Biological metabolism in living cells dramatically

Transcription

1 Mechanisms of Cryoinjury in Living Cells Doyong Gao andj. K. Critser Abstract Biological metabolism in living cells dramatically diminishes at low temperatures, a fact that permits the long-term preservation of living cells and tissues for either scientific research or many medical and industrial applications (e.g., blood transfusion, bone marrow transplantation, artificial insemination, in vitro fertilization, food storage). However, there is an apparent contradiction between the concept of preservation and experimental findings that living cells can be damaged by the cryopreservation process itself. The challenge to cells during freezing is not their ability to endure storage at very low temperatures (less than -180 C); rather, it is the lethality of an intermediate zone of temperature (-15 to -60 C) that a cell must traverse twice once during cooling and once during warming. Cryobiological research studies the underlying physical and biological factors affecting survival of cells at low temperatures (during the cooling and warming processes). These factors and mechanisms (or hypotheses) of cryoinjury and its prevention are reviewed and discussed, including the most famous "twofactor hypothesis" theory of Peter Mazur, concepts of cold shock, vitrification, cryoprotective agents (CPAs), lethal intracellular ice formation, osmotic injury during the addition/ removal of CPAs and during the cooling/warming process, as well as modeling/methods in the cryobiological research. Keywords: cryobiology; cryopreservation; cryoprotective agents; fundamental cryobiology; intracellular ice formation; principles/mechanisms of cryoinjury Introduction Cryobiology is a multidisciplinary science to study the physical and biological behaviors of cells and tissues (including their interactions with environment) at low temperatures (especially less than the freezing point of water, 0 C). Biological metabolism in living cells dramatically diminishes at low temperatures, a fact that permits the long-term preservation of cells and tissues for scientific research as well as many medical applications such as blood transfusion, bone marrow transplantation, artificial insemination, and in vitro fertilization. The major steps in the cryopreservation process can be summarized as follows: 1. Add cryoprotective agents (CPAs l ) to cells/tissues before cooling. 2. Cool the cells/tissues toward a low temperature (e.g., at -196 C, the liquid nitrogen temperature at pressure of 1 atm) at which the cells/tissues are stored. 3. Warm the cells/tissues. 4. Remove the CPAs from the cells/tissues after thawing. Major advantages of cryopreservation include the possible banking of a large quantity of cells and tissues for human leukocyte antigen typing and matching between recipients and donors; facilitating the transport of cells/tissues between different medical centers, and allowing sufficient time for the screening of transmissible diseases (e.g., HIV) in the donated cells and tissues before transplantation. However, there is an apparent contradiction between the concept of preservation and experimental findings that cells/tissues can be damaged by the preservation process itself. Injury to cells has been shown to be caused by each step or a combination of the steps described above. An extremely important part of the research in fundamental cryobiology is to reveal the underlying physical and biological mechanisms related to the injury of cells at low temperatures (so-called cryoinjury), especially those associated with the phase change of water in both extra- and intracellular environments. Understanding the mechanisms of cell cryoinjury will make it possible to establish biophysical/mathematical models describing cell responses to environmental change during the cryopreservation process and to use these models to develop the optimal procedures/devices for the long-term cryopreservation of living cells, preventing cryoinjury. The fundamental cryobiology theories associated with the mechanisms of the cryoinjury in cells and its prevention are reviewed and discussed below. Dayong Gao, Ph.D., is Associate Professor in the Department of Mechanical Engineering, Center for Biomedical Engineering, and the Department of Physiology, University of Kentucky, Lexington, Kentucky. J. K. Critser, Ph.D., is Director of the Cryobiology Institute and Professor in the Departments of Pediatrics and Immunology and Microbiology, Indiana University Medical School, Indianapolis, Indiana. 'Abbreviations used in this article: CPA, cryoprotective agent; E a, activation energy; IIF, intracellular ice formation; L, permeability coefficient of a cell to water; S/V, surface area to volume ratio. Volume 41, Number

2 Cryoinjury Mechanisms during the Freezing Processes Contrary to popular belief, the challenge to cells during cryopreservation is not their ability to endure storage at low temperature; rather, it is the lethality of an intermediate zone of temperature (-15 to -60 C) that cells must traverse twice once during cooling and once during warming (Mazur 1963). As cells are cooled to approximately -5 C, both the cells and surrounding medium remain unfrozen and supercooled. Between -5 and approximately -15 C, ice forms (either spontaneously or as a result of artificially introducing ice crystals to the solution) in the external medium; however, the cells' contents remain unfrozen and supercooled, presumably because the plasma membrane blocks the growth of ice crystals into the cytoplasm. The supercooled water in the cells has, by definition, a greater chemical potential than that of water in the partially frozen extracellular solution; and water thus flows out of the cells osmotically and freezes externally (Figure 1). The subsequent physical events in the cells depend on the cooling rate, as shown in Figure 1. If cells are cooled too rapidly, intracellular water is not lost fast enough to maintain equilibrium; the cells become increasingly supercooled, eventually attaining equilibrium by freezing intracellularly (Mazur 1963, 1990). Muldrew and McGann (1994) invoked an osmotic rupture hypothesis and model of intracellular freezing injury based on solid experimental data and theoretical derivations, indicating that the water flux across the cell membrane during the freezing/thawing can induce intracellular ice formation causing injury. Diller and Cravalho (1970) (also Diller 1982) developed novel microscopy to observe the cell behavior as well as intracellular ice formation (IIF 1 ) temperatures and HF dynamics during the freezing processes. More recently, based on heterogeneous nucleation theory and thermodynamics as well as experimental results, Toner and colleagues (1990, 1993) developed an IIF model +22 C -10 C and equations describing the probability of IIF as a function of cooling rate, temperature, and cryobiological characteristics of a specific cell type. The freezing behavior of the cells can be modified by the addition of CPAs, which affect the rates of water transport, nucleation, and crystal growth. Karlsson and coworkers (1994) further developed a model of diffusion-limited ice growth inside cells in the presence of CPAs. They successfully predicted the IIF phenomena as a function of cooling rate, temperature, and CPA concentration; and they also predicted optimum cooling protocols preventing IIF. In most cases, cells undergoing IIF are killed (Farrant 1970, 1977; Fujikawa 1980; Mazur 1977, 1984; Muldrew and McGann 1994; Rapatz et al. 1963; Steponkus and Wiest 1979; Stowell et al. 1965; Toner et al. 1990,1993; Trump et al. 1965). If cooling is sufficiently slow, the cells will lose water rapidly enough to concentrate the intracellular solutes sufficiently to eliminate supercooling. As a result, the cells dehydrate and do not freeze intracellularly. However, if the cells are cooled too slowly, they will experience a severe volume shrinkage and long-term exposure to high-solute concentrations (composed mainly of electrolytes) before they reach a eutectic temperature (when all components in a solution are solidified). Both volume shrinkage and long-term exposure may cause cell injury (Figure 1). Lovelock (1957) proposed that the increased concentration of solutes and cell dehydration have deleterious effects on the lipid-protein complexes of cell membranes, weakening them and increasing lipid and phospholipid losses. The cell is therefore rendered permeable to electrolytes and swells, eventually rupturing. Levitt (1962) proposed that loss of water from the protoplasm brings protein molecules into apposition, presenting the opportunity for formation of new chemical bonds previously too distant and rigidly structured in hydrated form to permit combination. Thawing has a disruptive force on the new combination, permitting unfolding and denaturation. Slow Cooling -30 C Optimal Cooling Rapid Cooling Figure 1 Schematic drawing of physical events in cells during freezing. 188 WAR Journal

3 Karow and Webb (1965) explained slow freezing injury as a consequence of the extraction of bound water from cellular structures for incorporation into ice crystals, denuding proteins of lattice-arranged bound water essential to cell integrity. The effects that may cause cell injury as a result of concentration of solutes, termed "solution effects," have been characterized collectively by Mazur and colleagues (1972), who suggest that these effects on cells are greatly enhanced in a slow cooling process when the exposure of cells to a highly concentrated solution is prolonged. They also indicated that hyperosmotic stress may cause a net leak/influx of nonpermeating solutes (mainly electrolytes); when cells are returned to isosmotic conditions, they swell beyond their normal isotonic volume and lyse. Meryman (1970, 1974) proposed that the slow freezing results in the decrease in intracellular volume beyond a critical volume. As the cell reduces in size in response to increasing extracellular osmolality, the compression of contents enhances the resistance to further shrinking. This compression results in a hydrostatic pressure difference across the cell membrane, incurring cell membrane damage. Steponkus and Wiest (1979) invoked a maximum cell surface area hypothesis: The cell shrinkage induces irreversible membrane fusion, hence the effective area of cell membrane is reduced; when returned to isotonic condition, the cells lyse before their normal volume is recovered. Some cells have been shown to exhibit a reversible inhibition of shrinkage at some point during a progressive increase in extracellular solute concentration, implying the development of membrane stress (Williams 1979). Alternate theories have embraced such mechanisms of slow-freezing injury as macromolecule denaturation from dehydration or phase transition of membrane lipids (Fishbein et al. 1978). Pegg and *? " W Diaper (1988) invoked a "packing effect," which is an inverse dependence of cell survival on the proportion of the initial sample volume occupied by the cells. Mazur and Cole (1985), however, have proposed that a considerable portion of the damage, at least in human erythrocytes, is due to the reduction in the size of the unfrozen channels, which results in increased cell-ice or cell-cell contacts or interactions. In general, a cooling rate that is either too high or too low can kill cells, although the mechanisms underlying cell damage are different. These cell responses to freezing were first expressed quantitatively by Mazur (1963) and directly linked with cell cryoinjury by the following two-factor hypothesis (Mazur et al. 1972): (1) At slow cooling rates, the cryoinjury occurs due to the solution effects (i.e., the solute/electrolyte concentration, severe cell dehydration, and the reduction of unfrozen fraction in the extracellular space); and (2) at high cooling rates, cryoinjury occurs due to the lethal IIF. The optimal cooling rate for cell survival should be low enough to avoid HF but high enough to minimize the solution effects. Based on these criteria, an optimal cooling rate for cell cryosurvival should exist between high and low rates, as shown in Figure 1. This scheme has been confirmed experimentally. As shown in Figure 2, cell survival plotted as a function of cooling rate produces a characteristic inverted U-shaped curve. There is an optimal cooling rate for a specific cell type. For a given cell type, a low cooling rate (relative to the optimal cooling rate) results in solution effects cryoinjury, whereas a high cooling rate (relative to the optimal cooling rate) results in lethal IIF. It is clear from the cryobiology theory described above that whether a given cooling rate is too high or low for a given cell type depends on the ability of water to move across the cell membrane (i.e., the water permeability of the cell COOLING RATE (C/min) Figure 2 Relation (inverted "U" shape) between percentage of survival and cooling rate in three types of cells. Stem cells refer to mouse bone marrow stem cells, and hamster cells refer to hamster oocytes. RBC refers to human red blood cells. Adapted from Mazur P Freezing and low temperature storage of living cells. In: Muhlbock O, ed. Proceedings of the 1974 Workshop on Basic Aspects of Freeze Preservation of Mouse Strains, Jackson Laboratory, Bar Harbor, ME. Stuttgart: Gustav Fisher Verlag. p Volume 41, Number

4 membrane, which is cell type dependent). Differences in water permeability account largely for the magnitude of difference in optimal cooling rates for different cell types (Figure 2). Numerous investigators (Diller 1982; Gao et al. 1994,1996; Levin et al. 1976; McGrath 1985; McGrath and Sherban 1989) have developed devices/modeling to determine cell membrane permeabilities. Devireddy and colleagues (1998) developed a novel differential scanning calorimeter method to determine cell permeabilities during the freezing processes. The rate of water transport across a cell membrane during the freezing process with different cooling rates can be described by four simultaneous equations (Mazur 1963). The first equation relates the loss of cytoplasmic water to the chemical potential gradient between intracellular supercooled water and external ice, expressed as a vapor pressure ratio: dv_l p ART P in, dt V» R where V is the volume of cell water ( am 3 /cell), t is time; L p is the permeability coefficient for water ((irn/min/atm) (hydraulic conductivity); A is the cell surface area ( Jm 2 ); R is the universal gas constant; T is temperature (K); v is the molar volume of water; and P e and P 4 are the vapor pressures of extracellular and intracellular water, respectively. The change in this vapor pressure ratio with temperature can be calculated from the Clausius-Clapeyron equation and Raoult's law, as follows: dln(p c /P i ) = L f dt RT 2 Nv dv (V + Nv o )VdT' Here, N is osmoles of intracellular solute, and L f is the latent heat of fusion of ice. Time and temperature are related by the cooling rate (B), which, if linear, is given by: Finally, L is related to temperature by the Arrhenius relation, dt = L po exp 5s. R IT T. where L is the permeability coefficient of water at a known temperature, T o. E a is the activation energy of L, and R is the universal gas constant. To avoid intracellular freezing, the water content of the cell given by the four equations above must, before reaching the intracellular ice-nucleation temperature (usually between (1) (3) (4) -5 and -40 C), have approached the equilibrium water content given by: V = Y+Nv. RT 273RJ where W { is the initial cell water volume and M ; is the initial osmolality of the extracellular solution. The quantitative expressions above permit one to calculate the extent of supercooling in cells as a function of the cooling rate, provided that one knows or can estimate the permeability coefficient of the cell to water (L *), its temperature coefficient or activation energy (E a J ), the osmoles of solute initially in the cell, and the ratio of the cell surface area to volume. The results of such calculations are commonly expressed as plots of the water content of a cell as a function of temperature relative to the normal or isotonic water volume. In Figure 3, for example, computed water loss curves for mouse ova are shown. The calculated extent to which a cell becomes supercooled is the number of degrees that any given curve is displaced to the right of the equilibrium curve at a given subzero temperature. In the example shown, mouse ova cooled at 4 C/min will be supercooled by 15 C as the temperature passes through -20 C. Three biological parameters have special influence on the position and shape of these curves: the L ; its E a ; and the size of the cell or, more properly, its surface area to volume ratio (S/V 1 ). An increase in L produces an effect similar to a decrease in cooling rate. The effect of cell size is the opposite. An increase in size reduces the cooling rate required to produce a given probability of intracellular freezing. Changes in L and S/V shift the positions of the curves. Changes in the E a have a major effect on the shape of the curves. For instance, the dashed line in Figure 3 represents the effect of changing the E a from 14 to 17 kcal/mol at a cooling rate of 4 C/min. All of the parameters needed to compute the curves can be estimated from permeability measurements. Once computed, these curves can be used to estimate the probability of intracellular freezing as a function of cooling rate. Cells that have dehydrated close to equilibrium before reaching their icenucleation temperature will have a zero probability of undergoing intracellular freezing. Cells that are still extensively supercooled when cooled to their nucleation temperature, and therefore still hydrated, have a high probability of undergoing intracellular freezing. It should be mentioned that in the text above, Mazur's theory and method have been introduced as an example to describe how to predict an optimal cooling rate to prevent IIF. Different parameters and cryobiological properties of cells will be needed to predict optimal cooling conditions to prevent IIF if different physical models are used, as described by Toner and coworkers (1990, 1993) and Muldrew and McGann (1994). (5) 190 WAR Journal

5 8C7min [2C7min;E a =17kcal/mol] TEMPERATURE ( C) Figure 3 Computed kinetics of water loss from mouse ova cooled at 2 to 8 C/min in 1 M dimethylsulfoxide. The curve labeled EQ represents the intracellular water content that the cell must achieve to maintain chemical potential between intra- and extracellular ice, as a function of temperature. The other solid curves, labeled 2, 4, or 8 C/min, were computed assuming the activation energy (E a ) of L to be 14 kcal/mol. Dashed curve reflects the effect of changing E a to 17 kcal/mol. Adapted from Mazur P Freezing of living cells: Mechanisms and implications. Am J Physiol M7(Cell Physiol 16):C125-C142. Preventing Cryoinjury during Slow Freezing As described above, we know that a "slow" cooling process should be used to avoid IIF. Although the avoidance of numerous and large intracellular ice crystals is necessary for cell survival, it is not sufficient. Most cells also require the presence of CPAs. The role of glycerol as an effective CPA for sperm and human erythrocytes was discovered approximately 50 yr ago by a group in Mill Hill, England (Polge and Lovelock 1952; Polge et al. 1949). Glycerol remains one of the more effective and commonly used CPAs to date, although others (e.g., dimethylsulfoxide, ethylene glycol, methanol, propylene glycol, and dimethylacetamide) have emerged since that time and are, in some cases, more effective. The most important characteristics of all of these solutes are that they are, for the most part, readily able to permeate the cells and they are relatively nontoxic to cells in concentrations approaching 1 M or more. The primary basis for their protective action is that they decrease the concentration of electrolytes during freezing and decrease the extent of osmotic shrinkage at a given low temperature (Mazur 1984; Woods et al. 1999). The extent of protection depends primarily on the molar ratio of the CPA to endogenous solutes inside and outside the cells, and the general protective mechanism of action is thus colligative (i.e., dependent only on the concentration of dissolved substances and not on their nature). The effectiveness of a given CPA for a given cell type usually depends on the permeability of that cell to that CPA and its toxicity. Although this class of CPAs must permeate to protect, their permeation before freezing and their removal after thawing generate osmotic volume changes, which alone can be damaging (Gao et al. 1995; Mazur and Schneider 1986; Schneider and Mazur 1988). This topic is discussed further in the following text (Critser and Mobraaten 2000). The second class of CPAs is composed of nonpermeating solutes. This class includes sugars and higher molecular weight compounds such as polyvinylpyrrolidone, hydroxyethyl starch, polyethylene glycols, and dextrans. In most cases, these solutes will not protect in the absence of a permeating CPA but will often substantially augment the effectiveness of a permeating CPA or permit the use of a lower concentration of permeating CPA. These nonpermeating CPAs may contribute to enhance vitrification of the solutions, stabilize proteins and membranes, and prevent progressive ice formation (Fahy 1986; Fahy et al. 1984; Takahashi et al. 1985). Preventing Cryoinjury by Vitrification All discussions above refer to the role of CPAs during slow equilibrium freezing at low enough cooling rates to prevent IIF. A diametrically different approach to cryopreservation is to use either high concentrations of certain CPAs or ultrarapid cooling rates (>10 6 C/min) to induce the cell cytoplasm to form a glass (i.e., to vitrify cells/tissues) rather than to crystallize (Rail and Fahy 1985). However, ultrarapid cooling rates are technically difficult to achieve. Several of the CPAs that are effective in ameliorating slow freeze injury also act to promote glass formation, but the required concentrations are so high (e.g., 4-8 M [Fahy et al. 1984]) that they can be very toxic to the cells/tissues (Fahy 1986; Fahy et al. 1984). Jackson and coworkers (1997) have demonstrated the significant effect of interaction between the microwave and CPA concentration on vitrification of aqueous solutions, Volume 41, Number

6 and they are developing a novel single-mode microwave cavity to enhance the vitrification with relatively low CPA concentrations and relatively low cooling rates. Cryoinjury during the Storage Period Usually, the cryopreserved samples are stored at or below -70 C. Although the question is often asked as to the length of time cells can be kept in the frozen state without damage, the question is probably moot if the storage temperature is less than -120 C and is certainly moot at -196 C (liquid nitrogen). Less than -120 C, chemical reactions cannot occur in human-relevant times. At -196 C, no thermally driven reactions can occur in less than geologically relevant times. The reactions that can occur are the slow accumulation of direct damage from ionizing radiation, but this accumulation would probably become significant only after centuries of storage (Mazur 1984). Cryoinjury Associated with the Warming Process A cell that has survived cooling to low subzero temperatures is still challenged during warming and thawing, which can exert effects on survival comparable with those of cooling (Mazur 1984). These effects depend on whether the prior rate of cooling has induced intracellular freezing or cell dehydration. In the former case, rapid thawing can rescue many cells, possibly because it can prevent the growth of small intracellular ice crystals into harmful large ice crystals (i.e., so-called recrystallization). Even when cells are cooled slowly enough to preclude intracellular freezing, the response to warming rate is often highly dependent on the freezing conditions and cell type and is difficult to predict a priori. For human sperm, previous reports used a wide variety of imprecise methods yielding a wide variety of warming rates. These methods include thawing in ambient air, in water baths with temperatures ranging between 5 and 37 C, in coat pockets, and thawing the samples in the investigator's hands. Mahadevan (1980) conducted a more comprehensive study in which warming rates between 9.2 and 2140 C/ min utilizing coat pocket, hands, air (5, 20, or 37 C), or water baths (5, 20, 37, 55, or 75 C) were examined in combination with a standard freezing rate of 10.0 C/min from +5 to -80 C. The results of that study indicated that slower warming in 20 or 35 C air resulted in more optimal cryosurvival of human sperm in terms of both motility and supravital staining; although there were significant differences in postthaw motilities only between the fastest two rates (1837 and 2140 C/min) and the other 12 rates. Additionally, Mahadevan (1980) examined the interaction between cooling and wanning rates. Slow warming was reported optimal for samples frozen slowly; however, there was little effect of warming rate when a fast cooling rate was used, probably because the cells had been killed by the IIF during the fast cooling process before the warming process started. Osmotic and Potential Toxic Effects of CPAs on the Cell Injury Survival of cells subjected to cryopreservation depends not only on the presence of CPAs but also on the concentration of the CPAs. For example, the percentage of mouse bone marrow stem cells that survive freezing after cooling at the optimal rate increases dramatically as the concentration of glycerol increases from 0.4 to 1.25 M (Leibo et al. 1970). The results are similar for other cells (Mazur 1984). Two important procedures related to the use of permeating CPAs in the cryopreservation of cells are (1) the addition of a CPA to the cells before freezing, and (2) the removal of the CPA from the cells after thawing. Cells transiently shrink when a CPA is added and then return to near-normal volume as the CPA permeates. They undergo transient volume expansion during the removal of the CPA, the magnitude of which depends on how the removal is effected and on the inherent permeability of the cell to water and CPA. In several cell types (e.g., oocytes), the addition of cryoprotectants (dimethylsulfoxide and 1,2-propanediol) has been shown to cause disruption of the cytoskeleton chemically (Johnson and Pickering 1987; Joly et al. 1992). Sperm of many species (e.g., bull, boar, and human) are damaged by exposure to glycerol (Watson 1979). Sherman (1964) reported that glycerol exposure caused a reduction in human sperm motility. Critser et al. (1988b) have demonstrated that exposure of human sperm to 7.5% glycerol for as little as 15 min causes a dramatic reduction in sperm motility, accounting for approximately 50% of the motility loss observed 24 hr after thawing. Historically, bull spermatozoa were exposed to medium containing glycerol for an extended period of time, which was referred to as "equilibration time." However, Berndtson and Foote (1972a,b) reported that glycerol was able to permeate bull sperm at either 25 or 5 C within 3 to 4 min, and they further found that maximum postthaw motility occurred when the sperm were exposed to glycerol for the shortest time measured (10 sec). These data suggest that bull spermatozoa are either highly permeable to glycerol or, unlike other cells, permeation is not required for protection. What then is the role of "equilibration time"? Is it, as some have proposed, a matter of allowing for protective membrane alterations (Watson 1979)? It is important to note that human sperm differ from the sperm of most domestic species and do not show enhanced cryosurvival with increasing delay between addition of CPA and freezing (Sherman 1964). Karow and colleagues (1992) found that the survival of human sperm in vitro at 3 C was significantly enhanced by 20 mm K + in Tyrode's solution relative to Tyrode's solution with less K +. A metabolizable sugar such as glucose was essential to maintain sperm viability in K + -free media. The 192 WAR Journal

7 addition of raffinose to media containing glucose improved motility of sperm stored at 3 C for 6 hr. The detrimental effect of glycerol on sperm can be alleviated to a certain extent by reducing the temperature at which the cryoprotectant is added in both bull and human sperm cells (Critser et al. 1988a,b; Sherman 1973). To what extent the sensitivity of sperm to glycerol reflects the osmotic consequence of its addition and removal, or to what extent it reflects chemical toxicity, is a current subject of investigation. Recently, it has been shown that for human sperm, the adverse effect of glycerol is attributable mainly to the osmotic volume changes induced during addition and removal of glycerol rather than specific chemical action (Gao et al. 1995). In this study, a systematic investigation was also conducted to determine the human sperm volume excursion (shrinkage or expansion) tolerance limits beyond which the cells will be irreversibly injured. The tolerance limits determined for human sperm in combination with the permeability data of cell membrane to water and CPA have been used to develop the optimal multistep procedures for addition and removal of glycerol in sperm to prevent cell osmotic injury (Gao et al. 1995). Cold Shock Injury to Sperm at Low Temperatures Higher Than the Water Freezing Point Currently, most procedures for cryopreservation of sperm, across species, utilize an initial slow cooling rate between body temperature or the temperature at which the sample is collected (often ambient temperature) and +5 C (Foote 1975; Graham 1978; Sherman 1973; Watson 1979). In general, this choice is due to the fact that some species of mammalian spermatozoa are sensitive to temperature changes in this range, with abrupt cooling (a high cooling rate) causing a high frequency of the cells to become irreversibly damaged. This phenomenon has been termed "cold shock" or "chilling" injury (Watson 1979, 1981, 1995). The sensitivity of sperm to cold shock varies among species: Bull, ram, boar, and stallion sperm are highly sensitive; dog and cat sperm are moderately sensitive; and rabbit and human sperm are relatively resistant (Critser etal. 1988b; Watson 1979,1995). The resistance of human sperm to cold shock is of particular importance in the context of developing new approaches to human sperm freezing. It is also an important example of the differences in fundamental cryobiology of cell types among species, highlighting the fact that simple modifications of procedures developed for freezing of domestic animal sperm are unlikely to yield optimal results in other species. Currently, the mechanism of cold shock is unclear. Other Potential Mechanisms of Cryoinjury In many previous studies to date, cells have generally been considered simple compartments of cytoplasmic solution enclosed by a semipermeable plasma membrane. The emphasis has been on how cell survival is related to the physical responses of the whole cell to the physical-chemical events involved in the cryopreservation process. Survival has generally been defined in terms of its viability (e.g., the intactness of cell membrane) and/or the ability of the cells to undergo subsequent growth and development (e.g., hematopoietic stem cells or embryos). Intactness of the plasma membrane is obviously necessary for functional cell survival. However, in many cases (e.g., granulocytes [Armitage and Mazur 1984]), an intact plasma membrane is not sufficient. Findings like the latter emphasized that cells are not simply bags of cytoplasm bound by a plasma membrane. Rather, within the bounds of the plasma membrane are other membranebound structures and organelles essential to cell functions. Little is known about how these intracellular structures and organelles respond to freezing, partly because it is difficult to assay their state and function in situ in an unambiguous fashion (McGann et al. 1988). The spermatozoon has been used as an especially good model for investigating the cryobiological role of both intracellular structures and the plasma membrane because sperm possess two clearly defined and measurable characteristics, motility and the acrosome reaction phenomenon, both of which depend on both functional integrity of organelles and the plasma membrane. An especially attractive feature of the acrosome reaction endpoint in cryobiology is that it involves membrane fusional events (Yudin et al. 1988). Investigation of the effects of freezing on such a reaction may provide an experimental model for the growing view that membrane fusion plays a major role in cryobiological injury (Mazur and Cole 1985, 1989). To be fully functional, sperm must reach the site of fertilization and the oocyte; they must capacitate, undergo the acrosome reaction, penetrate the zona pellucida, and fuse with the plasma membrane of the oocyte. Freezing could interfere with or ablate the capacity of the sperm to undergo one or more of these steps (Critser et al. 1987a,b). Although a motile spermatozoon is not necessarily a fully functional cell, a nonmotile sperm or one exhibiting a damaged plasma membrane is almost certainly nonfunctional (in the context of unassisted fertilization). Consequently, multiple levels of biological function must be considered in assessing the role of the biophysical parameters that have been shown to be critical to cell survival. Summary Although there is great diversity in the cryobiological response of different cell types or given cells among different mammalian species, cryosurvival requires that cell freezing and thawing be carried out within certain biophysical and biological limits defined by the following cryobiology principles: Cells must be frozen in such a way that little or none of their water freezes intracellularly. Volume 41, Number

8 Cells must be warmed in such a way that any unfrozen intracellular water remains unfrozen during warming, or that small ice crystals formed during cooling remain small during warming (Mazur 1984). Even when the conditions described above are met, most cells will not survive unless substantial concentrations of CPAs are present. These CPAs must be introduced before freezing and removed after thawing in ways that do not exceed osmotically tolerable limits. Their concentrations also must not be toxic. Although these general limits are necessary, they may not be sufficient for one or more possible reasons. Cells may be injured by factors such as cold shock that have nothing to do with ice formation or CPA damage. Another reason is that cell viability limits are defined primarily in terms of an intact plasma membrane that retains normal, semipermeable properties. It is possible that conditions that allow the plasma membrane to "survive" may not allow the "survival" of critical organelles inside cells. As reviewed above, with the progress in understanding mechanisms of injury to cells during cryopreservation, various physical modeling and mathematical formulations have been developed in cryobiology research to simulate a cell's response to environmental change during the cryopreservation process and to predict optimal cryopreservation conditions (Mazur 1963, 1990; Muldrew and McGann 1994; Pitt and Steponkus 1989; Toner et al. 1990). The accuracy of these predictive models has been confirmed using data collected on cell types ranging from plant protoplasts to mammalian embryos (Gao et al. 1995; Mazur 1984; Mazur and Schneider 1986; Muldrew and McGann 1994; Pitt and Steponkus 1989; Toner et al. 1990). Application of these models to a given cell type requires the determination of a series of cell cryobiological characteristics, including (1) osmotic behavior of the cells, cell volume and surface area, as well as osmotically inactive intracellular water fraction; (2) temperature dependence of cell membrane permeability coefficients to water and the CPA; (3) temperature dependence of safe volume excursion limits of cells; (4) intracellular ice formation temperatures; and (5) kinetic and thermodynamic heterogeneous nucleation parameters. These characteristics of many cell types have been investigated only recently. Determination of these cryobiological characteristics as well as further understanding the mechanisms of cryoinjury and cryoprotection of CPAs are the subjects of ongoing research in fundamental (i.e., rather than empirical) cryobiology. Other important ongoing research areas in fundamental cryobiology include the investigations of mechanisms and modeling of cryoinjury and cryoprotection to multicellular systems (i.e., tissues and organs) (Bischof et al. 1990; Rubinsky and Pegg 1988) and mechanisms of new potential CPAs such as antifreeze proteins (Rubinsky et al. 1992) and inhibitors of cell apoptosis (Baust et al. 1998). Acknowledgments This work was supported, in part, by grants from the American Heart Association of the Ohio Valley ( V), American Cancer Society (RPG LBC), National Science Foundation (NSF ), Whitaker Foundation (to D.Y.G), and National Institutes of Health (R24-RR and U01-RR to J.K.C.). References Armitage WJ, Mazur P Osmotic tolerance of human granulocytes. Am J Physiol 247(Cell Physiol 16):C373-C381. Baust JM, Hollister W, Van Buskirk R, Baust JG Cryopreservation outcome is enhanced by intracellular-type media and inhibition of apoptosis. Cryobiology 37: Berndtson WE, Foote RH. 1972a. Bovine sperm cell volume at various intervals after addition of glycerol of 5 C. Cryobiology 9: Berndtson WE, Foote RH. 1972b. The freezabilily of spermatozoa after minimal pre-freezing exposure to glycerol or lactose. Cryobiology 9: Bischof J, Hunt CJ, Rubinsky B, Burgess A, Pegg DE The effect of cooling rate and glycerol concentration on the structure of the frozen kidney: Assessment by cryo-scanning electron microscopy. Cryobiology 27: Critser JK, Arneson BW, Aaker DV, Ball GD. 1987a. Cryopreservation of human spermatozoa, n. Post-thaw chronology of motility and of zonafree hamster ova penetration. Fertil Steril 47: Critser JK, Huse-Benda AR, Aaker DV, Arneson BW, Ball OD. 1987b. Cryopreservation of human spermatozoa. I. Effects of holding procedure and seeding on motility, fertilizability, and acrosome reaction. Fertil Steril 47: Critser JK, Arneson BW, Aaker DV, Huse-Benda AR, Ball GD. 1988a. Factors affecting the cryosurvival of mouse two-cell embryos. J Reprod Fertil 82: Critser JK, Huse-Benda AR, Aaker DV, Arneson BW, Ball GD. 1988b. Cryopreservation of human spermatozoa. III. The effect of cryoprotectants on motility. Fertil Steril 50: Critser JK, Mobraaten LE. Cryopreservation of murine spermatozoa. ILAR J 41: Devireddy RV, Raha D, Bischof JC Measurement of water transport during freezing in cell suspensions using a differential scanning calorimeter. Cryobiology 36: Diller KR Quantitative low temperature optical microscopy of biological systems. J Microsc 126:9-28. Diller KR, Cravalho EG A cryomicroscope for the study of freezing and thawing process in biological systems. Cryobiology 7: Fahy GM The relevance of cryoprotectant "toxicity" to cryobiology. Cryobiology 23:1-13. Fahy GM, MacFarlane DR. Angell CA, Meryman HT Vitrification as an approach to cryopreservation. Cryobiology 21: Farrant J Mechanisms of injury and protection in living cells and tissues at low temperatures. In: Smith AU, ed. Current Trends in Cryobiology. New York: Plenum, p Farrant J Water transport and cell survival in cryobiological procedures. Phil Trans R Soc Lond B278: Fishbein WM, Winkert JW Parameters of biological freezing damage in simple solutions: Catalase 11 demonstration of an optimum recovery cooling rate curve in a membraneless system. Cryobiology 15: Foote RH Semen quality from the bull to the freezer: An assessment. Theriogenology 3: ILAR Journal

9 Fujikawa S Freeze-fracture and etching studies on membrane damage on human erythrocytes caused by formation of intracellular ice. Cryobiology 17: Gao DY, Benson CT, Liu C, McGrath JJ, Critser ES, Critser JK Development of a novel microperfusion chamber for determination of cell membrane transport properties. Biophys J 71: Gao DY, Liu C, Benson C, Liu J, Lin S, Critser ES, Critser JK Theoretical and experimental analyses on the optimal experimental design for determination of hydraulic conductivity of cell membrane. In: Hayes L, Roemer RB, eds. Advances in Heat and Mass Transfer in Biological System. Vol New York: ASME Press, p Gao DY, Liu J, Liu C, McGann LE, Watson PF, Kleinhans FW, Mazur P, Critser ES, Critser JK Prevention of osmotic injury to human spermatozoa during addition and removal of glycerol. Hum Reprod 10: Graham EG Fundamentals of the preservation of spermatozoa. In: The Integrity of Frozen Spermatotoa. Washington DC: National Academy Press, p Jackson TH, Ungan A, Critser JK, Gao DY Novel microwave technology for cryopreservation of biomaterials by suppression of apparent ice formation. Cryobiology 34: Johnson MH, Pickering SJ The effects of dimethylsulfoxide on the microtubular system of the mouse oocyte. Development 100: Joly C, Behini O, Boulekbache H, Testart J, Maro B Effects of 1,2- propanediol on the cytoskeletal organizaton of the mouse oocyte. Hum Reprod 7: Karlsson JM, Cravalho EG, Toner M A model of diffusion-limited ice growth inside biological cell during freezing. J Appl Phys 75: Karow AM Jr, Gilbert WB, Black JB Effects of temperature, potassium and sugar on human spermatozoa motility: A cell preservation model from reproductive medicine. Cryobiology 29: Karow AM, Webb WR Tissue freezing: A theory for injury and survival. Cryobiology 2: Leibo SP, Farrant J, Mazur P, Hanna MG, Smith LH Effects of freezing on marrow stem cell suspensions: Interactions of cooling and warming rates in the presence of PVP, sucrose or glycerol. Cryobiology 6: Levin RL, Cravalho EG, Huffins CG A membrane model describing the effect of temperature on the water conductivity of erythrocyte membranes at subzero temperatures. Cryobiology 13: Levitt J A sulfhydryl disulphide hypothesis of frost injury and resistance in plants. J Theor Biol 3:355. Lovelock JE The denaturation of lipid-protein complexes as a cause of damage by freezing. Proc R Soc Lond B 147: Mahadevan M Cryobiological and Biochemical Studies of Human Semen. Ph.D. Thesis. Melbourne, Victoria, Australia: Monash University. Mazur P Kinetics of water loss from cells at subzero temperatures and the likelihood of intraocular freezing. J Gen Physiol 47: Mazur P Freezing and low temperature storage of living cells. In: Muhlbock O, ed. Proceedings of the 1974 Workshop on Basic Aspects of Freeze Preservation of Mouse Strains, Jackson Laboratory, Bar Harbor ME. Stuttgart: Gustav Fisher Verlag. p Mazur P The role of intracellular freezing in the death of cells cooled at supraoptimal rates. Cryobiology 14: Mazur P Freezing of living cells: Mechanisms and implications. Am J Physiol M7(Cell Physiol 16):C125-C142. Mazur P Equilibrium quasi-equilibrium, and nonequilibrium freezing of mammalian embryos. Cell Biophys 17: 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 22: Mazur P, Cole KW Roles of unfrozen fraction, salt concentration, and changes in cell volume in the survival of slowly frozen human erythrocytes. Cryobiology 26:1-29. Mazur P, Leibo SP, Chu EHY A two-factor hypothesis of freezing injury. Exp Cell Res 71: Mazur P, Schneider U Osmotic response of preimplantation mouse and bovine embryos and their cryobiological implications. Cell Biophys 8: McGann LE, Yang H, Walterson M Manifestations of cell damage after freezing and thawing. Cryobiology 25: McGrath JJ Preservation of biological material by freezing and thawing. In: Shitzer A, Eberhart RC, eds. Heat Transfer in Medicine and Biology. New York: Plenum, p McGrath JJ Membrane transport properties. In: McGrath JJ, Diller KR, eds. Low Temperature Biotechnology: Emerging Applications and Engineering Contributions. New York: ASME Press, p McGrath JJ, Sherban KD Determination of the temperature dependence of biomembrane-passive transport using a microdiffusion chamber. In: Roemer RB, McGrath JJ, Bowman HF, eds. Bioheat Transfer Applications in Hyperthermia, Emerging Horizons in Instrumentation and Modeling. New York: American Society of Mechanical Engineers, Heat Transfer Division (Vol 126) and Biomedical Engineering Division (Vol 12). p Meryman HT The exceeding of a minimum tolerable cell volume in hypertonic suspension as a cause of freezing injury. In: Wolstenhoime OEW, O'Connor M, eds. The Frozen Cell. Ciba Foundation Symposium. London: Churchill, p Meryman, HT Freezing injury and its prevention in living cells. In: Mullins LJ, Hagins WA, Stryer L, Newton C, eds. Annual Review of Biophysics and Bioengineering. Palo Alto: Annual Reviews, p Muldrew K, McGann LE The osmotic rupture hypothesis of intracellular freezing injury. Biophysics 166: Pegg DE, Diaper MP The mechanism of injury to slowly-frozen erythrocytes. Biophys J 54: Pitt RE, Steponkus PL Quantitative analysis of the probability of intracellular ice formation during freezing of isolated protoplasts. Cryobiology 24: Polge C, Lovelock JE Preservation of bull sperm at -70 C. Vet Rec 64: Polge C, Smith AU, Parkes AS Revival of spermatozoa after vitrification and dehydration at low temperatures. Nature 164: Rail WF, Fahy GM Ice-free cryopreservation of mouse embryos at -196 C by vitrification. Nature 313: Rapatz G, Nath J, Luyet B Electron microscope study of erythrocytes in rapidly frozen mammalian blood. Biodynamica 9: Rubinsky B, Arav A, DeVries AL The cryoprotective effect of antifreeze glycopeptides from Antarctic fishes. Cryobiology 229: Rubinsky B, Pegg D A mathematical model for the freezing process in biological tissue. Proc R Soc Lond 234: Schneider U, Mazur P Osmotic consequences of cryoprotectant permeability and its relation to the survival of frozen-thawed embryos. Theriogenology 21: Sherman JK Improved methods of preservation of human spermatozoa by freezing and freezing-drying. Fertil Steril 14: Sherman JK Synopsis of the use of frozen human semen since 1964: State of the art of human semen banking. Fertil Steril 24: Steponkus PL, Wiest SC Freeze-thaw induced lesions in the plasma membrane. In: Lyons JM, Graham DG, Raison JK, eds. Temperature Stress in Crop Plants: The Role of the Membrane. New York: Academic Press, p Stowell RE, Young DE, Arnold EA, Tromp BF Structural, chemical and functional alterations in mammalian nucleus following different conditions of freezing, storage and thawing. FASEB Fed Proc 24(Suppl 15):S115-S141. Takahashi T, Inada S, Pommier CG, O'Shea JJ, Brown EJ Osmotic stress and the freeze-thaw cycle cause shedding of Fc and C3b receptors by human polymorphonuclear leukocytes. J Immunol 134:4062. Volume 41, Number

10 Toner M, Cravalho EG, Karel M Thermodynamics and kinetics of intracellular ice formation during freezing of biological cells. J Appl Phys 67: Toner M, Cravalho EG, Stachecki J, Fitzgerald T, Tompkins RG, Yarmuch ML, Armant DR Nonequilibrium freezing of one-cell mouse embryos. Biophys J 64: Trump BF, Young DF, Arnold EA, Stowell RE Effects of freezing and thawing on the structure, chemical constitution and function of cytoplasmic structures. FASEB Fed Proc 24(Suppl 15):S144-S167. Watson PF The preservation of semen in mammals. In: Finn CA, ed. Oxford Reviews of Reproductive Biology. London: Oxford University Press, p Watson PF The effects of cold shock on sperm cell membranes. In: Morris OJ, Clark A, eds. Effects of Low Temperatures on Biological Membranes. New York: Academic Press, p Watson PF Recent developments and concepts in the cryopreservation of spermatozoa and the assessment of their post-thawing function. Reprod Fertil Dev 7: Williams RJ The mechanisms of cryoprotection in the intestinal mollusk Mylilus. Cryobiology 4: Woods E, Zieger M, Gao DY, Critser JK Equations for obtaining melting points for the ternary system ethylene glycol/sodium chloride/ water and their application to cryopreservation. Cryobiology 38: Yudin AL, Gottlieb W, Meizel S Ultrastructural studies of the early events of the human sperm acrosome reaction as initiated by human follicular fluid. Gamete Res 20: ILAR Journal