The results presented here demonstrate that the recovery of cryopreserved human spermatozoa can be significantly improved by controlling the concentration gradients experienced by the cells during freezing. The rate of change in solute concentration is clearly a major factor affecting sperm recovery. In those experiments where the only independent variable was the time rate of change of the solute concentration of the unfrozen fraction, cell survival was directly and closely correlated with the percentage of time this time rate of change was decreasing, as shown in Figure 2
. The `controlled concentration' method (6.i) where the rate of change of solute concentration was decreasing for about 90% of the time, shows sperm recovery of 88% and 100% in experiments #1 and #3 respectively, where pooled spermatozoa were used. Spermatozoa from individual patients demonstrated similar behaviour.
When linear cooling was applied the recovery was significantly less than that achieved with `controlled concentration' and was similar to that reported in other studies (Serafini and Marrs 1986
; Ragni et al., 1990; Henry et al., 1993). It is significant to note that using the vapour phase cooling apparatus samples which were not manually nucleated had significantly lower motility than nucleated samples, which is consistent with the report by Crister et al (1987). It has been suggested that different sub-populations of spermatozoa may differ in their freezing sensitivity (Gao et al., 1993; Curry and Watson, 1994) and it appears that the `controlled concentration' protocol may provide successful cryopreservation of sensitive sub-populations.
This investigation provides significant insights into the response to freezing and thawing of spermatozoa. Freeze fracture electron microscopy and freeze substitution extends previous observations made by light cryomicroscopy (Korber et al., 1984; Holt et al., 1992) that spermatozoa and solutes migrate either entirely into the freeze-concentrated matrix or are entrapped near to the interface of the ice and the freeze-concentrated material. In some cases sperm tails may be associated with ice crystals whilst the sperm heads are within the freeze-concentrated matrix. In extreme cases the head of the sperm was situated in one region of the freeze-concentrated matrix with the end of the sperm tail in another zone with the tail bridging through an ice crystal. This may occur because the dimensions of the freeze-concentrated matrix make it difficult to accommodate the sperm head and tail except when lying in the plane of the matrix, or because the tail section of the spermatozoon has surface properties which made it less likely to be excluded from the ice crystal matrix. It is important to note that the relative sperm cell recovery on thawing is not correlated with the structure of the ice crystal network because this structure is essentially fixed by the temperature of ice nucleation in the undercooled straws. All samples nucleated at the same temperature (–5°C) formed a similar initial ice structure upon which all additional ice was subsequently deposited.
Major differences in the eutectic structure between different treatments were made apparent by freeze substitution. Following `controlled concentration' freezing the freeze-concentrated matrix contained large ice crystals, which were absent from the matrix of the other less successful freezing treatments. Further studies are required to determine at what temperature these ice crystals form within the freeze-concentrated matrix, but it is apparent that sperm cells within this matrix are in close association with ice crystals (Figure 4b). It is of interest that gross ice formation within the eutectic is observed in the sample with the highest recovery on thawing. The establishment of spatial gradients within freeze-concentrated materials has been clearly demonstrated by light cryomicroscopy (Korber et al., 1984) and is well documented in metals (e.g. Davies, 1973).
Studies of freeze-substituted lymphocytes (Farrant et al., 1977) have demonstrated that intracellular ice may be visualized and correlates with loss of viability on thawing. Freeze-substitution electron microscopy of frozen sperm has been restricted to an analysis of mouse spermatozoa frozen in the tail of the epididymis (Sherman and Liu, 1982). In this study, intracellular ice crystals were apparent and cellular dehydration was evident, particularly as voids between the acrosome and nuclear membrane and in the midpiece and tail. However as there were no motile cells in any sample following thawing it was not possible to correlate injury with cellular structure. In the current study freezing appeared to have little effect on the cell morphology as revealed by freeze fracture and freeze substitution. No osmotic shrinkage was evident nor was intracellular ice apparent. The possibility exists that cells contain micro-crystalline intracellular ice, beyond the limits of resolution of the ultrastructural techniques employed. Further studies are being undertaken to thermally cycle the samples, which would be expected to increase the dimensions of any intracellular ice present. The only significant structure observed in these studies was the ring of material surrounding sectioned spermatozoa, apparent by electron microscopy. The nature of this material and its formation require further investigation.
It has been demonstrated (Du et al., 1993) that human spermatozoa behave as ideal osmometers, within the range 250 to 1500 mOsm of sodium chloride and that 13% of the isotonic water is osmotically inactive. Using models of the osmotic behaviour of spermatozoa during freezing it has been suggested (Curry et al., 1995) that following a linear cooling at 10°C/min, less than 10% of the cellular water would remain in the cell at –10°C. However, this major loss is not observed here: spermatozoa in Figures 3 and 5 exhibit no osmotic shrinkage. Although the low water content of sperm cells, combined with their flat, non-spherical shape, could allow large changes in cellular water content to cause little modification in the surface area, there is no evidence of any membrane alterations consistent with osmotic dehydration in any of the micrographs examined. However, in this study spermatozoa were frozen in the presence of glycerol, which has a high permeability to human spermatozoa (Gao et al., 1992), and has also been demonstrated to reduce the water permeability of human spermatozoa (Noiles et al., 1992). It is also of potential significance that an aquaporin (AQP7) which mediates water permeability in spermatozoa is also involved in glycerol transport (Ishibashi et al., 1997). The lack of cellular shrinkage would be consistent with the cells effectively being in equilibrium with glycerol at all temperatures during freezing. Similarly the apparent absence of intracellular ice (Figure 5) could also be associated with a high intracellular glycerol concentration.
Whilst experimental treatments give significantly different levels of viability any correlation with cell morphology in the frozen state is lacking, and it is the controlled parameter, namely the rate of change in solute concentration, which is the major factor affecting sperm recovery. It is of interest to speculate precisely why this is the case, and Figure 6a and b shows the estimated form of the local undercooling of the cells during the freezing process for the three distinct cases `controlled concentration' (method 6.i), linear ice fraction (method 6.ii) and standard linear cooling (method 4), whose concentration histories are shown in Figure 1. These estimates of local undercooling have been made by a simple computation of the mass transport across the cell membrane. Figure 6a shows the form of the undercooling when the mass transfer is dominated by water transport, and Figure 6b shows the effect when the glycerol transport dominates.
From Figure 6a
it can be seen that although in the successful `controlled concentration' the undercooling rises to a higher level initially, it remains approximately constant throughout the rest of the freezing whereas the linear cooling rate and linear ice cases rise significantly towards the end of the freezing. This would be expected to increase the likelihood of intracellular ice in these two cases. If instead the mass transport is dominated by the transport of glycerol then it can be seen from Figure 6b
that all cases remain close to equilibrium at the start, but the local undercooling subsequently increases. This increase is almost linear in the `controlled concentration' case but stays low and rises rapidly in the later stages in the other two cases. The values of the local undercooling given in Figure 6a and b
are comparable for similar membrane permeabilities in the two cases. The local undercooling would be comparatively higher in the glycerol case if the permeability were less.
However, it must be noted that intracellular ice is not apparent and it is therefore likely that other physical events determine viability during freezing and thawing. The corresponding estimated changes in mass transport are shown in Figure 7a and b for water movement and glycerol movement respectively. These values are again comparable for similar membrane permeabilities, but would scale relatively for different relative permeabilities. These estimates show the relative form of mass transport where the transport is dominated by either the water or glycerol transfer, and it would be expected that the cells may experience a combination of the two effects over the course of the freezing. Cell viability may be determined by a combination of potential cytotoxic events at high sub-zero temperatures together with restrictions on transport at low temperatures due to low permeability, high viscosity etc. The `controlled concentration' treatment would minimize the time of exposure at high sub-zero temperatures and allow extended periods at lower temperatures to compensate for reduction in transport processes.
Finally, in addition to providing a novel insight to understanding the cellular response to freezing and thawing, the data presented here clearly demonstrate that a simple method exists for the cryopreservation of ejaculated human sperm, which gives better results than vapour freezing or by freezing at a linear cooling. The general applicability of this new approach to cryopreservation is being examined for other cell types including human testicular and epididymal spermatozoa, spermatozoa of different species, and also for embryos and oocytes.
Answers to all your Biology Questions
