- Email: [email protected]
Effects of cryobiological variables on the survival of skin using a defined murine model
CRYOBIOLOGY
27, 164-170 (MO)
Effects of Cryobiological J. N. KEARNEY,*
Variables on the Survival Defined Murine Model L. A. WHELDON,*
AND
of Skin Using a
G. GOWLANDt
*Yorkshire Regional Tissue Bank, Pinderfields General Hospital, Wakefield; and tDepartment of Immwrology, University of Leeds, Leeds, United Kingdom
Skin from an inbred strain of hairless mouse was used as a homogeneous model tissue for studies of skin cryopreservation. Tetrazolium reductase enzyme activity was used to assess tissue viability. Hepes-buffered 199 tissue culture medium was confirmed to be a suitable basal medium, to which cryoprotectants were added. Addition of serum to the cryoprotective cocktail had no beneficial effect. There was no Three cryoprotectants, dimethyl sulfoxidc, ethanediol, and glycerol were ewhdkd. evidence of specific toxicity attributable to the cryoprotective agents during the permeation period; however, short permeation times at low temperature were associated with maximum skin viability. Following freezing and thawing, higher viabilities were obtained when using a slow ( - 1°C min- ‘) or medium (-M”C min- ‘) rather than a fast (immersion in liquid nitrogen) cooling rate. Dimethyl sulfoxide was a marginally better cryoprotectant overall, although this difference was not statistically significant. 0 1990 Academic Press, Inc.
Cryopreservation methods for the storage of allograft skin are desirable in view of the unpredictability of donor availability and patient requirements, thus necessitating a long-term skin storage technique. This becomes even more important if major histocompatibility complex (MHC) tissue matching is adopted. For autograft skin, the potential benefits of cryopreserving “excess” harvested grafts are that viable skin remains available for areas of the burn not yet ready to take grafts and that any failed grafts may be replaced at the bedside rather than requiring further operating theatre sessions to obtain additional skin. This potential benefit can only be realized if the cryopreservation technique is reproducible and retains a high cell viability on each occasion. Many different protocols have been advocated for the freezing of human skin (2,3, 9, 10, 11, 13, 15, 16). We have evaluated a number of these protocols and have been Received April 6, 1989; accepted June 25, 1989.
001l-2240/90 $3.00 Copy~@t All rights
0 1590 by Academic Press, Inc. of reproduction in any form reserved.
disappointed with the variability in clinical graft success and other viability measurements, both within and between skin samples (unpublished data). These evaluations could only be carried out when harvested skin was found to be in excess of clinical requirements. Because of this heterogeneity between skin samples and the influence of other variables associated with the recipient (e.g., wound bed vascularity and microbiological flora), it was impossible to adequately assess the effects of changes in the cryopreservation protocol on grafl success or graft viability. To achieve this, we have developed a defined inbred murine skin model (8), thus avoiding variability due to heterogeneous skin samples. A simple prognostic viability index, the measurement of tetrazolium reductase (TR) activity has been shown to correlate with graft success and oxygen consumption (8) and was therefore used to optimize aspects of the cryopreservation protocol. This is the first study of skin cryopreservation to use genetically homogeneous skin in a number of large factorial design experiments.
CRYOPRESERVATION
MATERIALS
AND
METHODS
Inbred Swiss Albino hairless mice were used as skin donors. The mice were bred continuously within the Department of Immunology animal house and fed an ad libiturn diet. Only female mice 34 months of age were used in order to minimize variability (8). The preparation of full thickness trunk skin has been described elsewhere (8). Reagents
The following reagents were used: 199 tissue culture medium containing 25 mM Hepes buffer (Gibco, Paisley, Scotland); Dulbecco’s phosphate-buffered saline (PBS; Oxoid, England); dimethyl sulfoxide (Me,SO), ethanediol, glycerol (Analar grade, BDH, Poole, England): and 0.9% saline (Steriflex, Boots, Nottingham, England). All other chemicals were obtained from Sigma (Poole, England).
OF MURINE
SKIN
165
device (Spembly Technical Products, Ltd., Sittingbourne, England). Medium cooling (- 60°C min- ‘) was achieved by lowering the nylon packet (long axis horizontal) to a predetermined position (usually 2.5 cm) above the level of liquid nitrogen in a liquid nitrogen refrigerator (Taylor-Wharton 3K, Jencons, England), Fast cooling was achieved by plunging the packets directly into liquid nitrogen. In each case nucleation was spontaneous. With the first two cooling methods, on reaching - lOO”C, the packets were transferred rapidly into liquid nitrogen for subsequent storage (- 196°C). Factors Evaluated
Basal medium. Ten replicate skin pieces were soaked for 2 hr at 4°C in each of the following basal solutions: 0.9% saline, PBS, Hepes-buffered 199 medium, Hepesbuffered 199 medium containing 15% (v/v) glycerol. Standard nylon packets containing the skin were subsequently heat sealed and cooled in the vapor phase of liquid nitrogen at - 60°C min-‘. After 24 hr, the Assays packets were thawed rapidly (immersion in TR activity was measured using Carney’s a 37°C water bath) and the skin was assayed (4) modification of Hershey’s (7) proce- immediately for TR activity. Serum. Skin pieces (IO replicates per dure, as described previously (8). For skin freezing experiments between 5 and 10 treatment) were soaked for 2 hr at 4°C in pieces of full thickness trunk skin 0.5-l cm2 Hepes-buffered 199 medium containing in size were placed into standard nylon 15% (v/v) glycerol and 0, 10, or 50% fetal calf serum. The nylon packets were subse(Jencons, Leighton Buzzard, England) packets (13 x 2.5 cm) on strips of sterile quently cooled at either a slow (- 1°C gauze. Two milliliters of cooled (4°C) cryo- min-‘) or a medium (-60°C min-‘) rate. protectant solution was added and the After 24 hr of storage in liquid nitrogen the packet was heat sealed. Various perme- skin was thawed rapidly and assayed for ation temperatures and permeation periods TR activity. Cryoprotectantlpermeation time, A were used. Cooling rates were determined three-way factorial experiment was carried using an electronic thermometer (Type K, CP Instrument Co.) attached to a chart re- out to assess the effects of cryoprotectants corder (Datatrace, Gallenkamp, Loughbor- (15% (v/v) and 25% (v/v) glycerol, Me,SO, and ethanediol in I99 medium, and a 199 ough, England). Type K thermocouples were inserted into the skin tissue. Three medium control), permeation temperature cooling rates were used in these experi- (4, 23, and 37”(Z), and permeation time (15 ments. Slow cooling (- 1°C min- ‘) was min, 2 hr, and 24 hr). Five replicate skin achieved using a controlled rate freezing pieces were used for each combination of
166
KEARNEY,
WHELDON,
factors. Following treatment the skin was assayed for TR activity, Cooling
ratelcryoprotectantipertneatioa
time. Three experiments were carried out using the following cooling rates: 1°C mm-’ (slow), -60°C min-’ (medium), and direct immersion in liquid nitrogen (fast). Each experiment was a two-way factorial design to assess the effects of cryoprotectants (15% (v/v), 25% (v/v), and 50% (v/ v) glycerol, Me,SO, and ethanediol in 199 medium, and a 199 medium control) and permeation time at 4°C (15 min and 2 hr). After cooling to at least - 100°C the skin packets were transferred to liquid nitrogen. Following storage for at least 24 hr, the skin was thawed rapidly (immersion in 37°C water bath) and immediately assayed for TR activity. Ten replicate skin pieces were used for each factor combination.
AND GOWLAND TABLE 1 Effects of Basal Media on TR Activity after Freezing (- 60°C min- ‘) and Thawing: (a) Analysis of Variance and (b) Comparison of Subgroup Means Using the T-Method (a)
df
MS
F
P
Media Within (error)
4 45
3876.44 66.87
57.97
Minimal significant difference = 13.5% (b)
Control
199 + glycerol
, 104% ,
,9X%+
199 PBS
1"
42%
Saline
37%,
Note. Underlined means are not significantly different at P < 0.05.
pared with saline, although these differences were not statistically significant at P < 0.05.
Statistical
Analysis
Addition of serum to the cryoprotective cocktail did not increase the viability of the Data from the factorial experiments was frozen skin, in fact the opposite was obsubjected to analysis of variance. A posteserved at the higher serum concentration riot-i multiple comparisons of means were (Table 2). There was no difference in the carried out using the T-method (14). All staviability of skin cooled at a slow or a metistical analyses were performed using the dium rate. Under slow cooling conditions, TR enzyme activity data expressed as abthere was a statistically significant decrease sorbance at 490 nm (X 103) . mg skin-t * ml in viability with increasing percentages of solvent. Following analysis the data were serum, while under medium cooling conditransformed to percentages, taking the origtions there was not. inal fresh skin enzyme activity as 100%. Data from the three-way factorial experiment to evaluate the effects of cryoproRESULTS tectant medium, permeation temperature, Following freezing and thawing, a signif- and permeation time on skin viability reicant decrease in viability was observed re- sulted in significant F values for two of the gardless of the medium used, compared factors, temperature and time (P < 0.05), with the enzyme activity of fresh skin (Ta- when compared with the pooled error and ble 1). A combination of glycerol with 199 interaction (temperature X time) mean medium gave the highest post-thaw viabil- squares. The third factor, media, had no ity. In the absence of a specific cryoprotec- significant effect. Short permeation times at tive agent, there was a trend toward a 4°C were superior to the other combinahigher retained viability for the media pos- tions (Fig. 1). sessing buffering capacity (PBS and 199) The mean percentage enzyme activities and additional beneficial effects of a com- of skin subjected to three cooling rates, plex nutritional environment (199) com- with cryoprotectants at three concentra-
CRYOPRESERVATION
ISmtn
TABLE 2 Effects of Serum on TR Activity after Freezing by Slow and Medium Cooling and Thawing: (a) Two-Way Analysis of Variance and (b) Comparison of Subgroup Means Using the T-Method (a)
df
MS
F
P
Cooling rate (A) Percentage serum (S)
1 2 2 54
137.92 910.94 35.93 129.61
1.06 7.03 0.28
N.S. co.05 N.S.
AxB
Within (error)
0))
0% serum
10% serum
50% serum
Slow cooling
65.7%
51.4%
49.3%
1 Medium cooling
I N.S. 61.‘s%
10(1bOM)% IOzoo-
Minimal significant difference = 15.3%
’
167
OF MURINE SKIN
N’S 51.
%$i N:S. 54.2%
l. :.
100-
:
M-
; cd2 % 402 a zLOc $
o-
,oo- z44hr
1 Nore. Underlined means are not significantly different at P < 0.05.
no%
bO-
40zo-
tions and two permeation times, are presented in Fig, 2, All three cryoprotectants provided some degree of protection compared with the 199 medium control. Following short permeation times the use of Me,SO frequently resulted in higher viability compared with the other two cryoprotectants, but this was statistically significant only for certain comparisons (e.g., following slow cooling with a 15% (v/v) concentration of the three cryoprotectants, (Fig. 2A). Comparing permeation times, shorter permeation appeared generally superior and was statistically significant at the higher cryoprotectant concentrations. Slow and medium cooling rates resulted in higher viability than rapid cooling. DISCUSSION
Because all of the skin used in this study was genetically homogeneous, large factorial experiments could be undertaken using pooled skin taken from a number of donors. In this way the “main effects” and
0-
IW
15% $$
25w GIV/VI in 199
15% %
a* t%
15* I Iv/4 in I*
25% ;;;g
FIG. 1. Effect of permeation temperature (E3, 4°C; 0, 23°C; n , 37°C) and permeation time of cryoprotectant (G, glycerol; D, Me,SO; E, ethanediol)containing medium on mean percentage TR activity (bars = ti minimal significant difference). Statistically significant differences occur where bars do not over-
“interaction” between variables in the cryopreservation procotol were assessed independently of the variation normaliy found in outbred populations. To study the effects of different cryoprotective agents and additives, a basal salt medium was required. Saline has been used for clinical skin storage, although its lack of buffering capacity could lead to a significant lowering of pH during the cryopreservation process when storing tissues that have a large population of metabolizing cells. It is therefore not surprising that saline performed worse than the other salt-
168 KU-J-A
KEARNEY,
WHELDON,
AND
GOWLAND
improved the residual viability of the skin. From this experiment it was decided to use 199 medium as the basal medium for the remainder of the study. The addition of fetal calf serum to 199/ glycerol medium had no beneficial effect at either slow or medium cooling rates. At the higher serum concentrations (50%) a significant reduction in viability was seen comJOQ-8 pared with the control. Using guinea pig ear T T skin, Lawrence (9) found no consistent improvement in viability after cryopreservation with 5-50% fetal calf serum. In the clinical situation these results are fortuitous in view of the potential for immune responses to xenogeneic proteins present in fetal calf serum or for problems associated with the acquisition of safe homologous serum or isogroup plasma as recommended by Blondet (3). A variety of cryoprotectant concentrations, permeation temperatures, and permeation times have been advocated for subsequent cryopreservation of skin with little evidence of systematic investigation. These include 2% Me,SO for 15 min at 20°C followed by 10% Me,SO at 10°C (3), 15% Fro. 2. Effect of cryoprotectants and permeation glycerol at 4°C for 2 hr (1 l), or 20% glycerol times (I3 = 15 min; 0 = 2 hr) on TR activity following freezing and thawing, using (A) slow, (B) medium, and at 0°C for 2-4 hr (6), and many others (2, 13, (C) fast cooling rates (bars = L/2minimal significant 15, 16). The optimal combination of these difference). Statistically significant differences occur factors depends on the temperature/ where bars do not overlap. time/toxicity interactions of the various cryoprotectants on skin cells, and the difcontaining media. Phosphate can be used to fusion and permeability characteristics of buffer saline (PBS); however, phosphate skin cells embedded in their tissue matrix, has a poor buffering capacity above a pH of toward each type of cryoprotectant. For 7.5 and tends to precipitate many polyva- mouse skin, the main effects and interaclent cations. The 199 medium retained tions of these variables on skin viability slightly higher viability and in this case were assessed by a factorial experiment. buffering capacity was provided by Hepes There were significant differences between which is known to be efficient even at low permeation temperatures and times, but temperatures (12). The use of 25 mA4Hepes surprisingly no differences between cryoin tissue culture medium for the hypother- protectants compared with 199 medium. mic (4°C) storage of skin has previously Therefore, in contrast to Lawrence’s findbeen reported by Cram et al. (5), where it ings for guinea pig skin, where both glycdemonstrated improved buffering capacity erol and Me,SO (15% in saline) reduced without any detrimental effects. Addition skin viability at room temperature (9), the of glycerol to the 199 medium significantly cryoprotectants did not decrease the tetra-
CRYOPRESERVATION
zolium reductase activity of murine skin at any of the temperatures tested. Whether the cryoprotectants caused sublethal damage predisposing the skin to freeze injury when used in high concentrations was assessedin the subsequent experiment. In this case, skin subjected to combinations of permeation time and cryoprotectant concentrations was cooled at three different rates. In the previous experiment, a permeation time of 24 hr was highly detrimental to the skin and was therefore omitted from this experiment, whereas an extra cryoprotectant concentration (50%) was included, and a low permeation temperature (4°C) was used. All three cryoprotective agents gave viabilities higher than those of the 199 medium control. However, even in the absence of a cryoprotective agent, skin cryopreserved in 199 medium retained a significant proportion of TR activity, particularly following a short permeation period. Similarly, short permeation times were generally superior for media containing cryoprotectants including glycerol, which has been reported to require longer permeation periods for many cell types to achieve a satisfactory cryoprotective concentration intracellularly (1). For Me,SO, the large decrease in viability following a 2-hr permeation time at the highest concentration (50%) may indicate cytotoxicity; however, this would need to be confirmed by further experiments. For the other cryoprotectants, and for all cryoprotectants following short permeation times, there was no evidence that higher concentrations of cryoprotectants either directly or indirectly reduced skin viability. The viabilities obtained following fast cooling were lower than those for the other two cooling rates. There was, however, no difference between slow (- 1°C min-‘) and medium (-60°C min- ‘) cooling. Therefore, the commonly recommended cooling rate of - 1°C min-‘, which can only be accurately achieved with sophisticated cooling devices, is not essential to the maxi-
OF MURINE
SKIN
169
mum attainable viability of skin following cryopreservation, and this would explain why other simpler cooling methods have yielded acceptable results for both human and animal skin (9, 10). This observation has been reported previously for human skin by several authors (see review in Ref. (2)), suggesting that the results obtained using this murine model may be applicable to the human clinical situation. ACKNOWLEDGMENT
We thank the Yorkshire Regional Health Authority for supporting this work. REFERENCES
1. Bank, H. L., and Brockbank, K. G. M. Basic principles of cryobiology. J. Cardiac Surg. l(3), 137-143 (1987). 2. Baxter, C., Aggarwal, S., and Diller, K. R. Cryopreservation of skin: A review. Transplant. Pruc. 17(6), 112-120 (1985). 3. Blondet, R., Gibert-Thevenin, M. A., Pierre, C., and Ehrsam, A. Skin preservation by programmed freezing. &it. J. Plast. Surg. 35,53& 536 (1982). 4. Carney, S. A., Hall, M., and Ricketts, C. R. The succinic dehydrogenase and cytochrome c oxidase activities of guinea-pig skin after mild heat damage. &it. J. Dermatol. 94, 295-299 (1976). 5. Cram, A. E., Domayer, M. A., and Scupham, R. Preservation of human skin: A study of two media using the athymic (nude) mouse model. J. Trauma 25, 128-130 (1985). 6. Graham, W. P., Hamilton, R. W., and Lehr, H. B. Versatility of skin allografts: Desirability of a viable frozen tissue bank. J. Trauma 11, 494-501 (1971). 7. Hershey, F. B., Cruickshank, C. N. D., and Mullins, L. 1. The quantitative reduction of 2,3,5triphenyl tetrazolium chloride by skin in vitro. J. Hisrochem. Cyruchem. 6, 191-l% (1957). 8. Keamey, J. N., Wheldon, L. A., and Gowiand, G. Cryopreservation of skin using a murine model: Validation of a prognostic viability assay. Cvobiology, 27, 24-30 (1990). 9. Lawrence, J. C. Storage and skin metabolism, &it. J. Plast. Surg. 25, 440453 (1972). 10. May, S. R., Guttman, R. M., and Wainwright, J. F. Cryopreservation of skin using an insulated heat sink box stored at -70°C. Cryobiology 22, 205-214 (1985). 11. Ninneman, J. L., Fisher, J. C., and Frank, H. A. Clinical skin banking: A simplified system for
170
KEARNEY,
WHELDON,
processing, storage and retrieval of human allografts. J. Trauma 18, 723-725 (1978). 12. Pegg, D. E. Perfusion technology. In “Organ Preservation for Transplantation” (A. M. brow and D. Pegg, Ed.) 2nd ed., pp. 477495. Dekker, New York, 1981. 13. Perry, V. P. A review of skin preservation. Clyobiology 3, lO!M30 (1966). 14. Sokai, R. R., and Rohlf, F. J. “Biometry: The
AND GOWLAND Principles and Practise of Statistics in Biological Research.” Freeman, San Francisco, 1981. 15. Volkova, N. A., and Sandomirsky, B. P. Organ culture of human skin preserved by freezing. Cryo-Lett.
3, 227-238 (1982).
16. Wachtel, T. L., Ninnemann, J. L., Fisher, J. C., Frank, H. A., and Inancsi, W. Viability of frozen allografts. Amer. J. Surg. 138, 783-787 (1979).
27, 164-170 (MO)
Effects of Cryobiological J. N. KEARNEY,*
Variables on the Survival Defined Murine Model L. A. WHELDON,*
AND
of Skin Using a
G. GOWLANDt
*Yorkshire Regional Tissue Bank, Pinderfields General Hospital, Wakefield; and tDepartment of Immwrology, University of Leeds, Leeds, United Kingdom
Skin from an inbred strain of hairless mouse was used as a homogeneous model tissue for studies of skin cryopreservation. Tetrazolium reductase enzyme activity was used to assess tissue viability. Hepes-buffered 199 tissue culture medium was confirmed to be a suitable basal medium, to which cryoprotectants were added. Addition of serum to the cryoprotective cocktail had no beneficial effect. There was no Three cryoprotectants, dimethyl sulfoxidc, ethanediol, and glycerol were ewhdkd. evidence of specific toxicity attributable to the cryoprotective agents during the permeation period; however, short permeation times at low temperature were associated with maximum skin viability. Following freezing and thawing, higher viabilities were obtained when using a slow ( - 1°C min- ‘) or medium (-M”C min- ‘) rather than a fast (immersion in liquid nitrogen) cooling rate. Dimethyl sulfoxide was a marginally better cryoprotectant overall, although this difference was not statistically significant. 0 1990 Academic Press, Inc.
Cryopreservation methods for the storage of allograft skin are desirable in view of the unpredictability of donor availability and patient requirements, thus necessitating a long-term skin storage technique. This becomes even more important if major histocompatibility complex (MHC) tissue matching is adopted. For autograft skin, the potential benefits of cryopreserving “excess” harvested grafts are that viable skin remains available for areas of the burn not yet ready to take grafts and that any failed grafts may be replaced at the bedside rather than requiring further operating theatre sessions to obtain additional skin. This potential benefit can only be realized if the cryopreservation technique is reproducible and retains a high cell viability on each occasion. Many different protocols have been advocated for the freezing of human skin (2,3, 9, 10, 11, 13, 15, 16). We have evaluated a number of these protocols and have been Received April 6, 1989; accepted June 25, 1989.
001l-2240/90 $3.00 Copy~@t All rights
0 1590 by Academic Press, Inc. of reproduction in any form reserved.
disappointed with the variability in clinical graft success and other viability measurements, both within and between skin samples (unpublished data). These evaluations could only be carried out when harvested skin was found to be in excess of clinical requirements. Because of this heterogeneity between skin samples and the influence of other variables associated with the recipient (e.g., wound bed vascularity and microbiological flora), it was impossible to adequately assess the effects of changes in the cryopreservation protocol on grafl success or graft viability. To achieve this, we have developed a defined inbred murine skin model (8), thus avoiding variability due to heterogeneous skin samples. A simple prognostic viability index, the measurement of tetrazolium reductase (TR) activity has been shown to correlate with graft success and oxygen consumption (8) and was therefore used to optimize aspects of the cryopreservation protocol. This is the first study of skin cryopreservation to use genetically homogeneous skin in a number of large factorial design experiments.
CRYOPRESERVATION
MATERIALS
AND
METHODS
Inbred Swiss Albino hairless mice were used as skin donors. The mice were bred continuously within the Department of Immunology animal house and fed an ad libiturn diet. Only female mice 34 months of age were used in order to minimize variability (8). The preparation of full thickness trunk skin has been described elsewhere (8). Reagents
The following reagents were used: 199 tissue culture medium containing 25 mM Hepes buffer (Gibco, Paisley, Scotland); Dulbecco’s phosphate-buffered saline (PBS; Oxoid, England); dimethyl sulfoxide (Me,SO), ethanediol, glycerol (Analar grade, BDH, Poole, England): and 0.9% saline (Steriflex, Boots, Nottingham, England). All other chemicals were obtained from Sigma (Poole, England).
OF MURINE
SKIN
165
device (Spembly Technical Products, Ltd., Sittingbourne, England). Medium cooling (- 60°C min- ‘) was achieved by lowering the nylon packet (long axis horizontal) to a predetermined position (usually 2.5 cm) above the level of liquid nitrogen in a liquid nitrogen refrigerator (Taylor-Wharton 3K, Jencons, England), Fast cooling was achieved by plunging the packets directly into liquid nitrogen. In each case nucleation was spontaneous. With the first two cooling methods, on reaching - lOO”C, the packets were transferred rapidly into liquid nitrogen for subsequent storage (- 196°C). Factors Evaluated
Basal medium. Ten replicate skin pieces were soaked for 2 hr at 4°C in each of the following basal solutions: 0.9% saline, PBS, Hepes-buffered 199 medium, Hepesbuffered 199 medium containing 15% (v/v) glycerol. Standard nylon packets containing the skin were subsequently heat sealed and cooled in the vapor phase of liquid nitrogen at - 60°C min-‘. After 24 hr, the Assays packets were thawed rapidly (immersion in TR activity was measured using Carney’s a 37°C water bath) and the skin was assayed (4) modification of Hershey’s (7) proce- immediately for TR activity. Serum. Skin pieces (IO replicates per dure, as described previously (8). For skin freezing experiments between 5 and 10 treatment) were soaked for 2 hr at 4°C in pieces of full thickness trunk skin 0.5-l cm2 Hepes-buffered 199 medium containing in size were placed into standard nylon 15% (v/v) glycerol and 0, 10, or 50% fetal calf serum. The nylon packets were subse(Jencons, Leighton Buzzard, England) packets (13 x 2.5 cm) on strips of sterile quently cooled at either a slow (- 1°C gauze. Two milliliters of cooled (4°C) cryo- min-‘) or a medium (-60°C min-‘) rate. protectant solution was added and the After 24 hr of storage in liquid nitrogen the packet was heat sealed. Various perme- skin was thawed rapidly and assayed for ation temperatures and permeation periods TR activity. Cryoprotectantlpermeation time, A were used. Cooling rates were determined three-way factorial experiment was carried using an electronic thermometer (Type K, CP Instrument Co.) attached to a chart re- out to assess the effects of cryoprotectants corder (Datatrace, Gallenkamp, Loughbor- (15% (v/v) and 25% (v/v) glycerol, Me,SO, and ethanediol in I99 medium, and a 199 ough, England). Type K thermocouples were inserted into the skin tissue. Three medium control), permeation temperature cooling rates were used in these experi- (4, 23, and 37”(Z), and permeation time (15 ments. Slow cooling (- 1°C min- ‘) was min, 2 hr, and 24 hr). Five replicate skin achieved using a controlled rate freezing pieces were used for each combination of
166
KEARNEY,
WHELDON,
factors. Following treatment the skin was assayed for TR activity, Cooling
ratelcryoprotectantipertneatioa
time. Three experiments were carried out using the following cooling rates: 1°C mm-’ (slow), -60°C min-’ (medium), and direct immersion in liquid nitrogen (fast). Each experiment was a two-way factorial design to assess the effects of cryoprotectants (15% (v/v), 25% (v/v), and 50% (v/ v) glycerol, Me,SO, and ethanediol in 199 medium, and a 199 medium control) and permeation time at 4°C (15 min and 2 hr). After cooling to at least - 100°C the skin packets were transferred to liquid nitrogen. Following storage for at least 24 hr, the skin was thawed rapidly (immersion in 37°C water bath) and immediately assayed for TR activity. Ten replicate skin pieces were used for each factor combination.
AND GOWLAND TABLE 1 Effects of Basal Media on TR Activity after Freezing (- 60°C min- ‘) and Thawing: (a) Analysis of Variance and (b) Comparison of Subgroup Means Using the T-Method (a)
df
MS
F
P
Media Within (error)
4 45
3876.44 66.87
57.97
Minimal significant difference = 13.5% (b)
Control
199 + glycerol
, 104% ,
,9X%+
199 PBS
1"
42%
Saline
37%,
Note. Underlined means are not significantly different at P < 0.05.
pared with saline, although these differences were not statistically significant at P < 0.05.
Statistical
Analysis
Addition of serum to the cryoprotective cocktail did not increase the viability of the Data from the factorial experiments was frozen skin, in fact the opposite was obsubjected to analysis of variance. A posteserved at the higher serum concentration riot-i multiple comparisons of means were (Table 2). There was no difference in the carried out using the T-method (14). All staviability of skin cooled at a slow or a metistical analyses were performed using the dium rate. Under slow cooling conditions, TR enzyme activity data expressed as abthere was a statistically significant decrease sorbance at 490 nm (X 103) . mg skin-t * ml in viability with increasing percentages of solvent. Following analysis the data were serum, while under medium cooling conditransformed to percentages, taking the origtions there was not. inal fresh skin enzyme activity as 100%. Data from the three-way factorial experiment to evaluate the effects of cryoproRESULTS tectant medium, permeation temperature, Following freezing and thawing, a signif- and permeation time on skin viability reicant decrease in viability was observed re- sulted in significant F values for two of the gardless of the medium used, compared factors, temperature and time (P < 0.05), with the enzyme activity of fresh skin (Ta- when compared with the pooled error and ble 1). A combination of glycerol with 199 interaction (temperature X time) mean medium gave the highest post-thaw viabil- squares. The third factor, media, had no ity. In the absence of a specific cryoprotec- significant effect. Short permeation times at tive agent, there was a trend toward a 4°C were superior to the other combinahigher retained viability for the media pos- tions (Fig. 1). sessing buffering capacity (PBS and 199) The mean percentage enzyme activities and additional beneficial effects of a com- of skin subjected to three cooling rates, plex nutritional environment (199) com- with cryoprotectants at three concentra-
CRYOPRESERVATION
ISmtn
TABLE 2 Effects of Serum on TR Activity after Freezing by Slow and Medium Cooling and Thawing: (a) Two-Way Analysis of Variance and (b) Comparison of Subgroup Means Using the T-Method (a)
df
MS
F
P
Cooling rate (A) Percentage serum (S)
1 2 2 54
137.92 910.94 35.93 129.61
1.06 7.03 0.28
N.S. co.05 N.S.
AxB
Within (error)
0))
0% serum
10% serum
50% serum
Slow cooling
65.7%
51.4%
49.3%
1 Medium cooling
I N.S. 61.‘s%
10(1bOM)% IOzoo-
Minimal significant difference = 15.3%
’
167
OF MURINE SKIN
N’S 51.
%$i N:S. 54.2%
l. :.
100-
:
M-
; cd2 % 402 a zLOc $
o-
,oo- z44hr
1 Nore. Underlined means are not significantly different at P < 0.05.
no%
bO-
40zo-
tions and two permeation times, are presented in Fig, 2, All three cryoprotectants provided some degree of protection compared with the 199 medium control. Following short permeation times the use of Me,SO frequently resulted in higher viability compared with the other two cryoprotectants, but this was statistically significant only for certain comparisons (e.g., following slow cooling with a 15% (v/v) concentration of the three cryoprotectants, (Fig. 2A). Comparing permeation times, shorter permeation appeared generally superior and was statistically significant at the higher cryoprotectant concentrations. Slow and medium cooling rates resulted in higher viability than rapid cooling. DISCUSSION
Because all of the skin used in this study was genetically homogeneous, large factorial experiments could be undertaken using pooled skin taken from a number of donors. In this way the “main effects” and
0-
IW
15% $$
25w GIV/VI in 199
15% %
a* t%
15* I Iv/4 in I*
25% ;;;g
FIG. 1. Effect of permeation temperature (E3, 4°C; 0, 23°C; n , 37°C) and permeation time of cryoprotectant (G, glycerol; D, Me,SO; E, ethanediol)containing medium on mean percentage TR activity (bars = ti minimal significant difference). Statistically significant differences occur where bars do not over-
“interaction” between variables in the cryopreservation procotol were assessed independently of the variation normaliy found in outbred populations. To study the effects of different cryoprotective agents and additives, a basal salt medium was required. Saline has been used for clinical skin storage, although its lack of buffering capacity could lead to a significant lowering of pH during the cryopreservation process when storing tissues that have a large population of metabolizing cells. It is therefore not surprising that saline performed worse than the other salt-
168 KU-J-A
KEARNEY,
WHELDON,
AND
GOWLAND
improved the residual viability of the skin. From this experiment it was decided to use 199 medium as the basal medium for the remainder of the study. The addition of fetal calf serum to 199/ glycerol medium had no beneficial effect at either slow or medium cooling rates. At the higher serum concentrations (50%) a significant reduction in viability was seen comJOQ-8 pared with the control. Using guinea pig ear T T skin, Lawrence (9) found no consistent improvement in viability after cryopreservation with 5-50% fetal calf serum. In the clinical situation these results are fortuitous in view of the potential for immune responses to xenogeneic proteins present in fetal calf serum or for problems associated with the acquisition of safe homologous serum or isogroup plasma as recommended by Blondet (3). A variety of cryoprotectant concentrations, permeation temperatures, and permeation times have been advocated for subsequent cryopreservation of skin with little evidence of systematic investigation. These include 2% Me,SO for 15 min at 20°C followed by 10% Me,SO at 10°C (3), 15% Fro. 2. Effect of cryoprotectants and permeation glycerol at 4°C for 2 hr (1 l), or 20% glycerol times (I3 = 15 min; 0 = 2 hr) on TR activity following freezing and thawing, using (A) slow, (B) medium, and at 0°C for 2-4 hr (6), and many others (2, 13, (C) fast cooling rates (bars = L/2minimal significant 15, 16). The optimal combination of these difference). Statistically significant differences occur factors depends on the temperature/ where bars do not overlap. time/toxicity interactions of the various cryoprotectants on skin cells, and the difcontaining media. Phosphate can be used to fusion and permeability characteristics of buffer saline (PBS); however, phosphate skin cells embedded in their tissue matrix, has a poor buffering capacity above a pH of toward each type of cryoprotectant. For 7.5 and tends to precipitate many polyva- mouse skin, the main effects and interaclent cations. The 199 medium retained tions of these variables on skin viability slightly higher viability and in this case were assessed by a factorial experiment. buffering capacity was provided by Hepes There were significant differences between which is known to be efficient even at low permeation temperatures and times, but temperatures (12). The use of 25 mA4Hepes surprisingly no differences between cryoin tissue culture medium for the hypother- protectants compared with 199 medium. mic (4°C) storage of skin has previously Therefore, in contrast to Lawrence’s findbeen reported by Cram et al. (5), where it ings for guinea pig skin, where both glycdemonstrated improved buffering capacity erol and Me,SO (15% in saline) reduced without any detrimental effects. Addition skin viability at room temperature (9), the of glycerol to the 199 medium significantly cryoprotectants did not decrease the tetra-
CRYOPRESERVATION
zolium reductase activity of murine skin at any of the temperatures tested. Whether the cryoprotectants caused sublethal damage predisposing the skin to freeze injury when used in high concentrations was assessedin the subsequent experiment. In this case, skin subjected to combinations of permeation time and cryoprotectant concentrations was cooled at three different rates. In the previous experiment, a permeation time of 24 hr was highly detrimental to the skin and was therefore omitted from this experiment, whereas an extra cryoprotectant concentration (50%) was included, and a low permeation temperature (4°C) was used. All three cryoprotective agents gave viabilities higher than those of the 199 medium control. However, even in the absence of a cryoprotective agent, skin cryopreserved in 199 medium retained a significant proportion of TR activity, particularly following a short permeation period. Similarly, short permeation times were generally superior for media containing cryoprotectants including glycerol, which has been reported to require longer permeation periods for many cell types to achieve a satisfactory cryoprotective concentration intracellularly (1). For Me,SO, the large decrease in viability following a 2-hr permeation time at the highest concentration (50%) may indicate cytotoxicity; however, this would need to be confirmed by further experiments. For the other cryoprotectants, and for all cryoprotectants following short permeation times, there was no evidence that higher concentrations of cryoprotectants either directly or indirectly reduced skin viability. The viabilities obtained following fast cooling were lower than those for the other two cooling rates. There was, however, no difference between slow (- 1°C min-‘) and medium (-60°C min- ‘) cooling. Therefore, the commonly recommended cooling rate of - 1°C min-‘, which can only be accurately achieved with sophisticated cooling devices, is not essential to the maxi-
OF MURINE
SKIN
169
mum attainable viability of skin following cryopreservation, and this would explain why other simpler cooling methods have yielded acceptable results for both human and animal skin (9, 10). This observation has been reported previously for human skin by several authors (see review in Ref. (2)), suggesting that the results obtained using this murine model may be applicable to the human clinical situation. ACKNOWLEDGMENT
We thank the Yorkshire Regional Health Authority for supporting this work. REFERENCES
1. Bank, H. L., and Brockbank, K. G. M. Basic principles of cryobiology. J. Cardiac Surg. l(3), 137-143 (1987). 2. Baxter, C., Aggarwal, S., and Diller, K. R. Cryopreservation of skin: A review. Transplant. Pruc. 17(6), 112-120 (1985). 3. Blondet, R., Gibert-Thevenin, M. A., Pierre, C., and Ehrsam, A. Skin preservation by programmed freezing. &it. J. Plast. Surg. 35,53& 536 (1982). 4. Carney, S. A., Hall, M., and Ricketts, C. R. The succinic dehydrogenase and cytochrome c oxidase activities of guinea-pig skin after mild heat damage. &it. J. Dermatol. 94, 295-299 (1976). 5. Cram, A. E., Domayer, M. A., and Scupham, R. Preservation of human skin: A study of two media using the athymic (nude) mouse model. J. Trauma 25, 128-130 (1985). 6. Graham, W. P., Hamilton, R. W., and Lehr, H. B. Versatility of skin allografts: Desirability of a viable frozen tissue bank. J. Trauma 11, 494-501 (1971). 7. Hershey, F. B., Cruickshank, C. N. D., and Mullins, L. 1. The quantitative reduction of 2,3,5triphenyl tetrazolium chloride by skin in vitro. J. Hisrochem. Cyruchem. 6, 191-l% (1957). 8. Keamey, J. N., Wheldon, L. A., and Gowiand, G. Cryopreservation of skin using a murine model: Validation of a prognostic viability assay. Cvobiology, 27, 24-30 (1990). 9. Lawrence, J. C. Storage and skin metabolism, &it. J. Plast. Surg. 25, 440453 (1972). 10. May, S. R., Guttman, R. M., and Wainwright, J. F. Cryopreservation of skin using an insulated heat sink box stored at -70°C. Cryobiology 22, 205-214 (1985). 11. Ninneman, J. L., Fisher, J. C., and Frank, H. A. Clinical skin banking: A simplified system for
170
KEARNEY,
WHELDON,
processing, storage and retrieval of human allografts. J. Trauma 18, 723-725 (1978). 12. Pegg, D. E. Perfusion technology. In “Organ Preservation for Transplantation” (A. M. brow and D. Pegg, Ed.) 2nd ed., pp. 477495. Dekker, New York, 1981. 13. Perry, V. P. A review of skin preservation. Clyobiology 3, lO!M30 (1966). 14. Sokai, R. R., and Rohlf, F. J. “Biometry: The
AND GOWLAND Principles and Practise of Statistics in Biological Research.” Freeman, San Francisco, 1981. 15. Volkova, N. A., and Sandomirsky, B. P. Organ culture of human skin preserved by freezing. Cryo-Lett.
3, 227-238 (1982).
16. Wachtel, T. L., Ninnemann, J. L., Fisher, J. C., Frank, H. A., and Inancsi, W. Viability of frozen allografts. Amer. J. Surg. 138, 783-787 (1979).