The History and Principles of Cryopreservation D.E. Pegg, M.D.1 Downloaded by: Collections and Technical Services Department. Copyrighted material. ABSTRACT The ability of glycerol to protect cells from freezing injury was discovered accidentally. The subsequent development of cryopreservation techniques has had a huge impact in many fields, most notably in reproductive medicine. Freezing injury has been shown to have two components, direct damage from the ice crystals and secondary damage caused by the increase in concentration of solutes as progressively more ice is formed. Intracellular freezing is generally lethal but can be avoided by sufficiently slow cooling, and under usual conditions solute damage dominates. However, extracellular ice plays a major role in tissues. Cryoprotectants act primarily by reducing the amount of ice that is formed at any given subzero temperature. If sufficient cryoprotectant could be introduced, freezing would be avoided altogether and a glassy or vitreous state could be produced, but osmotic and toxic damage caused by the high concentrations of cryoprotectant that are required then become critical problems. The transport of cryoprotectants into and out of cells and tissues is sufficiently well understood to make optimization by calculation a practical possibility but direct experiment remains crucial to the development of other aspects of the cryopreservation process. KEYWORDS: Cryopreservation, cryoprotectant, freezing T he accidental discovery by C. Polge, A.U. Smith, and A.S. Parkes in 1948 that glycerol would enable fowl spermatozoa to survive freezing to 70°C1 initiated a phase of dramatic development in the techniques we now know as “cryopreservation.” Stories abound as to precisely how this discovery was made: at least it is clear that a mistake in the labeling of a bottle of solution in a refrigerator led to some fowl semen being frozen in a mixture of glycerol, albumen, and water rather than the intended solution of levulose. The levulose solution was ineffective but the glycerol solution was highly effective. At first the researchers were unaware of the composition of their apparently miraculous solution but, with the help of an analytical chemist, its true identity was revealed and a few more experiments then showed that glycerol was the active ingredient.2 A truly serendipitous discovery, but recall the observation by Hans Krebs (I think) that “Chance favours the prepared mind!” Indeed it did, and the subsequent work at the Medical Research Council’s Mill Hill Institute, particularly by Audrey Smith and a succession of young colleagues, was crucial to establishing this new field of science. It is interesting to note that the title of the paper describing that fundamental observation was “Revival of Spermatozoa after Vitrification and Dehydration at Low Temperatures”; yet these experiments did not in fact produce vitrification in the sense that is now meant by that term. Their method would now be termed the “classical freezing” approach. It is also interesting to note that much of the subsequent research and perhaps the greatest practical impact of all cryopreservation work has been in the areas of reproductive medicine, animal The Cryobiology of Assisted Reproduction: Gametes and Gonads; Editor in Chief, Bruce R. Carr, M.D.; Guest Editors, S.L. Tan, M.D., Roger G. Gosden, Ph.D., D.Sc. Seminars in Reproductive Medicine, Volume 20, Number 1, 2002. Address for correspondence and reprint requests: Professor D.E. Pegg, Medical Cryobiology Unit, Department of Biology, University of York, PO Box 373, York YO10 5YW, United Kingdom. 1Medical Cryobiology Unit, Department of Biology, University of York, York, United Kingdom. Copyright © 2002 by Thieme Medical Publishers, Inc., 333 Seventh Avenue, New York, NY 10001, USA. Tel: +1(212) 584-4662. 1526-8004,p;2002,20,01,005,014,ftx,en;sre00149x. 5 SEMINARS IN REPRODUCTIVE MEDICINE/VOLUME 20, NUMBER 1 2002 husbandry, and now the conservation of endangered species. The introduction of the glycerol technique for freezing bull spermatozoa revolutionized the cattle breeding industry, as described by Audrey Smith herself in 19613: Farmers could (now) select semen from bulls which might not have been available when required. Valuable semen was no longer wasted, and semen from bulls of uncertain value could be stored for several years until their progeny had matured and been tested for milk yield and other qualities. Another potential advantage of the method was that, during epidemics of infections such as foot and mouth disease, there would be a reserve of semen previously collected which could be used until all danger of further contagion had passed. The banks of semen at low temperatures would permit economy in the number of bulls used for breeding within one country. In addition, the possibility was opened up of vigorous international trade in frozen semen to improve stock throughout the world. This review will consider the mechanisms by which the classical freezing method seeks to preserve the viability of cells and tissues and how the more recent approach of vitrification differs. Following that initial discovery, nearly all the subsequent developments of the classical freezing method have relied upon the addition of a cryoprotective compound, such as glycerol, and the empirical optimization of a number of variables that have been shown experimentally to affect survival. These include the nature and concentration of the cryoprotectant and the temperature at which it is added, the rates of cooling and warming, the storage temperature, and the temperature and rate at which the cryoprotectant is removed. Despite its largely empirical nature, this approach was strikingly successful, and effective methods were soon developed for a wide range of cells, including the spermatozoa of several species, erythrocytes, lymphocytes, and hemopoietic cells, various endocrine cells, and many strains of tissue culture cell. These early successes are well documented in A.U. Smith’s 1961 monograph “Biological Effects of Freezing and Supercooling.”3 These practical successes stimulated a considerable volume of more fundamental work that has uncovered a number of the mechanisms that are involved: the fundamental importance of the total quantity of ice that is formed, the location of the ice crystals in relation to the cells, the toxicity of cryoprotectants and the temperature dependence of that toxicity, and the magnitude of osmotically induced changes in volume. These factors will be considered in more detail and then the ability of predictive modeling to specify preservation methods without recourse to experimentation will be addressed. DAMAGE BY ICE When a dilute aqueous solution is frozen, the ice that forms is essentially pure crystalline water; ice has negligible ability to dissolve solutes. Solutes are therefore rejected and concentrate in the dwindling volume of unfrozen liquid, depressing the chemical potential of water and achieving equilibrium with ice at each temperature. Thus, the familiar freezing point depression curve of an aqueous solution (that for NaCl/water is shown in Fig. l) also describes the dependence of solution composition upon temperature. In the presence of ice and at equilibrium, composition is determined solely by temperature and is therefore independent of initial composition; however, initial composition does control the amount of ice that forms and the factor by which the concentration increases at a given temperature. It is important to realize that the increase in concentration of NaCl that occurs during the freezing of isotonic saline is enormous, actually 32-fold at 21°C. The question then arises: Is it the ice, the elevated salt concentration, or both that damage cells during progressive freezing? A rather definite answer was given in the early 1950s4–7 by Jim Lovelock, who was a colleague of Polge, Smith, and Parkes at Mill Hill. Lovelock showed that the extent of hemolysis observed in human erythrocytes that were frozen to various temperatures in solutions of sodium chloride could be accounted for quantitatively by the effect of the salt concentration produced by freezing to each temperature. A duplication of that experiment in our laboratory is illus- Figure 1 The phase diagram of sodium chloride/water. From left to right the curve shows the freezing points of solutions of sodium chloride of increasing strength. The lowest possible freezing point is 21.4°C, at which temperature the concentration is almost 24% in w/w terms. If solutions less concentrated than this are cooled, when they reach the curve, sufficient ice will crystallize to hold the solution composition on the line. For concentrations greater than 24% w/w, the curve describes the dependence upon temperature of the concentration of saturated solutions of sodium chloride. Downloaded by: Collections and Technical Services Department. Copyrighted material. 6 Figure 2 Graph showing the extent of hemolysis when human erythrocytes suspended in isotonic saline were frozen to and thawed from the indicated temperatures (solid line). For comparison, the broken line shows the hemolysis produced by exposing the cells to the corresponding concentrations of sodium chloride followed by resuspension in isotonic saline. trated in Figure 2, and it does indeed show an impressive correspondence between the effect of the two treatments. However, it should be noted that an experiment of this sort cannot prove causation: I am told that there is a very close correlation between the production of pig iron in the United States in the 1980s and the birth rate in the United Kingdom, yet I doubt if anyone would argue that either one caused the other! But in this case a rational causative mechanism was postulated and the close correlation convinced most workers. Lovelock went on to explain the cryoprotective action of glycerol and of other highly soluble, nontoxic, penetrating solutes by their ability to moderate the increase in salt concentration that occurs during freezing; this is demonstrated for the system NaCl/glycerol/water in Figure 3 and tended to reinforce acceptance of the “saltdamage” theory of freezing injury. In addition, we have shown that the addition of glycerol has the same influence on the effect of exposure to high concentrations of salt as it does on the effect of salt concentration by freezing,8 and we have argued that this makes it even more probable that the salt-damage theory is valid. However, the fundamental logical position is unchanged and Mazur has pointed out9 that there is a fixed relationship between salt concentration and the amount of ice formed at any given subzero temperature when solutions of a given cryoprotectant in isotonic saline are frozen: although cryoprotectants actually moderate the Figure 3 The effect of glycerol on the increase in solute concentration during progressive freezing. Each solution contained isotonic saline plus the initial concentration of glycerol that is shown against each curve in mole fraction terms. The corresponding molar concentrations are 0 and approximately 1, 2, 3, and 4 mol/L. The concentration factor for each solution at each temperature applies both to glycerol and sodium chloride. rise in salt concentration by reducing the amount of ice, the existence of the phenomenon of cryoprotection does not support either the salt- or the ice-damage theory. Further, Mazur has provided extensive experimental evidence that ice may indeed have a direct damaging action.9–12 He did this by freezing erythrocytes in solutions of glycerol and sodium chloride in which the initial concentration of salt varied between 0.6 X and 4.O X isotonic, which made it possible to separate, to a considerable extent, the increase in salt concentration from the quantity of ice that formed. Mazur found a stronger correlation between severe damage and the amount of ice rather than the salt concentration. One problem with this sort of experiment is that the means used to separate the variables (different initial tonicities) may also affect the behavior of the cells: cells destined to be subjected to low “unfrozen fractions” enter the experiment in a swollen state, and we have suggested that such cells may behave differently from normal cells.13 In fact, we were able, with erythrocytes suspended in solutions of NaCl/glycerol/water, to demonstrate remarkably similar responses to freezing and thawing as to exposure to equivalent concentrations of solute and redilution.8 When similar experiments were carried out in the absence of glycerol, the correspondence between the effects of freezing and solution exposure was good for cells in isotonic and 2 X isotonic saline, but freezing was substantially more damaging than solution exposure when the cells were suspended in 0.6 X or 4 X saline.14 The mechanism of this effect is unclear, but it is consistent with the proposition that 7 Downloaded by: Collections and Technical Services Department. Copyrighted material. THE HISTORY AND PRINCIPLES OF CRYOPRESERVATION/PEGG SEMINARS IN REPRODUCTIVE MEDICINE/VOLUME 20, NUMBER 1 2002 cells suspended in a range of tonicities of saline do not behave as a uniform population. The matter remains unresolved; the general view is that ice probably has no direct role in slow freezing injury to erythrocytes, but that possibility cannot be excluded. In the foregoing discussion we have ignored the fact that real biological systems are compartmentalized, the intracellular spaces being separated from the extracellular space by semipermiable cell membranes. Therefore it is important to consider whether ice forms inside or outside the cells or both. Freezing is a nucleationinduced event, and the probability of nucleation increases directly with the degree of supercooling and with the volume. Thus, as our compartmentalized model cools below the equilibrium freezing point of the saline, it is inevitable that a nucleation event will occur in the single, large extracellular compartment before very many of the cells have nucleated. Once that has happened, further cooling will result in growth of that extracellular ice, concentration of the extracellular solution, and dehydration of the cells by osmosis through their semipermeable membranes. Thus, providing the cooling rate is sufficiently low and the water permeability of the cells sufficiently high, the intracellular spaces will remain free of ice. This fundamental phenomenon was discovered and elucidated by Mazur15 at Oak Ridge National Laboratory. He provided a quantitative analysis of the effect of cooling rate on water transport during progressive cooling and correlated the predicted extent of intracellular supercooling with the known effect of cooling rate on cell survival: the greater the degree of supercooling, the greater the probability of freezing. It was shown that cooling rates that produced significant supercooling with each cell also caused the survival to fall. The phenomenon is best illustrated by Figure 4, compiled by Leibo,16 which shows, for three types of cell with differing water permeabilities, an inverse correlation between intracellular freezing and survival. This is the basis of the so-called “two-factor hypothesis” of freezing injury: solute damage at low cooling rates where extracellular ice is probably innocuous to cells in suspension; intracellular freezing at high cooling rates, which is generally lethal. Each cell/cryoprotectant combination is associated with an optimum cooling rate. This discussion has dealt with a simplified model of cells in suspension and the simplest of all mammalian cells (the erythrocyte), together with a few examples of more typical mammalian cells in suspension. But many of us would like to be able to cryopreserve more complex systems—multicellular, organized tissues and even organs. Several factors conspire to make this a vastly more difficult task. Most obviously, tissues consist of many types of cells that may differ in their requirements for optimal preservation; the dimensions of the tissue or organ will impose restrictions on the rates of cooling and heating that can be used; and above all, the function of an organ or tissue depends upon the interrelations and interconnections between the cells and the preservation of an adequate three-dimensional arrangement with intact nonliving intercellular structures such as basement membranes, glycosaminoglycans, and collagen fibers. The point to be emphasized is that ice forming outside the cells may nevertheless be within the system we wish to preserve and may therefore produce lethal injury. In fact, we have concluded that damage due to extracellular ice is the single most serious obstacle to the extension of cryopreservation techniques to structured multicellular systems.17,18 In the case of vascularized tissues we have obtained both experimental and theoretical evidence for the crucial role of intravas- Figure 4 The effect of cooling rate on the percentage of cells that were seen to freeze internally (solid line) and the percentage that survived freezing and thawing (broken line). Mouse ova were frozen in 1 M DMSO, HeLa cells in tissue culture medium, and erythrocytes in 1.5 M glycerol.16 Downloaded by: Collections and Technical Services Department. Copyrighted material. 8 cular freezing: rabbit kidneys, frozen after equilibration with 2 M glycerol, were found to have ruptured glomerular capillaries19; mathematical modeling of the process of freezing in a Krogh cylinder model of vascularized tissue showed that the capillaries would have to expand eight-fold to accommodate the required volume of ice, which is far beyond their elastic limit.20 We conclude that ice is exceedingly damaging to tissues, through a number of mechanisms, even in the presence of concentrations of cryoprotectant that would ordinarily be expected to provide a high degree of cell protection. The route to effective cryopreservation of such systems would therefore seem to be via ice-free cooling, or vitrification. The term vitrification may be defined for aqueous systems as the conversion of the system from a fluid to a solid solely by an increase in viscosity, without any crystallization of water and therefore in the complete absence of ice. Before considering vitrification in more detail we must consider two other factors that are important in conventional freeze-preservation, but which acquire even greater importance in vitrification where the concentrations of added solute are much greater: the chemical toxicity of the additives and the osmotic consequences of their addition and removal. THE TOXICITY OF CRYOPROTECTANTS Cryoprotectants are of necessity tolerated in high concentrations; glycerol is, in this sense, far less harmful than NaCl, but at a sufficiently high concentration any compound will be “toxic.” This has two important consequences: the highest concentration that the tissue will tolerate prior to preservation is limited, and, during freezing, the concentration will increase as ice separates. In vitrification, as opposed to freezing, a much higher initial concentration is needed but no further concentration occurs during cooling because freezing does not occur. In both techniques one seeks the highest tolerable concentration, in freezing to reduce the salt concentration and in vitrification to achieve the vitreous state without freezing. In practice it is found that the maximum concentration that can be achieved without impairment of viability is dependent on the temperature and rate of addition and removal; temperature dependence is due partly to chemical toxicity (which is reduced by reduction in temperature) and partly to osmotic effects (which are increased by reduction in temperature). Consider chemical toxicity first. Cryoprotectants are frequently added at 0 to 4°C rather than at room temperature to take advantage of the positive temperature coefficient of chemical toxicity, and this has assumed a far greater importance in attempts at vitrification. Thus, for the vitrification of mouse embryos, Rall and Fahy used a two-stage incubation with their cryoprotectant mixture, first at 20°C with one quarter of the final concentration and then at 4°C with the final concentration.21 Much earlier than this, Elford and Walter22 increased the concentration of dimethyl sulphoxide in smooth muscle to a final level of 50% w/v by a series of step increases starting at 37°C (to 20%), then after cooling to 7°C (to 30%), then at 14°C (to 40%), at 22°C (to 50%), and finally at 39°C (to 60%), following which the muscles were cooled to 79°C. Reversal of this stepwise process during warming permitted full recovery of contractile function, and no ice was formed at 79°C. Thus, in this system, a very high concentration of cryoprotectant was tolerated when the temperature and rate of addition and removal were appropriately optimized. Empirical experiments in our laboratory have shown that rabbit kidneys will tolerate up to 4 M glycerol,23 and rabbit corneas will recover after exposure to 4.25 M dimethyl sulphoxide.24 Unfortunately, the concentrations of cryoprotectant needed to vitrify aqueous systems are usually in excess of 5.5 M and cannot be achieved without severe toxicity. With some cryoprotectants, even using conventional freezing methods of cryopreservation, it seems that toxic levels are reached. This was first pointed out by Lovelock in his study of the cryopreservation of erythrocytes with methanol, which, irrespective of the initial concentration, caused recovery to fall to negligible levels when the temperature fell below 55°C. With some systems a similar, if less severe, phenomenon occurs with propan-1,2-diol (propylene glycol); for example with human platelets25 and rabbit kidneys,26 but other systems, including rabbit cornea27 and human embryos,28 seem to tolerate this cryoprotectant more readily. It seems very likely that all cryoprotectants produce some damage at the concentrations used in practice. Certainly this is true of glycerol and human erythrocytes,8 but where the additive is effective this toxic action is clearly less than the cryoprotective effect. The crucial points are that there always is a toxic limit to the concentration of cryoprotectant that can be used and that the apparent toxic limit is strongly influenced by the conditions under which the cryoprotectant is added and removed. OSMOTIC EFFECTS OF CRYOPROTECTANTS The most effective cryoprotectants penetrate cell membranes but they do so more slowly than water, which means that some osmotic imbalance is inevitable during the addition or removal of these compounds. Gross osmotic shock results in cell damage, ultimately in lysis, and it is therefore both logical and of proven efficacy to control changes in cell volume so that acceptable limits are not transgressed. An important step in designing cryopreservation methods is therefore to measure the 9 Downloaded by: Collections and Technical Services Department. Copyrighted material. THE HISTORY AND PRINCIPLES OF CRYOPRESERVATION/PEGG SEMINARS IN REPRODUCTIVE MEDICINE/VOLUME 20, NUMBER 1 2002 volume response to changes in external osmolality of a nonpermeating solute and to correlate each volume with subsequent structural and functional damage. The next step is to measure or, failing that, to estimate the water and solute permeability for the chosen cryoprotectant. Two stratagems for controlling volume during the addition and removal of penetrating solutes are available: the first is to use low rates of change in concentration or, what amounts to the same thing, to change the concentration in several small steps; the other approach, which is applicable only to the more critical removal phase, is to incorporate solutes that do not penetrate the cells and therefore function as “osmotic buffers” by restricting the inflow of water as the concentration of cryoprotectant is reduced. These highly effective procedures are illustrated for the addition and removal of dimethyl sulphoxide with human oocytes29 in Figure 5. Such processes are even more important when concentrations of penetrating agents sufficient to achieve vitrification are used. VITRIFICATION OF CELLS AND TISSUES Vitrification occurs when the viscosity of the solution reaches a sufficient value to inhibit the crystallization of ice. Although the concentration of solute in the remaining liquid phase increases during progressive freezing, a temperature will eventually be reached with many systems at which that residual liquid vitrifies in the presence of ice. It has already been pointed out that under the cooling conditions that are usually used, ice does not form inside the cells, and in particular because intracellular protein promotes vitrification, it is actually the case that the cells in conventionally frozen material are vitrified. Polge, Smith, and Parkes were not, after all, in error in the title of their paper,1 although the presence of extracellular ice would render the term vitrification inappropriate in today’s usage. Had it been possible to reach the concentration of glycerol required to vitrify before cooling was initiated, true vitrification of the whole system would have occurred, but this would require about 80% w/w, far beyond the toxic limit at 0°C. There are, however, more favorable solutes than glycerol, perhaps the most encouraging recent additions to the list being propan-1,2-diol and butan-2,3-diol,30 which will vitrify at a concentration less than 50%. This approach may be termed the equilibrium approach to vitrification: the system will vitrify no matter how slowly it is cooled. There is, however, another approach that is best illustrated by Figure 6: the first stage in the process of freezing is nucleation, and nucleation has an unusual temperature dependence in that it becomes more active rather than less active with reduction of temperature, until it is limited by viscosity. However, the growth of ice crystals has a more usual temperature dependence, in that it is slowed and eventually arrested by cooling. The interesting point is the manner in which these two processes interact: the rate of ice crystal growth in the solution illustrated in Figure 6 has reached low values before the temperature zone for active nucleation is entered. Consequently three possibilities exist: if cooled sufficiently rapidly the sample may escape both nucleation and ice crystal growth; if cooling is more rapid the sample may be nucleated but without ice crystals; or, with slower cooling, the sample may nucleate and ice then may form. Critical cooling rates (that avoid the crystal- Figure 5 Calculated changes in cell volume for mature human oocytes as they are exposed in sequence to 0.8, 1.5, 0.64, and 0 M DMSO in isotonic media. This scheme will restrict the osmotically induced excursions in cell volume to 30%, during both the addition and the removal of DMSO. Downloaded by: Collections and Technical Services Department. Copyrighted material. 10 Figure 6 A diagram to illustrate the dependence of the nucleation of ice and the growth of ice crystals on time and temperature. The material is a thin film of 50% polyvinylpyrrolidone. The lines with arrows indicate cooling rates that will avoid nucleation (300°C/s), will allow nucleation but avoid ice crystal growth (80°C/s), or will allow ice crystals both to form and to grow (20°C/s). lization of ice) for a wide range of solutes have been published by Sutton.31–33 Figure 6 carries another danger message and this concerns warming: as the temperature is raised the sample traverses the nucleation zone first and then the zone of ice crystal growth, so the stage is set for freezing to occur during warming, even if it was avoided during cooling. The obvious remedy for this situation is ultrarapid heating, and irradiation with microwaves is the obvious technique to use because of its potential for depositing energy uniformly as well as rapidly, even in bulky samples. However, the critical warming rates are orders of magnitude greater than the corresponding critical cooling rates. Considerable effort is now being devoted to the definition of the problem of microwave heating of appropriate systems from very low temperatures34,35 and to its solution. Other maneuvers that may help include the use of extremely high hydrostatic pressures to depress both the freezing and the nucleation temperature36 and the addition of materials that retard the growth of ice crystals, such as the so-called “antifreeze” proteins found in some fishes and insects37,38 and certain synthetic so-called “ice-blockers.”39 THE ROLE OF MATHEMATICAL MODELING Following the mathematical analysis by Mazur15 in which he showed how intracellular freezing was controlled by the cooling rate and the permeability of the cell membrane to water, there has been steady progress in the modeling of cryopreservation in a variety of cells and even in tissues. Such studies have been very useful as a means of exploring mechanisms and testing ideas that can then be subjected to experimental investigation. To be useful as predictors, such models must be physically realistic and be supplied with reliable numerical data: this cannot always be done at the present time. The most secure use of models is in predicting safe methods for the addition and removal of cryoprotectants with isolated cells at a fixed temperatures above 0°C. Here one needs some standard chemical data on the cryoprotective solute and knowledge of the temperature, the size of the cells, their water content, the time course of their volume response to known concentrations of the cryoprotectant, and an impermeant solute and the limits of their osmotic tolerance to that impermeant solute. These data permit the calculation by means of standard solute and solvent flux equations40 of osmotically active volume, hydraulic conductivity, and solute permeability. It is then possible to use the derived data to predict volume time courses for other regimes of addition and removal: this has been both useful and reliable.25,41 However, to predict the fluxes that occur during cooling and their potential effects requires knowledge of the temperature dependence, not only of the permeability coefficients but also of the effective osmotic volume and the susceptibility of the cells to changes in volume. Extrapolation of permeability constants to subzero temperatures from measurements made at suprazero temperatures seems hazardous, and data concerning the 11 Downloaded by: Collections and Technical Services Department. Copyrighted material. THE HISTORY AND PRINCIPLES OF CRYOPRESERVATION/PEGG SEMINARS IN REPRODUCTIVE MEDICINE/VOLUME 20, NUMBER 1 2002 other temperature coefficients is simply absent. In this author’s opinion, there is no alternative at this time to direct experimentation for the determination of optimal cooling rates. The situation concerning rewarming is even less satisfactory; the mechanisms are not even well understood qualitatively, let alone quantitatively. Measurements of solute permeation into tissues have also been published and are useful in guiding the design of experimental protocols, but with less precision than with cells.42,43 11. 12. 13. 14. CONCLUSIONS Cryopreservation does have a rational, scientific basis but there are still many unanswered questions and some cells and tissues remain refractory. Vitrification remains an enticing vision but the problems of vitrifying organized systems remain formidable. Yet it is toward vitrification that the signposts for preservation of these refractory systems clearly point. At this time most if not all of the systems that have been preserved by vitrification (such as early embryos, monocytes, pancreatic islets, and blood vessels [see Refs. 21 and 44]) can also be preserved by conventional freeze-preservation methods. The problem is to identify solutes that can be included in acceptable concentrations and that will vitrify at realizable rates of cooling and warming, yet will also be compatible with viability. 15. 16. 17. 18. 19. 20. REFERENCES 1. Polge C, Smith AU, Parkes AS. Revival of spermatozoa after vitrification and dehydration at low temperatures. Nature (Lond) 1949;164:666 2. Parkes AS. Preservation of living cells and tissues at low temperatures. Proc III Internat Congress Animal Reproduction, Cambridge, 1956:69 3. Smith AU. Biological Effects of Freezing and Supercooling. London: Edward Arnold; 1961 4. Lovelock JE. The haemolysis of human red blood cells by freezing and thawing. Biochim Biophys Acta (Amst) 1953; 10:414–426 5. Lovelock JE. The mechanism of the protective action of glycerol against haemolysis by freezing and thawing. Biochim Biophys Acta (Amst) 1953;11:28–36 6. Lovelock JE. The protective action by natural solutes against haemolysis by freezing and thawing. Biochem J 1954;56: 265–270 7. Lovelock JE. Haemolysis by thermal shock. Brit J Haemat 1955;1:17–29 8. Pegg DE, Diaper MP. On the mechanism of injury to slowly frozen erythrocytes. Biophys J 1988;54:471–488 9. Mazur P, Rall WF, Rigopoulos N. The relative contributions of the fraction of unfrozen water and of salt concentration to the survival of slowly frozen human erythrocytes. Biophys J 1981;36:653–675 10. Mazur P, Rigopoulos N. Contributions of unfrozen fraction and of salt concentration to the survival of slowly frozen 21. 22. 23. 24. 25. 26. 27. 28. 29. human erythrocytes: influence of warming rate. Cryobiology 1983;20:274–289 Mazur P, Cole KW. Influence of cell concentration on the contribution of unfrozen fraction and salt concentration to the survival of slowly frozen human erythrocytes. Cryobiology 1985;22:505–536 Mazur P, Cole KW. Roles of unfrozen fraction, salt concentration and changes in cell volume in the survival of frozen human erythrocytes. Cryobiology 1989;26:1–29 Pegg DE, Diaper MP. The “unfrozen fraction” hypothesis of freezing injury to human erythrocytes: a critical examination of the evidence. Cryobiology 1989;26:30–43 Pegg DE, Diaper MP. The effect of initial tonicity on freezethaw injury to human red cells suspended in solutions of sodium chloride. Cryobiology 1989;26:579–560 Mazur P. Kinetics of water loss from cells at subzero temperatures and the likelihood of intracellular freezing. J Gen Physiol 1963;47:347–369 Leibo SP. Preservation of mammalian cells and embryos by freezing. In: Simatos D, Strong DM, Turk M, eds. Les Colloques de I’Institut National de la Sante et de la Recherche Medicale: 62. Cryoimmunologie. Paris: Inserm; 1977: 311–334 Pegg DE, Jacobsen IA. Current status of cryopreservation of whole organs with particular reference to the kidney. In: Marberger M, Dreikorn K, eds. Renal Preservation. Baltimore: Williams and Wilkins; 1983:301–322 Pegg DE. Ice crystals in tissues and organs. In: Pegg DE, Karow AM Jr, eds. The Biophysics of Organ Cryopreservation. New York and London: Plenum Press; 1987:117–140 Jacobsen IA, Pegg DE, Starklint H, et al. Effect of cooling and warming rate on glycerolised rabbit kidneys. Cryobiology 1984;21:637–653 Rubinsky B, Pegg DE. A mathematical model for the freezing process in biological tissue. Proc R Soc Lond B Biol Sci 1988;234:343–358 Rall WF, Fahy GM. Ice-free cryopreservation of mouse embryos at 196°C by vitrification. Nature (Lond) 1985;313: 573–575 Elford BC, Walter CA. Effects of electrolyte composition and pH on the structure and function of smooth muscle cooled to 79°C in unfrozen media. Cryobiology 1972;9:82–100 Jacobsen IA, Pegg, DE, Wusteman MC, Robinson SM. Transplantation of rabbit kidneys perfused with glycerol solutions at 10°C. Cryobiology 1978;15:18–26 Taylor MJ. Equilibrium approach to ice-free cryopreservation of corneas: tolerance of endothelium to high concentrations of dimethyl sulphoxide. Cryobiology 1988;25:533 Arnaud FG, Pegg DE. Cryopreservation of human platelets with propane-1,2diol. Cryobiology 1990;27:130–137 Jacobsen IA, Pegg DE, Starklint H, Hunt CJ, Diaper MP. Introduction and removal of cryoprotectants with rabbit kidneys: assessment by transplantation. Cryobiology 1988;25: 285–299 Rich SJ, Armitage WJ. Propane-1,2-diol as a potential component of a vitrification solution for corneas. Cryobiology 1990;27:42–54 Testart J. High pregnancy rate after early embryo freezing. Fertil Steril 1986;46:268 Newton H, Pegg DE, Barrass R, Gosden RG. Osmotically inactive volume, hydraulic conductivity and permeability to dimethyl sulphoxide of human mature oocytes. J Reprod Fertil 1999;117:27–33 Downloaded by: Collections and Technical Services Department. Copyrighted material. 12 30. Boutron P. Non-equilibrium formation of ice in aqueous solutions: efficacy of polyalcohol solutions for vitrification. In: Pegg DE, Karow AM Jr, eds. The Biophysics of Organ Cryopreservation. New York and London: Plenum Press; 1987:201–236 31. Sutton RL. Critical cooling rates to avoid ice crystallization in solutions of cryoprotective agents. J Chem Soc Faraday Trans 1991;87:101–105 32. Sutton RL. Critical cooling rates to avoid ice crystallization in aqueous cryoprotectant solutions containing polymers. J Chem Soc Faraday Trans 1991;87:3747–3751 33. Sutton RL. Critical cooling rates for aqueous cryoprotectants in the presence of sugars and polysaccharides. Cryobiology 1992;29:585–598 34. Marsland TP, Evans S, Pegg DE. Dielectric measurements for the design of an electromagnetic rewarming system. Cryobiology 1981;24:311–323 35. Robinson MP, Pegg DE. Rapid electromagnetic warming of cells and tissues. IEEE Trans Biomed Eng 1999;46:1413–1425 36. Fahy GN, MacFarlane DR, Angell CA, Meryman HT. Vitrification as an approach to cryopreservation. Cryobiology 1984;21:407–426 37. De Vries AL. Biological antifreeze agents. Am Rev Physiol 1983;45:245 38. Sutton RL, Pegg DE. Devitrification in butane-2,3-diol solutions containing anti-freeze peptide. Cryo-Letters 1993;14: 13–20 39. Wowk B, Darwin M, Harris SB, Russell SR, Rasch CM. Effects of solute methoxylation on glass-forming ability and stability of vitrification solutions. Cryobiology 1999;39: 215–227 40. Kleinhans FW. Membrane permeability modelling; KedemKatchalsky vs a two-parameter formalism. Cryobiology 1998; 37:271–289 41. Pfaff RT, Agca Y, Liu J, Woods EJ, Peter AT, Critser JK. Cryobiology of rat embryos 1: determination of zygote membrane permeability coefficients for water and cryoprotectants, their activation energies and the development of improved cryopreservation methods. Biol Reprod 2000;63: 1294–1302 42. Bateson EAJ, Busza AL, Pegg DE, Taylor MJ. Permeation of rabbit common carotid arteries with dimethyl sulphoxide. Cryobiology 1994;31:393–397 43. Newton H, Fisher J, Arnold JRP, Pegg DE, Faddy MJ, Gosden RG. Permeation of human ovarian tissue with cryoprotective agents in preparation for cryopreservation. Hum Reprod 1998;13:101–105 44. Song YC, Khirabadi BS, Lightfoot FG, Brockbank KGM, Taylor MJ. Vitreous cryopreservation maintains the function of vascular grafts. Nat Biotechnol 2000;18:296–299 13 Downloaded by: Collections and Technical Services Department. Copyrighted material. THE HISTORY AND PRINCIPLES OF CRYOPRESERVATION/PEGG