Cryopreservation: Advanced

The only method of stable and long-term (practically infinite) preservation and storage of any perishable biological materials, particularly cells, is to keep them in the glassy (vitreous) state. This was apparent to Father Luyet when he titled his pioneering work “The vitrificationof organic colloids and of protoplasm” and “Revival of frog's spermatozoa vitrified in liquid air"(1, 2). He and other “pioneers of the cryobiological frontiers” including Lovelock, Meryman, Mazur, Polge, Smith, Levitt, Farrant, and Willadsen, clearly understood some 40-70 years ago that only a glassy state would insure stable and non-lethal preservation of cells. With time, we saw the development of a variety of biopreservation methods, such as slow-cooling (which is just a way of achieving glassy state inside the cell) (3). It was Luyet’s work, which would make cryopreservation a science. From the outset, he recognized that ice damage must be avoided and vitrification could be a method for long-term preservation of cell viability (2).

In order to understand rapid-cooling or modern vitrification techniques, let us compare them to the slow-cooling method. During slow-cooling, using a programmable freezer, embryos are exposed to relatively low concentrations of cryoprotectants (1.5M PrOH and usually some sucrose, around 0.2M) equilibrated for 10-25 min at room temperature, loaded into a straw or vial, sealed and placed into a controlled-rate freezer. Ice formation is initially induced extracellularly by seeding at a temperature whereby ice can perpetuate (around -5.5°C or lower) and, as a result of the solute gradient created, freezable water flows out of the cells, minimizing the chance of intracellular ice formation during cooling. As the temperature is gradually lowered, the concentration of cryoprotectant in the liquid phase, which includes the intracellular fluid, increases correspondingly until a level is reached at which additional formation and growth of ice crystals, although possible, are unlikely, even if the temperature drops further (4). Rather, the liquid phase turns into a glassy substance that solidifies without further crystal formation as the temperature continues to decrease. The unfrozen liquid phase remaining within the cells when they are plunged into LN2 should ideally consist of this glassy substance with all the original cell solutes remaining in solution (4). This suggests that when we slow-cool cells using a penetrating cryoprotectant such as PrOH, and standard slow-cooling protocols, we are actually vitrifying the cells. Indeed, when we slow-cool human embryos, typical survival rates range between 80% and 100%, for many IVF centers. These survival rates would not be possible, at least according to Mazur and company, if intracellular ice formation were occurring (5). This correlates well with the theory that slow-cooling is vitrification. Of course this does not mean that IIF does not or cannot occur, it simply suggests that in conventional embryo freezing protocols, IIF is not a major source of cell damage. However, if water eflux is inhibited and does not occur in equilibrium, as suggested, and despite a slow cooling rate, IIF could pose a real problem (6).

Intracellular Ice Formation and Osmotic Effects

Slow cooling protocols derived and modified from mouse and bovine embryo freezing work in the early 1970’s do not work well for the storage of unfertilized oocytes, but have been able to produce a low percentage of offspring. Oocytes can therefore be frozen effectively, but only if the stresses imposed on them are reduced or removed. New information is needed by using very different approaches within the same context of dehydrating the cell, binding the remaining water with a cryoprotectant, and cooling to specific temperature(s) before storage in liquid nitrogen. For success, we will need to identify problem areas and then modify our approach to work around them.

However, there are a number of unknowns. When a cell lyses after thawing, the typical logic is that intracellular ice formation (IIF) and/or osmotic effects were responsible. However, commonly this analysis is not based on experimentation, but on the seemingly logical, potential problems of IIF and osmotic effects. Hence, the freezing protocol is either modified, or abandoned because "it does not work". In 1990 and 1991 Toner et al., specifically investigated IIF in mouse oocytes cooled in PBS without cryoprotectants and confirmed two distinct mechanisms of IIF: the first catalyzed by extracellular ice and the second heterogeneously by intracellular structures (Toner et al., 1990; 1991). They discuss these mechanisms in the presence of cryoprotectants and conclude that both seeding (mechanism I) and heterogeneous nucleation (mechanism II) are rendered ineffective by the addition of cryoprotectants. Leibo, McGrath, and Cravalho, in 1975, and again in 1977 and 1978 show that mouse oocyte survival is inversely related to IIF. Their graph shows that the curve denoting IIF crosses the X axis and becomes zero (0% IIF) at and below a cooling rate of 1°C/min. This is substantially faster than the current protocol standard rate of 0.3°C/min used for freezing mammalian embryos and oocytes, and accordingly IIF would not occur when a slower rate is used. In another study, Ashwood-Smith et al. 1998 looked at IIF in mouse oocytes using a cryomicroscope, reconfirming previous literature concluding that if IIF was present (occurred at –14°C using a solution of PBS, 10% MeOH & 10% dextran), the cells would lyse upon thaw due to a number of mechanical forces, including internal ice expansion rupturing membranes. They also say that cells could have been pierced by extracellular ice crystals, but eariler Luyet and Gehenio 1940 denied the existence of this mechanistic failure, stating “experience has shown that cells/tissues can tolerate the presence of extracellular ice and despite morphological appearances exerts no mechanical pressure on the cells and certainly does not puncture them”. Ashwood-Smith et al. 1998, argued that mammalian eggs are much larger than the red blood cells used in Luyet’s paper, and physical forces could still play a significantly damaging role. But if these forces were lethal it would not be possible to obtain near 100% survival of mouse embryos using a variety of cryoprotectants and protocols currently employed.

Therefore we suggest that other stresses cause cell death; although mechanical forces would still contribute to the overall stress upon the cell, especially re-expansion or sheering forces upon rewarming (causing zona cracking for instance). The aforementioned studies collectively indicate that IIF is not necessarily a serious problem in modern protocols where penetrating cryoprotectants, slow-cooling, and appropriate rewarming strategies are used.

The other potential problem of “solution effects” involves the change in a solution or cytoplasm that result from dehydration, increased solute concentration, pH changes, and precipitation of solutes (Mazur 1965, Mazur et al., 1984). Half a century ago, Lovelock (1953), showed that the increasing extracellular solute concentration that occurs during cooling when water freezes out as ice, was responsible for red blood cell lysis (>0.8M NaCl) and that the addition of a cryoprotectant (glycerol) reduced the amount of ice formed at any temperature; thereby effectively reducing the concentration of electrolytes produced. It has indeed been known for a long time that “solution effects” must be reduced to such a degree that cells can withstand cryopreservation. However, several more recent studies have reported higher tolerance levels to osmotic stress. Agca et al. (2000)exposed bovine oocytes to PBS with increasing concentrations of sodium chloride to increase the osmolarity to supra-physiological concentrations (up to 4800 mOsm). They found that oocytes exposed to 2400mOsm or lower developed to the blastocyst stage, albeit at slightly reduced rates, as compared to untreated controls. Van Os and Zeilmaker (1986) exposed mouse zygotes to solutions up to 3100 mOsm (added NaCl) apparently without detriment to blastocyst formation.

Others have exposed human or mouse oocytes and embryos to varying concentrations of cryoprotectants or sugars without cooling and found cells have a considerable tolerance to heightened osmotic conditions(Hotamisligil et al., 1996; McWilliams et al., 1995; Oda et al., 1992). When Toner et al. (1993) cooled mouse zygotes to –40°C in moderate to high concentrations of NaCl (300-2400mOsm) without cryoprotectants all the embryos lysed after rewarming. However, if the osmolarity of the PBS solution was increased (up to 2400mOsm) using another salt (choline chloride) the majority of the cells remained intact despite the absence of cryoprotectants. Furthermore these zygotes were plunged directly into the highly concentrated solutions and not slowly dehydrated over a range of increasing osmotic concentrations, yet they survived.

This study showed that cellular demise from osmotic stress could be due to the type of stressor and not necessarily the ability of the cell to handle osmotic shock. Indeed, during the normal course of cryopreservation, cells (mouse embryos for example) are exposed to very high osmolarities yet >90% survive. If osmotic stress were as lethal as originally suggested, the majority of the cells would die; yet this does not happen. Even when we consider human oocyte freezing, around half of the cells are intact after thawing and around 70% of the blastomeres in human embryos survive freezing.

Therefore, lethal IIF and osmotic effects may only play a small part in cellular demise during cryopreservation than generally thought, even though they can be lethal if the cells are not treated properly. There are likely other problems that may have been concealed by IIF and osmotic effects.


The basic principles of removing water from the cell, adding a cryoprotectant to bind the remaining water to prevent ice crystal formation, and cooling to liquid nitrogen temperatures is the same as it has been in the past. Since one could vitrify an embryo by slow-cooling early studies calculated the intracellular cryoprotectant concentration that occurred after cooling to minus -30°C and exposed embryos to these concentrated solutions at room temperature. A variety of cryoprotectants including DMSO, proplyene glycol, glycerol, etc. were used at various concentrations. Early studies in the cow, pig, sheep, rabbit, and mouse led to a better understanding of what was possible(19, 25, 28-38).

Despite these successes, most of the solutions being used were more or less toxic to the cells and although offspring could be produced, the technique overall, did not work very well. The main reason was that for equilibrium vitrification to occur the cells needed to be exposed to high cryoprotectant concentrations at or above room temperature, and for a reasonable amount of time to dehydrate the cells and load the cell with cryoprotectant. Dimethylsulfoxide (DMSO) was now being used as the primary or one of the primary cryoprotectants because of its rapid passage through cell membranes, despite its toxic potential (39-42). The elevated cryoprotectant concentrations (6-8 Molar), temperature (23°C to 39°C), and duration, all increase toxicity. So the challenge was to reduce the concentration enough to allow for survival without allowing intracellular ice to form and kill the cell. It is likely that the cells did vitrify, however, many died or were dead upon rewarming.

Modifications in media composition and reduction of cryoprotectant concentration led to some improvement of survival but the big breakthrough came when very rapid cooling rates were used (25). Vajta found that by using minute volumes of 1ul or less along with direct submersion into LN2, using an open container allowed for extremely rapid cooling rates on the order of >10,000°C/min. They found that despite using high cryoprotectant concentrations and relatively high temperatures of 23°C or higher, the extremely rapid cooling rate did not allow enough time for ice to form, even if it could. Longer times in the solutions and slower cooling rates proved detrimental to survival, possibly because of cytotoxicity (42).

From 1998 until the present there have been hundreds of publications on human blastocyst, embryo and oocyte vitrification. Despite posing a greater risk from the potential toxicity of the highly concentrated cryoprotectants and the relatively high exposure temperature, rapid-cooling, in most instances, has met with greater success (42, 43). The cryoprotectant solutions most often used consist of an equilibration solution of 7.5% DMSO and 7.5% Ethylene Glycol and a final vitrification solution of twice that concentration (44). To combat the potentially cytotoxic effects, exposure to the final vitrification solution is usually limited to around 45-90 seconds or less before plunging (45-50). Also, as mentioned above, in order to obtain a rapid cooling rate in the order of >10,000°C/min it is necessary to use very small volumes of media, usually 1ul or less and direct exposure to LN2. Since most clinics used 0.25cc plastic straws for slow-cooling, new storage devices that could hold very small volumes of media and that allowed for direct contact with LN2 needed to be developed. Alternative storage containers including electron microscope grids, ultra small nylon loops, hemi-straws, and open pulled straws were tested (25, 51-59). In addition to these, over 15 new devices were developed including the hemi-straw, closed-pulled straw, cryo-top, cryo-tip, cryo-pette, cryo-loc, cryo-leaf, rapid-i, HSV straw, etc. These devices generally fall into one of two categories; 1) micro-sized straws that could be heat-sealed, or 2) a thin flat plastic blade with a handle. With the cryo-top device, for example, the cell(s) are placed on the end of the device and excess media is removed, leaving the cell(s) covered in a very thin film of media, before plunging directly into liquid nitrogen (44,60). This allows for an extremely rapid cooling rate of over 20,000°C/min as shown in Table 1.

Similar vitrification devices to those in Table 1 allow cooling rates of >15,000°C/min and have resulted in high survival rates (44, 47, 49, 52, 58, 61-67). In fact, the combination of cryoprotectants used in conjunction with very rapid cooling rates has allowed for these results, whereas slower cooling rates have yielded poor survival rates (68, 69). All of these devices, despite claims of "novel method for vitrification", all are based upon the same principle of ultra rapid cooling and warming rates. Some investigators have even gone as far as to use LN2 in a vacuum to create a "slush", increasing the cooling rate even more (65, 70, 71). All of these devices/methods are simply a modification of what Vajta described in 1998.

Potential Problems:

Despite the increase in survival and pregnancy rates, and the relative abundance of recent reports on vitrification, there are numerous potential shortcomings associated with these protocols that have prevented its widespread application and acceptance (72). To start with, viral contamination from direct contact to liquid nitrogen is a concern despite reports indicating that no such contamination has occurred to date (73-75). Other devices that are closed can be more appealing to use as they avoid direct contact with LN2. The cryo-tip, cryo-pette, closed pulled straw, and micro-secure are examples of closed devices (44, 76). Bielanski and Vajta discussed current concerns about the safety of using open containers for vitrification and reviewed the confirmed and theoretical hazards of these procedures in their 2009 manuscript (77). They also suggest methods to avoid these dangers when using current vitrification techniques/devices. Of primary importance is to use a vitrification method that is successful in your particular practice. Secondarily would be to use a secure closed system that would avoid potential contamination problems and that would follow good tissue practice regulations so that foreseeable changes in laboratory regulations would not prevent you from continuing to use that system.

Another drawback is that the technique of placing cells into a highly concentrated vitrification solution, loading them onto a minute container, and plunging into liquid nitrogen, all in less than 45-90 seconds remains technically challenging; and more importantly, leaves little or no room for error. In a recent paper, (67) reported an impressive >98% oocyte survival rate after thawing, however, they also mentioned that it took their lab over 5 months training to obtain such rates and that operator skill was crucial to guarantee the proficiency of the procedure. Because results are often based upon the technical skill of the person doing the vitrification procedure adaptability and consistency can be poor. Failed experiments or studies with low success rates are rarely, if ever published, thus giving a false impression of overall success rates. Despite these problems, vitrification has led to a marked improvement in blastocyst survival and higher pregnancy rates for many clinics (22, 78). Edgar and Gook recently published an extensive review of the literature comparing overall success rates with slow-cooling vs. rapid-cooling/vitrification (78).

Large-Volume Vitrification (ICE Vitrification):

There is another method to obtain a vitrified cell. Large volume vitrification is a somewhat different technique based on the basic cryopreservation principles described above. The use of a large container, a 0.25cc straw for example, that is sterile, that can be loaded and sealed easily in a timely manner, and that uses a significantly slower cooling rate of <2000°C/min, contradicts the idea that a faster cooling rate is better for vitrification. However, slow-cooling is also vitrification, and the cooling rate is very slow, relative to the rates achieved with current rapid-cooling protocols using micro-volume devices. Hence, we have known for decades that you do not need a fast cooling rate to achieve vitrification.

Almost 10 years ago a different method of vitrification using conventional 0.25cc or 0.5cc straws with relatively slow cooling rates was developed (79). This method formerly called S3vitrification and more recently ICE vitrification (after several improvements), allows the use of large volumes of media, no DMSO, and sterile, sealable straws to vitrify eggs and embryos (79, 80). This system falls somewhere between slow-cooling (equilibrium freezing) and rapid-cooling (kinetic vitrification) methodology. This method may represent an intermediate closer to equilibrium vitrification, whereby the cell equilibrates with the surrounding solution so that ice cannot form no matter how slow the cooling rate; than kinetic vitrification, whereby the cooling rate is quick enough that ice does not form crystals even though it could, as described above.

Based upon the work of Luyet, Whittingham, Mazur, and company, Stachecki realized that slow-cooling allowed for vitrification, if the conditions were correct. Therefore, he concluded that plunging from higher temperatures was possible and ultra-rapid cooling rates were not necessary as long as the vitrification media used would allow for, not only the cells to be vitrified, but for their survival upon rewarming. In order to reduce the potential toxicity of the solutions, Stachecki chose to avoid using DMSO and used ethylene glycol and glycerol instead as the major permeable cryoprotectants. This allowed for longer exposure times, which were needed to sufficiently dehydrate and load the cells with cryoprotectants. The time in the final vitrification solution ranged from 90 sec to 2 min and allowed for initial results of around 90% survival and >40% pregnancy rates among five different clinics (80). They further showed that even after 4 minutes in the final vitrification solution there was only a 33% decrease in cell survival, and that the solutions were not toxic enough to kill all the cells, despite the long exposure times. In his initial study Stachecki et al. (79) demonstrated further that rapid-cooling rates were not necessary for blastocyst vitrification by cooling them slowly at -100°C for 2 min prior to plunging and storage in LN2. Direct plunging from room temperature also worked well yet still acheived a cooling rate of near 2000°C/min, almost ten-fold slower than Vajta's OPS and other micro-volume devices (See Table 1). These studies demonstrated that despite what other investigators wrote and still write about what is necessary or beneficial to achieve high success rates of vitrified human embryos this new media did not follow convention and other methods of vitrification are possible. This large-volume vitrification system allowed for 1) extended time for embryo equilibration, 2) a large volume container could be used because rapid cooling rates were not necessary, and 3) similar success rates to other rapid-cooling vitrification systems. Stachecki hypothesized that because the large-volume vitrification system did not rely on cooling rate, either rapid or slow cooling would work. This would mean that any micro-volume device such as the cryo-top, cryo-loc, rapid-i, cryo-pette, micro-secure, etc. could also work using this media. Schiewe confirmed this by using the micro-secure device to obtain high survival and pregnancy rates with Stachecki's vitrification media (76). Their current 2011 success rates for vitrifying blastocysts are over 96% with a 57% ongoing/delivery rate (Schiewe, personal communication). Currently there are several clinics successfully using micro-volume devices including the rapid-i, cryo-loc, micro-secure, and cryo-top with this media (unpublished results), demonstrating the versatility of this vitrification media.

In conclusion, we have learned that slow-cooling and rapid cooling from room temperature both can lead to a vitrified cell. Furthermore, both are very similar in cryobiological theory to achieve a vitrified state. Slow-cooling is close to equilibrium vitrification, where by the cell can be vitrified because enough water has been removed and/or bound by cryoprotectants that no ice can form no matter how slow the cooling rate. Large volume vitrification with the ICE vitrification media is similar. There are, of course, reasons that equilibrium is not totally achieved in these systems and IIF can and does occur, in some instances. However, it is close enough to equilibrium that over 90% of the cells have no IIF or at least a non-lethal amount and pregnancy rates are similar to fresh rates. Rapid cooling or vitrification using conventional 30% DMSO/EG solutions, is closer to kinetic vitrification, wherby IIF can and will form if the cooling rate is not rapid enough. Chemical toxicity is a major problem with this system. However, if everything is performed correctly, which is sometimes not so easy, rapid cooling can yield very good results.