in samples, a bad sample being spread over several hours and standing a chance of being equalised by a specially good one spread out over the same time.

In order to eliminate differences in samples as far as possible the collecting doe was generally served by three or four different bucks. There are individual differences between bucks; some produce many sperm and others only a few and so this method was adopted in order to get average results. Statistically the mixed semen of four bucks should be less variable than that of one. The males were not used for about 4–7 days before the experiment so as to avoid any differences that might be due to lowered numbers or vitality in successive ejaculates (see Lewis, Stigler and a recent paper by Young which refers to the literature). Throughout every effort was made to keep all the factors constant or average except the one being tested : when quantitative results are required this is essential to success.

After each doe was inseminated a drop of the fluid remaining in the inseminator was examined under the microscope and the percentage of the sperm motile was estimated. The sperm of the rabbit is slowly motile at low temperatures (10° C.); warming to body temperature increases the rate of motility, but does not increase either it or the percentage motile to anything like the same extent as it does in the sperm of man or the horse, in which it is necessary to approach body temperature before the percentage motile can be judged.

Fertility as measured by the number of does which produced young. Table I shows the number and percentage of the does which produced litters from insemination after the sperm had been kept outside the body at different temperatures. At 35° C. there is a sharp fall from 40 per cent, in the first 6 hours to 8 per cent, in the third 6-hour period, no does producing young after 14 hours. At 10° C. fertility remains high up to about 76 hours and then proceeds to fall off, no doe producing young after 96 hours. At o° C. fertility falls fairly rapidly from 59 per cent, during the first 6 hours to 11 per cent, during the third 6-hour period, no doe producing young after 16 hours. The few tests made at the intermediate temperatures seem to indicate that the temperature curve of fertility approximates to a line joining these three points, since at 15° C. a high degree of fertility was attained at 60 hours while at 5°C. a high degree of fertility was obtained up to 18 hours, the few negative results obtained beyond this time not being significant since they were all from one sample (see Appendix A). It will be noted that at 10° C. there is a slight depression in fertility between 48 and 54 hours while the percentage increases again later up to 63 per cent, at 74–78 hours; reference to Appendix A will show that this is probably due to the time of year these times were being tested—September and October—in which normal matings often give low results. It may be concluded that there is an optimum temperature at or about 10° C. at which young are produced up to 96 hours, while at the higher and lower temperatures tested fertility falls off rapidly; at 35°C. young are only produced up to 14 hours while at o° C. only up to 16 hours.

The variations in the different samples are shown in Table II ; the sample is shown as fertile if any young were produced from it. As mentioned above, every effort was made to obtain uniform samples by using the mixed semen of different bucks and the results given in this table show that this was fairly successful. The one sample at 35° C. which gave no young at 2–6 hours consisted of the jelly part of the semen (secretions of the uterus masculinus) and might be expected to contain only a few sperm. Variation in the keeping qualities of the different samples is only shown towards the end of the life of the sperm; thus at 10° C. the first sample which gave no young occurred at 56–60 hours, the 15 other samples tested up to this time all giving positive results. On the other hand, one sample gave young after 98–102 hours, although 7 other samples from 86 hours onwards had failed to produce young. At the other temperatures (35° and o° C.) the rate of fall off in fertility is too sharp to show up much difference between samples. Even with all the precautions taken to prevent differences between samples it will be seen that some do exist; since the individual male used has been eliminated as a variable such differences as exist are probably due to other factors than temperature affecting the life of the sperm outside the body—what these are further work must show, but it is probable that the amount of exposure to air is an important one.

The necessity for eliminating the differences between individual males in an experiment to determine the effect of temperature is shown by a consideration of Table III. In rabbits agouti is a Mendelian dominant to black and as pure lines of these strains were being used the opportunity was taken of testing differences between bucks without upsetting the plan of the experiment. Recessive coloured (black) does were inseminated from a mixed sample obtained from one or two black and one or two agouti bucks. In Mendelian work where equal numbers of two sorts of sperm are produced by the same buck equal numbers of the two sorts of young are given and the totals at the bottom of Table III show that this is so when two strains are compared as a whole. When, however, only one buck of each strain is being used (Samples 4 to 8 in Table III) there is a great preponderance of the young produced from one only. In Samples 4 to 7 the bucks, as a result of these experiments, were carefully examined and two were found to have become sterile, although a few motile sperm were present in their semen; they had previously been fertile and had bred litters of young. Even where two bucks of each sort were compared it frequently happened that the majority of the young produced were predominantly from one sort only (see Samples 16, 17 and 18), this depending on the individuals rather than the strain, for as the number of bucks used on each side is increased so apparently is the chance of getting equal numbers of young of both sorts (Sample 15), as would be expected from eliminating individual differences. This problem of differences in male fertility has been under investigation for some time in another set of experiments which it is hoped will be published later. Apart from these individual differences there does not seem to be any variation between the two strains in the effect of temperature or time of survival ; the only case in which there is any suggestion of this is in Sample 11, Table III, in which the agouti sperm seemed to survive better, but this is countered by Sample 2. The numbers are too small, however, to draw definite conclusions. Differences in the vitality of semen from a normal and a partially fertile boar have recently been noted by Nordby.

The does used came from strains differing in fertility but the distribution of the does of the different strains is fairly uniform over the whole series so that, as there are not sufficient numbers to warrant a division between the strains, the results are given without subdivision in Table IV. It will be seen from this table that just as fertility as measured by the percentage of does fertile falls off with the time the sperm is kept, so the average size of the litter decreases ; thus at 10° C. during the first day the average litter size is 5·6 which falls to 4·3 and 4·1 in the second and third days respectively, while on the fourth day in spite of one large litter it is reduced to 3·9. Similarly, during the rapid fall off in fertility at 35°C. and o° C., the average litter size decreases appreciably from the first to third 6-hour period. In normal matings with fertile bucks there are many more sperm at the tops of the tubes than are required to fertilise all the eggs of any female, but in the semen which is kept outside the body and which is approaching the end of its fertility it is only reasonable to suppose that the number of sperm, capable of fertilisation, which reach the tubes often falls below the number of ova to be fertilised, as has been shown experimentally by Walton. Observations on the percentage motility of the sperm support this suggestion.

When small litters, from does which would normally have produced large litters, are obtained in this way the weight of the individual young at birth is greatly increased. Thus the largest of the litters of one weighed 104 grs. while the largest of a litter of twelve weighed 58 grs. Table V shows the distribution and average weights of the young from the litters of different sizes obtained during these experiments, the average weight of the litters of one being 86 grs., exactly double that of the litter of twelve, which was 43 grs. The variation in weight at each litter size is due to other factors such as strain, age of the doe, etc., but as these factors and their inter-relationships are being investigated in other series of experiments and in small litters obtained by other means full discussion of the results will be reserved for a future communication. It may be stated, however, that the young of the small litters are very much better developed at birth than those of large litters, having quite long hairs over the surface of the body.

The duration of pregnancy is also longer in the case of small litters obtained in this way (see Table VI) ; while a small part of the extra weight and development of the individuals of small litters is due to the prolongation of the pregnancy, the greater part is due to other causes. The variation in average duration of pregnancy for litters of different sizes is shown in Table VI ; while other factors such as strain, etc. cause much variation, the average duration for litters of one is 33·6 days and for litters of eleven 31·7 days; or put in another way the average litter size for those having a pregnancy of 35 days is 2·0 while for those having a pregnancy of 31 days it is 5·2. Since the factors which influence the duration of pregnancy are also being investigated in other series of experiments discussion of the cause will be deferred until these have been completed.

The observations made on the motility of the sperm are summarised in Table VII The numbers denote the samples and the average motility of each sample is recorded for each 6-hour period over which it was examined: thus at 35° C. sample 6 had over 20 per cent, motile at 8–12 hours, over I per cent, at 14-18 hours and none motile at 20–24 hours. The average motility at 35° C. fell off rapidly from 52 per cent, motile in the first 6 hours to 9 and o per cent, in the third and fourth 6-hour periods respectively, whereas at o° C. there appeared to be an initial depressing effect of the low temperature, so that during the first 6 hours there were only 42 per cent, motile ; but the motility after this did not fall off so rapidly, still being 15 per cent, in the third and fourth 6-hour periods. At 10° C. the percentage motile during the first 24 hours was high (80 per cent.) and fell gradually until at the fourth and fifth days it was 29 per cent, and 5 per cent, respectively. Since the fall off in motility at this temperature was slow and each sample was examined for from 2 to 16 hours after the tube was first opened (see Appendix A) a comparison of the results would show whether there was any decrease in motility owing to the effect of opening the tube or whether the motility was preserved better by keeping the tubes closed until required. Excluding results towards the end of the time (after about 82 hours) as being towards the end of the life of the sperm and, like those at the other temperatures, not suitable for this purpose because the other limiting factors for life were predominant, it would appear that opening the tube during the first 2 days led to a decrease in motility whereas opening it on the third to fourth days had no effect and even sometimes appeared to increase the motility slightly. In future work it may possibly be found that the sperm, like the abortion bacillus, survives best under partial anaerobic conditions—at the level of tissue respiration rather than that of external respiratory exchange ; for body fluids such as milk contain a high concentration of CO₂ when they are first drawn off but lose it on exposure to air. A comparison of the observations on motility with those of the percentage of does fertile and the size of litter at the different temperatures is shown in Fig. 1 ; broadly speaking the three characters move together at each temperature, so that the percentage motile is a guide to the percentage fertility and the size of the litter. When examined in detail, however, it will be seen that the percentage motile tends to over-estimate the percentage fertility at the high temperatures (35° C.) and to under-estimate it at the low temperatures (0° C.) in the early stages before fertility declines sharply: in the later stages when fertility is declining rapidly, however, the reverse seems to occur. It was noted that healthy sperm on standing under the coverslip for a short time became flocculated, with the heads together, usually round leucocytes, but these aggregations were easily broken up by moving the coverslip.

Since it is generally believed that the sperm are dimorphic, due to differences in sex chromosome content (although in the rabbit but little difference can be seen), it was thought possible that one or other sort might have a longer survival and so give rise to abnormal sex-ratios. The results are shown in Table VIII ; it will be seen that no striking and definite change in the sex-ratio was produced, although there seemed to be a slight preponderance of females towards the end of the period of life ; since the numbers are small, however, and the sex-ratio is therefore liable to large variations with each 6-hour period, the results considered as a whole appear negative. Moreover, as in the parallel series of experiments (see Walton, this Journal, 7, 201), the sex-ratio shows a deviation in the opposite direction, this method appears to offer no solution of the problem of sex control. Riddle and Behre have also shown that the staleness of the spermatozoa does not appreciably affect the sex-ratio in doves.

The rabbit is peculiarly susceptible to infection of the tops of the Fallopian tubes, this leading to adhesions to the ovaries and à blocking of the passage of the tubes, so causing sterility. Cases of this sort had been observed before insemination experiments were started (see Hammond and Marshall), but during the course of these experiments a very large number of infections were encountered (see Appendix A, marked S). In all 83 does went sterile through adherent tubes and, owing to the practice in the earlier experiments of making about 3 inseminations before testing the does for fertility by a normal mating, 151 results out of a total of 567 does inseminated had to be rejected for this cause. In the later experiments (after 14. v. 27) the does were tested by a normal mating after each insemination and the times of infection could be determined. These were as follows:

At 35° C. 9 out of 36 does became sterile after insemination at 10–20 hours after collection.

At 15° C. 1 out of 10 does became sterile after insemination at 52–60 hours after collection.

At 10° C. 2 out of 6 does became sterile after insemination at 94–102 hours after collection.

At o° C. 8 out of 17 does became sterile after insemination at 14–22 hours after collection.

Thus it would appear that the temperature and the time during which the sperm is kept have no differential effect on the infections. The bacterial content of the semen kept at 10° C. for 94–102 hours was often very high, much higher than it was at o° C. for 14–22 hours, and yet the percentage of infections was not very different. A number of infections has been observed also when the semen is immediately transferred from the collecting doe to the animal inseminated (see Hammond and Asdell).

Although only a small amount of data is available the percentage of infections appeared to be slightly greater where the percentage of sperm motile was larger, as the following results show ; but these are too small to warrant definite conclusions.

There were cases at each temperature where no animals inseminated from a particular sample were infected, among which may be mentioned one sample at 35° C. kept for 12–20 hours (19. vi. 27) and one sample at 10° C. kept for 66–74 hours (21. i. 27).

In the same stock of rabbits, mated to the same bucks as used to collect semen for these experiments, only very occasional cases of adherent tubes have occurred as a result of normal matings—only 2 or 3 out of several thousand matings.

As the does killed to obtain the semen were very frequently those which had been inseminated before it may have been that a pathogenic organism brought in by one such purchased doe has been the cause, but if so it has not been passed on by the male. In the cases in the parallel series where insemination has been made from semen obtained directly from the male only occasional cases of sterility due to this cause have been found.

Lush also found when inseminating rabbits with centrifuged sperm that many cases of sterile animals resulted, but we are not aware of any such cases among farm animals that have been inseminated.

Since references to most of the literature are given in the papers quoted and in the paper by Walton no detailed account of it will be given here. While the effects of various factors affecting the motility of the sperm have in a few cases been investigated systematically (see Yamane and Kato—on the optimum pH), most of the work on the duration of fertilising power has been in the nature of trial experiments with various methods of preserving the life of the sperm. Yamane and Kato obtained young after rabbit sperm had been kept for 24 hours in phosphate-buffered dextrose solution, and Iwanow, who kept the intact epididymis at 1–2° C., obtained young after 7 days in rabbits and 8 days in guinea-pigs.

Comparing the results given here with those obtained in a parallel series where the sperm were collected from the epididymis of the male and kept under liquid paraffin (see Walton, this Journal, 7, 201), it will be seen that there is an essential similarity as regards the optimum temperature for their life. There are differences, however, in the shapes of the curves which may be important in indicating the other factors which may affect the life of the sperm. The differences are briefly as follows, the table showing the time at which the last litter was produced.

While the time of survival is practically the same at 35° C. the relative time of survival is increased as the temperature is lowered in the paraffin series (I) as compared with the series exposed to some air in tightly-corked tubes (II). Whether the presence of air, the dilution with the secretions of the glands of the male and female tracts, or other factors are the cause future work must decide. It seems not unlikely that the factors which are responsible for the sudden fall in the temperature curve at and below 10° C. in both series are similar to those which cause the differences between the two series. Since for practical purposes it is easier to work at o° C. than to control temperature above this level, it is important to determine the cause for the fall in the curve below 10° C.

Whether these temperature effects on the length of life of the sperm are related in any way to the critical temperature of the animal as a whole, as determined by calorimetric work, must await further investigation on the comparison of species having different critical body temperatures. It may be that the metabolism of the single mammalian cell is affected in the same way as the body as a whole by the temperature of the environment. It would appear not unlikely that the problem of keeping the sperm outside the body is closely allied to that of the hibernating animal in which either the metabolism must be slowed down by one means or another or else nutrients supplied to balance loss.

The general shape of the temperature curve (see Walton, this Journal, 7, 201) with its peak lying between 10–15° C. may be of use in practice in that when the air temperature is above this level the sperm may be dispatched in a thermos at the lower temperature, so that fluctuations if any in transit may be under optimum conditions instead of quickly passing to unfavourable conditions for life.

Since X-rays have been found to be more effective in producing transmittible gene mutations in the mature spermatozoa than in the immature male germ cells (Harris), this technique of artificial insemination may be of use in attempting to produce mutations in the more valuable farm animals such as Muller has produced in Drosophila.

The experiments were performed at the Field Laboratories, Milton Road, Cambridge, in connection with the Institute of Animal Nutrition and the expenses were largely defrayed out of a grant made to the Institute by the Ministry of Agriculture and Fisheries.

To Mr S. Tadman and Mr V. Swann the writer’s thanks are due for the care and feeding of the animals during the course of the experiments.