Introduction

Cryoprotectant toxicity should be distinguished from other mechanisms of cryopreservation injury such as chilling injury (injury produced by too low temperatures as such) and cold shock  (injury produced by rapid cooling). Cryoprotectant toxicity itself can again be divided into general cryoprotectant toxicity and specific cryoprotectant toxicity. General cryoprotectant toxicity involves concentration (water substitution) effects of cryoprotectants and specific cryoprotectant toxicity involves the effects of individual compounds on cellular viability. General cryoprotectant toxicity presents a formidable obstacle for cryopreservation methods that require very high concentrations of cryoprotectant agents (such as vitrification).

Another mechanism of injury that is rarely discussed in the cryobiology literature but that can complicate cryopreservation of complex organs is “non-specific” dehydration injury. In light of the fact that the current generation of vitrification agents are delivered in hypertonic carrier solutions and contain non-penatrating cryoprotective agents which do not cross the blood brain barrier, this form of damage may be especially important in cryopreservation of the brain.

Systemic reviews of cryoprotectant toxicity are rare but some mechanisms for (specific) cryoprotectant toxicity have been proposed including, but not limited to, protein denaturation, modification of biomolecules, membrane injury, destabilization of the cytoskeleton, oxidative damage, and ATP depletion. It is important to stress that some of the mechanisms may be downstream effects of other mechanisms. For example, ATP depletion can cause oxidative damage. And as Gregory Fahy has pointed out, cryoprotectant toxicity should be distinguished from injury associated with the method of introduction and washout of the cryoprotectant. In 2004, Fahy, Wowk et al., proposed a compositional variable to predict general cryoprotectant toxicity.

Cryoprotectant toxicity can also vary by species and organ type. Cryoprotectants that are moderately toxic in one species can be highly toxic in others. Similarly, cryoprotectants that are moderately toxic in one organ can be highly toxic in others (or even between different types of cells within organs). This raises the question of whether universal non-toxic cryoprotective agents are attainable (a requirement for reversible vitrification in complex organisms).

Cryoprotectant toxicty can be investigated by cryopreserving an organ (or cell) and measuring its viability after rewarming and washout of the cryoprotective agent. To eliminate the influence of other mechanisms of injury associated with cryopreservation (such as ice formation), a cell can just be loaded and unloaded with the cryoprotectant without cryopreservation. The effects of hypothermia on viability can be eliminated altogether by normothermic perfusion of the organ. This, of course,  introduces a challenge for hypoxia sensitive organs such as the heart and the brain because cryoprotective agents may not be good oxygen carriers.

Papers

Baxter SJ, Lathe GH (1971). Biochemical effects of kidney of exposure to high concentrations of dimethyl sulphoxide.
Biochemical Pharmacology. Jun; 20(6): 1079-91.

Baxter and Lathe investigated the effect of high concentrations of DMSO on kidney preparations. In a series of illuminating experiments, the investigators established that anaerobic glycolysis was reduced in slices and homogenates as a result of increased activation of the gluconeogenesis enzyme Fructose 1,6-diphosphatase (FDPase). DMSO-induced activation of FDPase can be inhibited by adding an amide or lysine to DMSO. The finding that a combination of DMSO and an amide allows for less toxic cryoprotectants formed the basis of subsequent investigations of GM Fahy for potent vitrification solutions.

Fahy GM (1983). Cryoprotectant Toxicity Neutralizers Reduce Freezing Damage.
Cryo-Letters 4: 309-314.

In this paper GM Fahy reports the ability of toxicity neutralizers urea, formamide, and acetamide (all amides) to reduce injury of cryopreserved renal cortical slices with DMSO. In later research papers Fahy will establish that DMSO neutralizes the toxicity of formamide, and not the other way around.

Fahy GM (1984). Cryoprotectant toxicity: biochemical or osmotic?
Cryo-Letters 5: 79-90.

If osmotic stress is an important cause of injury during introduction and removal of cryoprotectant agents, improved viability can be obtained by reducing the rate of cryoprotective agent introduction and removal. Fahy reviews the literature and presents data obtained in renal cortical slices that indicate that substantial hypertonic osmotic stress does not produce major changes in viability. Conversely, reducing exposure time to higher concentrations of the cryoprotectant can contribute to improved viability. These results suggest that biochemical toxicity, not osmotic stress, is the major factor in cryoprotectant-induced injury.

Fahy GM (1984). Cryoprotectant toxicity: specific or non-specific?
Cryo-Letters 5: 287-294

Fahy reviews the argument (Morris, Cryoletters 4, 339-340, 1983) that the lower toxity of cryoprotectant solutions that contain DMSO and amides can be entirely explained by the lower absolute concentration of DMSO. Fahy points out that the original Bexter and Lathe experiments demonstrated that solutions with the same absolute amount of DMSO (4.6 M) but with or without amides had different effects on glucose utilization. The author also presents data showing that “simple substitution (“dilution”) of one agent for another strikingly fails to reduce overall toxicity over a very critical range of DMSO concentration.” Also briefly discussed is the possibility of mutual toxicity neutralization between DMSO and amides, a topic that would be further explored by Fahy in future research.

Fahy GM, MacFarlane DR, Angell CA, Meryman HT (1984). Vitrification as an approach to cryopreservation.
Cryobiology.  Aug ; 21(4): 407-26.

In this paper on vitrification as an alternative to conventional cryoprotection, Fahy et al., list a number of methods for reducing cryoprotectant toxicity:

Primary (direct) methods:

1. Maintain temperature as low as possible;
2. Select an appropriate carrier solution;
3. Keep exposure time at higher concentrations to a minimum;
4. When possible, employ specific cryoprotectant toxicity neutralizers.

Secondary (indirect) methods:

1. Avoid osmotic injury;
2. Mutual dilution of cryoprotectants may be helpful in some instances;
3. Use extracellular cryoprotectant to reduce exposure to intracellular cryoprotectant when possible.

The most important insights, some of which are still maintained in the current generation of vitrification solutions, concern toxicity neutralization, the choice of an appropriate carrier solution, and the use of extracellular cryoprotectants.

Fahy GM (1986). The relevance of cryoprotectant “toxicity” to cryobiology.
Cryobiology. Feb; 23(1) :1-13.

Fahy presents evidence that cryoprotectants themselves can present a source of injury. As a consequence, the advantages of higher concentrations of the cryoprotective agents does not necessarily produce higher viability after freezing, even when this allows for greater ice inhibition. He reviews data on “cryoprotectant-associated freezing injury” for DMSO, ethylene glycol, methanol, ethanol, and glycerol.  Because vitrification requires very high concentrations of cryoprotective agents, toxicity is the key limiting factor in reversible vitrification of organs.

Fahy GM, Lilley TH, Linsdell H, Douglas MS, Meryman HT (1990). Cryoprotectant toxicity and cryoprotectant toxicity reduction: in search of molecular mechanisms.
Cryobiology. Jun; 27(3): 247-68.

Fah,y et al., delineate 6 criteria that must all be met simultaneously in order for a putative mechanism of cryoprotectant toxicity to be implicated:

1. The relationship between observed biochemical alteration and cellular viability must be clear or easily plausible;
2. The maginitude of the cryoprotectant effect must be large enough to be significant;
3. The effect must be irreversible over a reasonable time span after removal of the cryoprotectant;
4. The time course of the observed effect must be consistent with the time course of observed injury;
5. The cryoprotectant effect must be possible under conditions that could reasonably be encountered inside a living cell being prepared for freezing or being subjected to freezing and thawing itself;
6. The cryoprotectant effect must be due to the cryoprotectant itself and not due to the technique of introduction and washout.

The authors investigate the proposed mechanisms for the biochemical effects of DMSO toxicity in the 1971 Baxter study and find that a) the effect of DMSO on FDPase activation is too small to affect the normal respiration of the cell and therefore fails to meet criterion 2 to be a significant mechanism of cryoprotectant toxicity; b) the presence of formamide does not affect the interaction between DMSO and lysine; and c) toxicity is not consistently reduced by blocking alteration of FDPase rather than substituting those compounds for DMSO.

The authors further present results that do not support the theory that generalized  protein denaturation is related to cryoprotectant toxicity.  The article ends with a referenced list of phenomena possibly related to mechanisms of cryoprotectant toxicity.

Fahy GM, da Mouta C, Tsonev L, Khirabadi BS, Mehl P,  Meryman HT (1995). Cellular injury associated with organ cryopreservation: Chemical toxicity and cooling injury.
Editors: John J. Lemasters, Constance Oliver. Cell Biology of Trauma, CRC Press

Fahy, et al., review different mechanisms of cryoprotectant toxicity with a particular focus on DMSO-medicated chemical injury. Mechanisms discussed include fructose-1,6-bisphosphatase activation, sulfhydryl oxidation, activation of extracellular proteinases and endothelial cell detachment and death. The article lists a number of interventions that do not change CPA-medicated injury such as inhibition calcium mediated injury or protein denaturation. The authors also report how the toxicity of formamide can be completely reversed by addition of DMSO.

Bakaltcheva IB,  Odeyale CO, Spargo BJ (1996). Effects of alkanols, alkanediols and glycerol on red blood cell shape and hemolysis.
Biochimica et Biophysica Acta. 1280: 73-80

In this elegant and thoughtful paper, the authors use the human red blood cell to study cryoprotectant toxicity. Morphological observations, quantification of hemolysis, measurements of the dielectric constant of the incubation medium (Ds) and the dielectric constant of the erythrocyte membrane in the presence of organic solutes (Dm), are used to investigate cryoprotectant toxicity in a series of alkanols, alkanediols, and glycerol. The authors propose that toxicity of a cryoprotectant is related to its ability to change the ratio of Ds/Dm. Changes in this ratio reflect changes in the difference between hydrophobicity of the solution and the membrane, with decreases in this ratio leading to increased exposure of membrane surface area and vesiculation, and increases in this ratio leading to decreased exposure of membrane surface area and cell fusion. The authors suggest that the design of less toxic cryoprotective agents should involve the maintenance of dielectric homeostasis of the medium and the membrane. Their findings also throw light on the observation that combinations of various cryoprotectant agents (such as DMSO and formamide) can reduce the overall toxicity of a solution.

Fahy GM, Wowk B, Wu J, Paynter S (2004). Improved vitrification solutions based on the predictability of vitrification solution toxicity.
Cryobiology. Feb; 48(1): 22-35.

This seminal paper on non-specific cryoprotectant toxicity represents a major contribution to the cryobiology literature in general, and enabled the authors to formulate less toxic vitrification solutions for the cryopreservation of whole organs. In the paper the authors propose a new compositional variable that reflects the strength of water-cryoprotectant hydrogen bonding called qv*. Contrary to the cryobiology wisdom to date, the authors found that weaker glass formers favor higher viability. As a consequence, vitrification agents with higher concentrations of cryoprotective agents are not necessarily more toxic. Although qv* is not helpful in predicting specific cryoprotectant toxicity, this paper, and the research that is reflected in it, suggests that non-specific cryoprotectant toxicity is mediated through the effects of penetrating cryoprotectant agents on the hydration of biomolecules.