CRYOBIOLOGY
27, 42-54 (1990)
Propane-1,2-dial as a Potential Component Vitrification Solution for Corneas S. J. RICH’ Department
of Ophthalmology,
of a
W. J. ARMITAGE
AND
University
of Bristol, Bristol, United Kingdom
Any method of cryopreservation of the cornea must maintain integrity of the comeal endothelium, a monolayer of cells on the inner surface of the cornea that controls corneal hydration and keeps the cornea thin and transparent. During freezing, the formation of ice damages the endothelium, and vitrification has been suggested as a means of achieving ice-free cryopreservation of the cornea. To achieve vitrification at practicable cooling rates, tissues must be equilibrated with high concentrations of cryoprotectants. In this study, the effects of propane-l ,2-diol on the structure and function of rabbit comeal endothelium were studied. Corneas were exposed to concentrations of propane-l ,Zdiol ranging from 10 to 30% v/v in a Hepes-buflered Ringer’s solution containing glutathione, adenosine, 5 mmollliter sodium bicarbonate, and 6% w/v bovine serum albumin. Endothelial function was assessed by monitoring cornea1 thickness during perfusion of the endothelial surface at 34°C for 6 hr. Exposure to IO-15% v/v propane-1,2-dial was well tolerated for 20 min at 4°C when the cryoprotectant was removed in steps or by sucrose dilution. However, exposure to 25% v/v propane-l,Zdiol for 20 min at 0 or -5°C was consistently tolerated only when 2.5% w/v chondroitin sulfate was included in the vehicle solution. Exposure to 30% v/v propane-l ,Zdiol was harmful at - 5 and - 10°C. The endothelial damage following exposure to 30% Y/V propane-l ,2-diol was probably the result of a toxic effect rather than osmotic stress. Although 25% v/v propane-1,2-dial does not vitrify at cooling rates that are practicable for corneas, it could at this concentration form a major component of a vitrification solution comprising a mixture of cryoprolectants. 0 1990 Academic Press, Inc.
Conventional methods of cryopreservation involving ice formation have not, as yet, been satisfactorily applied to the cornea (23). Freezing injury in an organized tissue such as the cornea is caused by two factors: namely, ice formation per se, which causes structural damage and hence compromises overall function; and increasing solute concentration, as a result of ice crystallization, which damages individual cells (16, 24). Ice formation in corneas particularly undermines the integrity of the endothelium, which is a monolayer of cells that maintain cornea1 transparency (14). Vitrification is a method of cryopreservation that should avoid damage related to ice formation (6). To achieve vitrification at practicable cooling rates requires the pres-
ence of very high concentrations of cryoprotectants (12). In 1985, Rail and Fahy (19) vitrified mouse embryos using a solution containing dimethyl sulfoxide (20.5% w/v), acetamide (15.5% w/v), propane-1,2-diol (10% w/v), and polyethylene glycol (6% w/v). Human monocytes (22) and islets of Langerhans (9) have since been shown to retain functional integrity after vitrification using the same cryoprotectant mixture, thus suggesting that this approach to cryopreservation is’applicable to a range of cell types. Preliminary experiments on vitrification of the cornea showed that cornea1 endothelium tolerated brief exposure to this solution (1). Propane- I ,2-diol has good glass-forming properties (4), an apparently low toxicity, and high cell permeability (18). The aim of the present study, therefore was to investiReceived December 12,1988; accepted May 2, 1989. ’ To whom reprint requests should be addressed at gate the effects of propane-1,2-diol on the structure and function of rabbit comeal enU.K. Transplant Service, Southmead Road, Bristol, BSlO 5ND, UK. dothelium. 42 0011-2240190$3.00 Copyrigbr Q 1990 by Academic F’rcss, Inc. All rights oE reprduction in any form reserved.
43
CORNEAL TOLERANCE OF PROPANE-l ,2-DIOL TABLE 1 Composition of Solutions”
NaCl KCI CaCl, * 6H,O MgC12 . 6Hz0 NaHC03 NaH,PO . tH,O Glucose Glutathione Adenosine Hepes buffer Bovine serum albumin Newborn calf serum
GBRb (mmol/liter)
Bicarbonate-free Ringe? (mmol/liter)
GBR,= (mmoliliter)
128.34 4.83 1.04 0.79 29.16 0.66 5.00 0.30 0.50 1.o% v/v
128.34 4.83 1.04 0.79 0.66 5.00 0.30 0.50 1.0% v/v
128.34 4.83 1.04 0.79 4.99 0.66 5.00 0.30 0.50 20.00 6.0% w/v -
a GBR and bicarbonate-free Ringer were gassed with 5% CO2 in oxygen. b Solutions for normothermic perfusion. c Cryoprotectant vehicle.
Solutions
METHODS
Preparation of Corneas New Zealand White rabbits (2-3 kg) were killed by an intravenous overdose of pentobarbitone sodium, The eyes were enucleated and the corneas removed by the method of Dikstein and Maurice (5). The corneas were secured on plastic support rings to minimize damage through wrinkling and distortion.
The composition of glutathione-bicarbonate Ringer (GBR), which is adequate for maintaining endothelial function for several hours during normothermic perfusion, is given in Table 1. This solution was supplemented with 1% v/v newborn calf serum (17). Two other versions of the solution were prepared: namely, without sodium bicarbonate, and with only 5 mmol/liter sodium bicarbonate (GBR,) (Table 1). In the former solution, the lack of sodium bicarbonate was not compensated for by another solute; TABLE 2 Protocol A: Stepwise Addition and Dilution of 10% v/v Propane-l ,t-diol
Perfusion
Time
(hours)
FIG. 1. Change in mean stromal thickness of four corneas (?SEM) during perfusion immediately after comeal isolation. Perfusion with isosmotic GBR was interrupted by 2 hr of bicarbonate-free perfusion (open circles). The rate of swelling between 2.5 and 4 hr was 34 + 5 p.m/hr and the rate of thinning between 4.5 and 6 hr was 21 + 1 pm/hr.
Step
Propane- 1,2-diol (% v/v)
Duration (n-W
1
0
10
2 3 4 5 6 7 8 9
5 10 8 6 4 2 1 0
10 20 5 5 5 5 5 10
Temperature ec,
44
RICH AND
ARMITAGE
TABLE 3 Protocol 8: Stepwise Addition and Sucrose Dilution of 10% v/v Propane-L,Zdiol Duration (tin)
Temperature ec,
0
10
0 0 1 0
10 20 15 15
4 4 4 25 25
Step
Propane-l ,2-diol (% v/v)
Sucrose W
1
0
2 3 4 5
5 10 0 0
thus its osmolality was less than that of GBR. The latter solution, which contained 6% w/v bovine serum albumin but no newborn calf serum, was used as the cryoprotectant vehicle. Chondroitin sulfate (2.5% w/v; Sigma, grade III, 99% mixed isomers, sodium salt from whale or shark cartilage) or 2.5% w/v dextran (Sigma, av mol wt 40,000 Da) and 1 mol/liter sucrose were also included in this solution as indicated. The molar concentration of electrolytes in GBR5 was maintained constant in all solutions containing propane- 1,2-diol or sucrose .
from the eye: these corneas served as a control group. Each experimental group also comprised four corneas. Differences between groups both in mean stromal thickness at various times during perfusion and in rates of change in thickness were com-
Assessment of Endothelial Function
After exposure to propane-l ,Zdiol, the comeal endothelium was perfused (1.25 ml/ hr, 13 cm H,O) for 6 hr at 34°C. The, epithelial surface was covered with silicone oil (Dow Corning 200/20 cs) to prevent evaporation. Endothelial function was assessed by measuring stromal thickness using a specular microscope with a calibrated fine focus. Measurements were made by focusing in turn on the endothelium and the epithelium, and mean stromal thickness was calculated from duplicate measurements taken at three random sites. Measurements were made at 30-min intervals throughout the perfusion period. The initial 2 hr of perfusion was with GBR. The perfusate was then changed to the bicarbonate-free Ringer solution for a further 2 hr before returning to GBR perfusion for the final 2 hr. Four untreated corneas were perfused, as described, immediately after their isolation
300! 0
, . ,
1
2 Pwfuskm
, . , , , , , 3 Time
4
6
B
hxml
FIG. 2. Each graph represents change in mean corneal thickness of four corneas (*SEM) during perfusion following exposure to 10% v/v propane-l,Z-diol. The 6&r GBR perfusion was intermpted by 2 hr of bicarbonate-free perfusion (open circles). (A) Stepwise dilution (protocol A). The rate of swelling between 2.5 and 4 hr was 37 2 3 cLm/hrand the rate of thinning between 4.5 and 6 hr was 31 + 4 qn/hr. (B) Sucrose dilution (protocol B). The rate of swelling between 2.5 and 4 br was 28 k 2 pm/br and the rate of thinning between 4.5 and 6 hr was 18 + 2 &hr.
45
CORNEAL TOLERANCE OF PROPANE-1,2-DIOL
FIG. 3. Scanning electron micrograph of the endothelial surface of a cornea following exposure to 10% v/v propane-1,2-diol (protocol B) and a 4hr perfusion. The endothelial cells appear normal, i.e., uniform in size and shape with clearly defined intercellular cell borders. Bar = 25 pm.
pared by one-way analysis of variance, with the level of significance set at 5%, and the Welsch step-up procedure (20). Means are quoted with the standard error of the mean.
buffer (pH 7.3) containing 3 mmol/liter CaCl,. The corneas were subsequently washed in the same buffer prior to criticalpoint drying and gold sputtering. The corneas were examined in a Philips 505 scanning electron microscope.
Scanning Electron Microscopy
After the 6-hr perfusion the corneas were Exposure of Corneas to Propane-l ,2-diol fixed overnight at 4°C with 2.5% v/v gluImmediately after excision, corneas were taraldehyde in 0.1 mol/liter cacodylate immersed in 50 ml GBRS in polystyrene TABLE 4 Protocol C: Stepwise Addition and Three-Step Sucrose Dilution of 15% v/v Propane-1,2-dial Propane-l ,2-diol (% v/v)
Sucrose WI
Duration (min)
Temperature (“Cl
cl 5 10 15 5 0 0
cl 0 0 0 1 1 0
10 10 10 20 15 15 15
4 4 4 4 25 25 25
46
RICH AND ARMITAGE RESULTS
Perfusion of Freshly Isolated Corneas
At the start of perfusion, these corneas were 420 + 10 pm thick. During the initial 2 hr of perfusion with GBR, they thinned slightly to 410 2 6 pm (Fig. I). During the 30 min immediately following the changeover to the bicarbonate-free perfusate, the 2 O 3 5 6 corneas swelled to 475 + 10 pm: they then Perfuslon Tkne (hours) FIG. 4. Change in mean comeal thickness of four continued to swell at a lower rate of 34 f 5 When perfusion with GBR was recorneas (?SEM) during perfusion following sucrose p&r. dilution of 15% v/v propane-l ,2-diol (protocol C). The sumed, there was a drop in thickness of 47 6-hr GBR perfusion was interrupted by 2 hr of bicar+ 4 pm during the first 30 min after changebonate-free perfusion (open circles). The rate of swell- over: this was followed by a period of tbining between 2.5 and 4 hr was 37 + 7 CLm/hrand the rate ning at a lower rate of 21 * 1 pm/hr. The of thinning between 4.5 and 6 hr was 19 2 3 km/hr. initial rapid changes in thickness that folpots. The corneas were then exposed to fi- lowed the changeover of perfusate were in nal concentrations of propane-l ,2-diol of part due to the difference in osmolality beIO, 15, 25, or 30% v/v. To reduce osmotic tween the two perfusates. The subsequent stress during the removal of cryopro- lower rates of swelling and thinning were tectant, the concentration of propane- similar to those reported by Hodson and 1,Zdiol was reduced in steps OF by a “su- Miller (8) for corneas in which the endothecrose dilution” technique (15). The proto- lial bicarbonate pump was first inhibited cols for the addition and subsequent and then restored, respectively. removal of propane-l ,2-diol are given in Tables 2-8. Corneas were immersed in 50 Stepwise Dilution of 10% vlv (I.4 M) Propane-i ,2-dial ml of solution for each step, and exposure temperatures of 0, -5, and - 10°C were Corneas were exposed to 10% v/v proattained with ice/water slush, an alcohol pane-1,2-diol (PD) for 20 min at 4°C accordbath (FTS Multicool), and a controlled-rate ing to protocol A (Table 2). Figure 2A freezer (Planer Biomed Kryo lo), respec- shows that during the initial 2 hr of GBR tively. perfusion, these corneas thinned slightly to
J.-v-T-
TABLE 5 Protocols D and E: Exposure of Corneas to 25% v/v Propane-l ,2-dial at WC (D) or at - 5°C (E)”
Step
Propane- 1,Zdiol (or0v/v) 0 5 10 15 25 LO 0 0
Sucrose (Ml
Temperature (X’)
Duration bin)
D
E
10 IO 10 20 20 15 15 15
4 4 4 4 0 0 25 25
4 4 4 4 -5 0 25 25
n The propane-1,2-diol was removed by a three-step sucrose dilution. These protocols were repeated with all solutions containing 2.5% w/v chondroitin sulfate or 2.5% w/v dextran.
CORNEAL TOLERANCE OF PROPANE-1,ZDIOL
47
FIG. 5. Scanning electron micrograph of corneal endothelium following exposure to 25% v/v propane-l ,2-diol (protocol E) and 6 hr perfusion. A central large hexagonal cell (a normal occurrence in rabbit endothelium) contrasts with the large cell to the right which is irregular in shape and appears to have resulted from intercellular membrane breakdown or cell loss. Bar = 25 km.
compensate for the swelling that occurred during the removal of PD. The corneas then demonstrated a period of swelling during perfusion with bicarbonate-free Ringer, followed by a period of thinning when perfusion with GBR was resumed. This behavior was indicative of a functioning endotheliurn. Cornea1 thickness and the rates of swelling and thinning were similar to the control values throughout the perfusion. All four corneas showed signs of endothelial disruption at the beginning of the perfusion in that dark areas were present where cells had apparently fallen away from Descemet’s membrane. This soon reorganized and a confluent endothelial mosaic was maintained for the remainder of the perfusion period. Sucrose Dilution of 10% vlv (1.4 M) Propane-i,2-dial
Corneas were exposed to 10% v/v PD for
20 min at 4°C according to protocol B (Table 3); but, unlike protocol A, the dilution of PD was carried out with a solution containing 1 mol/liter sucrose and in fewer steps. At the start of perfusion, these corneas were thinner than the controls (P < O.OS),presumably because they were partially dehydrated during the sucrose dilution, and Fig. 2B shows that they swelled rather than shrank (cf. Fig. 2A) during the first 2 hr of perfusion. During the ensuing perfusion, however, the mean thickness and rates of swelling and thinning were not significantly different from the control values. The reversal of swelling following cessation of the bicarbonate-free perfusion again demonstrated a functioning endothelial pump. The rate of thinning during the last 1.5 hr of perfusion was, however, lower (P < 0.05) than that of corneas exposed to 10% PD by protocol A (i.e., PD removed without sucrose dilution). All four corneas
48
RICH AND ARMITAGE
FIG. 6. Scanning electron micrograph of comeal endothelium exposed to 25% v/v propane-1,Zdiol (protocol E) in the presence of 2.5% w/v dextran. The endothelial cells are hexagonal and uniform in size with no visible sign of damage. Bar = 25 pm.
had a visibly intact endothelial mosaic throughout the 6-hr perfusion and scanning electron microscopy performed after the perfusion showed a normal endothelial mosaic with regularly sized hexagonal cells and intact intercellular cell borders (Fig. 3). Sucrose Dilution of 15% vlv (2.0 M) Propane-l ,2-diol
Corneas were exposed to 15%v/v PD for 20 min at 4°C according to protocol C (Table 4). Three of the four corneas in this group had a visibly normal endothelial mosaic throughout the perfusion period. AlI four corneas had a functioning endothelial pump (Fig. 4) despite the fourth cornea showing slight areas of endothelial damage during perfusion. These corneas were thinner than the controls at the start of perfusion (P < 0.05) but after 2 hr, the thickness and rates of swelling and thinning were not significantly different from the control values.
Sucrose Dilution of 25% vlv (3.4 M) Propane-l ,2-diol
Corneas were exposed to 25% v/v PD for 20 min at 0°C (protocol D) or at - 5°C (protocol E) (see Table 5). The four corneas exposed to 25% PD at 0°C were initially hazy but soon cleared to reveal a normal endothelial mosaic. By the end of the dhr perfusion, however, all four corneas were losing cells from the endothelial mosaic. One cornea showed no decrease in thickness on changeover from bicarbonate-free Ringer and went on to swell at 1.8 +n/hr during the ensuing GBR perfusion, which suggesteda lack of bicarbonate pump activity or a loss of barrier function. The mean rate of thinning for the three remaining corneas was 16.6 pm/hr (range: 30.2-22.8 prnlhr) . When corneas were exposed to 25% PD at - 5”C, three of the corneas, although iuitially hazy, soon cleared to show an intact endothelial mosaic that persisted for the re-
CORNEAL TOLERANCE OF PROPANE-1,2-DIOL
FIG. 7. Each graph represents change in mean corneal thickness of four corneas (-+SEM) following exposure to 25% v/v propane-1,2-diol with 2.5% w/v chondroitin sulfate. The 6-hr GBR perfusion was interrupted by 2 hr of bicarbonate-free perfusion (open circles). (A) Exposure to 25% v/v propane-1,2-dial plus chondroitin sulfate at 0°C. The rate of swelling between 2.5 and 4 hr was 37 & 2 Frn/hr and the rate of thinning between 4.5 and 6 hr was 11 * 2 prn/hr. (B) Exposure to 25% v/v propane-l ,2-diol plus chondroitin sulfate at - 5°C. The rate of swelling between 2.5 and 4 hr was 35 + 1 prn!hr and the rate of thinning between 4.5 and 6 hr was 15 + 4 cl.m/hr.
mainder of the perfusion. These three corneas demonstrated endothelial pump activity during the final 1.5 hr of perfusion by thinning at a mean rate of 17.8 cl,m/hr (range: 14.4-22.2 &hr). The fourth cornea, however, swelled during the final GBR perfusion, and had a persistently hazy appearance. Scanning electron microscopy of the corneas following perfusion showed some large irregular cells resulting from the breakdown of intercellular ceil membranes or cell loss (Fig. 5). Protocol E was repeated with the addition of 2.5% w/v dextran to all vehicle so-
49
lutions. Three corneas were subjected to this treatment, two of which showed a functioning bicarbonate pump by thinning during the final 1.5 hr of perfusion at 18.6 and 19.2 kmlhr, respectively. The third cornea swelled slightly during the final GBR petfusion (2.4 pm/hr) although subsequent microscopy showed a good endothelial mosaic with uniformly sized hexagonal cells (Fig. 6). Consistently good results were obtained with protocols D and E, however, when 2.5% w/v chondroitin sulfate was added to all vehicle solutions. In each case the initially hazy endothelium soon cleared and stayed clear, retaining a normal mosaic for the remainder of the perfusion. AU of these corneas demonstrated a functioning endothelial pump (Fig. 7). Apart from their initial thickness at the start of perfusion, which was less than the control values (P < O.OS),the thickness of these corneas and their rates of swelling and thinning were not significantly different from the control values. Scanning electron microscopy of these corneas following the 6-hr perfusion showed a good endothelial mosaic with uniformly sized hexagonal cells and intact intercellular cell borders (Fig. 8). Sucrose Dilution of 30% v/v (4.1 M) Propane-l ,2-diol Corneas were exposed to 30% v/v PD plus 2.5% w/v chondroitin sulfate for 10 min at either - 5 or - 10°C according to protocols F, G, and H (Tables 6 and 7). Protocols H and I (Tables 7 and S), involving exposure to 30% v/v PD at - lO”C, were performed with 2.5% w/v dextran in place of chondroitin sulfate. In all experiments involving exposure to 30% v/v PD the corneas were initially thin but showed some signs of endothelial damage in that cells were missing from the mosaic. Despite variations in the temperature of exposure to 30% v/v PD and in the rate of dilution, after about 1 hr of perfusion all of these corneas became so cloudy that stromal thickness
50
RICH AND ARMITAGE
FIG. 8. Scanning electron micrograph of the endothelial surface of a cornea exposed to 25% v/v propane-1,2-diol (protocol D) in the presence of 2.5% w/v chondroitin sulfate. The endothelial mosaic appears undamaged; the hexagonal cells are regular in size with intact intercellular borders. Bar = 25 urn.
measurements were no longer possible. The corneas remained opaque for the rest of the perfusion period. Scanning electron microscopy performed after the 6-hr perfu-
sion revealed extensive damage to these corneas as exemplified by Figs. 9 and 10. Large irregular cells resulting from cell loss or membrane breakdown (Fig. 9) and se-
TABLE 6 Protocols F and G: Exposure of Corneas to 30% v/v Propane-1,2-diol for 10 min at -5°C” Duration (min)
Step
Propane-l ,Zdiol (% v/v)
Temperature (“C)
F
G
1 2 3 4 5 6 7 8 9 10
0 5 10 15 25 30 25 10 0 0
4 4 4 4 -5 -5 -5 0 25 25
10 10 10 10 i0 10 15 15 15 1.5
10 10 10 10 IO 20 20 20 20
10
QThe propane-l ,2-diol was removed by sucrose dilution in 4 x 15-min steps (protocol F) or 4 x 20-min steps (protocol G). All solutions contained 2.5% w/v chondroitin sulfate.
CORNEAL
51
OF PROPANE-1,2-DIOL
TOLERANCE
TABLE 7 Protocol II: Exposure of Corneas to 30% v/v Propane-l ,2-diol at - 10°C for 10 mine Step
Propane- 1,2-diol (% v/v)
Sucrose (Ml
Temperature 03
Duration bin)
1
0
cl
4
10
2 3 4 5 6 7 8 9 10
5 10 15 25 30 25 10 0 0
0 0 0 0 0 1 1 1 0
4 4 4 -5 -10 -5 0 25 25
10 10 10 10 10 20 20 20 20
a The orouane-1 .2-diol was removed bv sucrose dilution in 4 X 20-min steps. All solutions contained 2.5% w/v _ chondrokn sulfate or 2.5% w/v dextran.
age than swelling (2); hence any damage due to osmotic stress is more likely to occur during the dilution of cryoprotectant. Osmotic stress during removal of the cryoprotectant can be mitigated by performing dilutions in steps or by balancing the decrease in concentration of extracellular permeating cryoprotectant by the simultaneous addition of a nonpermeating solute such as sucrose (so-called sucrose dilution). The advantage of sucrose dilution is that there can be larger decrements in the concentration of permeating cryoprotectant at each dilution step, thereby reducing the time that cells are exposed to potentially
verely contorted intercellular borders (Fig. 10) were typical in corneas exposed to 30% v/v PD under these conditions. DISCUSSION
High concentrations of permeating cryoprotectants can cause cellular damage by toxic effects or by osmotic stress created during addition and removal of cryoprotectant. Osmotic imbalances cause shrinkage of cells during addition of cryoprotectant and swelling during cryoprotectant removal. Cornea1 endothelial cells have been shown to be more tolerant of shrink-
TABLE 8 Protocol I: Exposure of Corneas to 30% v/v Propane-l ,2-diol at - 10°C for 10 mm* Step 1
2 3 4 5 6 7 8 9 10
Propane-l ,2-dial (% v/v)
Sucrose @fJ
0 5
0 0 0 0 0 0
10 15 25 30 15 5 0 0
1 1 1 0
Temperature eo 4 4 4 4 -5 - 10 0 25 25 25
Duration (mid 10 10 10 10 10 10 15 15 15 15
n The propane-1,2-diol was removed by sucrose dilution in 4 x 15min steps. All solutions contained 2.5% w/v dextran .
52
RICH
AND
ARMITAGE
FIG. 9. Scanning electron micrograph of corneal endothelium exposed to 30% v/v propane-1,2-diol (protocol F) with 2.5% w/v chondroitin sulfate. Irregularity in cell size and shape indicates endothelial disruption. Bar = 25 pm.
toxic concentrations of cryoprotectant (6, 19). Twenty minutes of exposure to 10% (1.4 M’) or 15% (2.0 AI) v/v propane-1,Zdiol at 4°C was well tolerated by corneal endothelium. Twenty minutes of exposure to 25% v/v (3.4 M) propane-1,2-diol, on the other hand, did cause some endothelial damage at both 0 and - 5°C. The addition of 2.5% w/v chondroitin sulfate to the cryoprotectant vehicle solution ameliorated the damage caused by exposure to 25% v/v propaneI ,2-diol. The substitution of chondroitin sulfate with 2.5% w/v dextran (which is similar in molecular weight to chondroitin sulfate) also had a beneficial effect; but dextran was not as effective as chondroitin sulfate in reducing damage caused by propane-l ,Zdiol. Chondroitin sulfate occurs naturally in comeal stroma (14) and is included in short-term storage solutions for human corneas because of an apparently
beneficial effect on the endothelium (IO, 11, 21), The corneas exposed to 10, 15, and 25% v/v propane-1,2-diol (the latter in the presence of 2.5% w/v chondroitin sulfate) showed no differences at the 5% level from the freshly isolated corneas in their rates of swelling and thinning when bicarbonate was restored to the perfusate. The rate of thinning of edematous corneas is dependent on the balance between the rate of pumping of bicarbonate ions by the endothelium and the passive leak of water and solute across the endothelium into the stroma. The leak is governed by the passive permeability of the endothelial monolayer and by the stromal swelling pressure, which is inversely related to corneal thickness (14). The similarity of swelling rates in these corneas exposed to propane- 1,2-diol suggested that the passive endothelial permeabilities, and hence the endotheIial barrier properties,
CORNEAL TOLERANCE OF PROPANE-1,2-DIOL
53
FIG. 10. Scanning electron micrograph of cornea endothelial surface following exposure to 30% v/v propane-1,2-diol (protocol H) with 2.5% w/v chondroitin sulfate. There is evidence of extensive damage to the endothelial mosaic. The intercellular borders are contorted resulting in apparent “holes” between some cells. Bar = 25 pm.
were similar to control values. Since these corneas were also of similar thickness to the controls at the beginning of the second period of GBR perfusion, the subsequent thinning (after the initial rapid fall in thickness) was presumed to be the result of endothelial pump activity. Exposure of 30% v/v (4.1 M) propane1,2-diol for 10 min at -5°C was not tolerated by corneal endothelium. Neither increasing the duration of the dilution steps from 15 to 20 min to reduce osmotic stress nor decreasing the propane-l ,2-diol concentration in larger steps to reduce the time of exposure to potentially toxic levels of the cryoprotectant improved endothelial function. In a further attempt to reduce toxicity of propane-l,Zdiol, exposure to 30% v/v was performed at a lower temperature ( - 10°C). There was, however, still no improvement in endothelial function, It is also possible that the use of higher concentrations of chondroitin sulfate may improve the function of these corneas.
Two molar propane-1,Zdiol has been reported to be damaging to comeal endothelium after only 5 min exposure at 20°C (13); however, a serial dilution technique was used. Limited tolerance of other tissues to propane- 1,2-diol has recently been reported, Reduced survival of human blood platelets occurs following exposure to greater than 2 mol/liter propane-1,2-diol at 20°C (3). Kidneys have been reported to tolerate only 2.5-3 M (7,18). Although corneas tolerated exposure to 25% v/v propane-l ,Zdiol, they were damaged by higher concentrations under the conditions studied in these experiments. Apparent vitrification of 35% v/v (4.8 M) propane-l,2-diol in the presence of 2.5% w/v chondroitin sulfate plus 6% w/v bovine serum albumin was achieved by plunging a glass vial containing 5 ml of the solution into liquid nitrogen (unpublished observation). Although no ice crystals were visible, nucleation would almost certainly have occurred upon cooling, necessitating ex-
54
RJCH AND ARMITAGE
tremely high warming rates to avoid devitrifrcation and recrystallization. Such high cooling and warming rates are impracticable for the cornea. Thus, a comeal vitrification solution based solely on this cryoprotectant is currently not feasible. However, since 25% v/v propane-1,Zdiol is well tolerated, this compound could form a major component of a vitrification solution containing a mixture of cryoprotectants. ACKNOWLEDGMENTS S.J.R. is supported by a B&ton Research Studentship from the National Eye Research Centre. W.J.A. is supported by a Royal National Institute for the Blind Research Fellowship. The work was funded by a grant from the T. F. C. Frost Charitable Trust. We thank Professor David Easty for his continued support of this work. REFERENCES 1. Armitage, W. J. Survival of comeal endothelium following exposure to a vitrification solution. Cryobiology26, 318-327 (1989). 2. Armitage, W. J., Moss, S. J., and Easty, D. L.
Effects of osmotic stress on rabbit comeal endothelium. Cryobiology 25, 425439 (1988). 3. Amaud, F., and Pegg, D. E. Attempts to cryopreserve platelets with propylene glycol. CryobioC Qgy 24, 587(1987). 4. Boutron, P., and Kaufmann, A. Stability of the
amorphous state in the system water-1,2propanediol. Cryobiology 16, 557-568 (1979). 5. Dikstein, S., and Maurice, D. M. The metabolic basis to the fluid pump in the cornea. J. fhysiol 221, 29-41 (1972).
6. Fahy, G. M., MacFarlane, D. R., Angell, C. A., and Meryman, H. T. Vitrification as an approach to cryopreservation. Cryobiology 21, 407-426(1984). 7. Halasz, N. A., and Collins, G. M. Studies in cryo-
preservation, II. Propylene glycol and glycerol. Cryobiology 21, 14147 (1984). 8. Hodson, S., and Miller, F. The bicarbonate ion pump in the endotbelium which regulates the hydration of rabbit cornea. J. Physiol. 263,563577 (1976). 9. Jutte, N. H. P. M., Heyse, P., Jansen, H. G.,
Bruining, G. J., and Zeilmaker, G. H. Vitriftcation of human islets of Langerhans. Cryobiology 24,403-411 (1987). 10. Kaufman, H. E., Vamell, E. D., Kaufman, S.,
Beuermau, R. W., and Barron, Et. A. K-Sol corneal preservation. Amer. J. Ophthaimol. 100, 299-304(1985).
11. Lindstrom, R. L., Skelnik, D. L., Mindrup. E. A., and Doughman, D. J. Organ culture corneal preservation with chondroitin sulphate. In “ARVO Abstracts.” Invest. Ophthalmol. Vis. Sci. Suppl., p. 266 (1984). 12. MacFarlane, D. R. Physical aspects of vitrification in aqueous solutions. Cryobiology 24, 181195 (1987). 13. Madden, P. W. The assessment of endothelial integrity by scanning electron microscopy and fluorescein diacetate staining following treatment with cryoprotective additives. Curr. Eye Res. 8, 17-36 (1989). 14. Maurice, D. M. The cornea and sclera. ZR “The Eye” (H. Davson, Ed.), 3rd ed., Vol. lb, Chap. 1, pp. L-158. Academic Press, Orlando, FW London, 1984. 15. Mazur. P. Fundamental cryobiology and the preservation of organs by freezing. In “Organ Preservation for Transplantation” (A. M. Karow and D. E. Pegg, Eds.), 2nd ed., pp. 143-175. Marcel Dekker, New York, 1981. 16. Mazur, P. Freezing of living cells: Mechanisms and implications. Amer. J. Physiol. 247, 125142 (1984). 17. Neuwirth Lux, O., and Dikstein, S. Survival of isolated rabbit cornea and free radical scavengers. Curr. Eye Res. 4, 15S-154 (1985). 18. Pegg, D. E., Jacobsen, I. A., Diaper, M. P., and Foreman, J. Perfusion of rabbit kidneys with solutions containing propane-l ,2-dial. Crjlobiology 24, 42B-428 (1987). 19. Rail, W. F., and Fahy, G. M. Ice-free cryopreservation of mouse embryos at - 196°Cby vitrllication. Nature (London) 313, 573-575 (1985). 20. Sokal, R. R., and Rohlf, F. J. “Biometry,” 2nd ed. Freeman, New York, 1981. 21. Stein, R. M., Boume, W. M., andcampbell, R. J. Chondroitin sulfate for comeal preservation at 4°C. Evaluation by electron microscopy. Arch. Ophrhalmol. 104, 13%1361(1986). 22. Takahashi, T., Hirsch, A., Erbe, E. F., Brass, J. B., Steere, R. L., and Williams, R. J. Vitrification of human monocytes. Cryobiology 23, 103-115 (1986). 23. Taylor, M. J. Clinical cryobiology of tissues: Preservation of corneas. Cryobiology 23, 323-353 (1986). 24. Taylor M. J., and Pegg, D. E. The effect of ice formation on the function of smooth muscle tissue stored at - 21 or -60°C. Cryobiology 20, 36-N-I (1983).
27, 42-54 (1990)
Propane-1,2-dial as a Potential Component Vitrification Solution for Corneas S. J. RICH’ Department
of Ophthalmology,
of a
W. J. ARMITAGE
AND
University
of Bristol, Bristol, United Kingdom
Any method of cryopreservation of the cornea must maintain integrity of the comeal endothelium, a monolayer of cells on the inner surface of the cornea that controls corneal hydration and keeps the cornea thin and transparent. During freezing, the formation of ice damages the endothelium, and vitrification has been suggested as a means of achieving ice-free cryopreservation of the cornea. To achieve vitrification at practicable cooling rates, tissues must be equilibrated with high concentrations of cryoprotectants. In this study, the effects of propane-l ,2-diol on the structure and function of rabbit comeal endothelium were studied. Corneas were exposed to concentrations of propane-l ,Zdiol ranging from 10 to 30% v/v in a Hepes-buflered Ringer’s solution containing glutathione, adenosine, 5 mmollliter sodium bicarbonate, and 6% w/v bovine serum albumin. Endothelial function was assessed by monitoring cornea1 thickness during perfusion of the endothelial surface at 34°C for 6 hr. Exposure to IO-15% v/v propane-1,2-dial was well tolerated for 20 min at 4°C when the cryoprotectant was removed in steps or by sucrose dilution. However, exposure to 25% v/v propane-l,Zdiol for 20 min at 0 or -5°C was consistently tolerated only when 2.5% w/v chondroitin sulfate was included in the vehicle solution. Exposure to 30% v/v propane-l ,Zdiol was harmful at - 5 and - 10°C. The endothelial damage following exposure to 30% Y/V propane-l ,2-diol was probably the result of a toxic effect rather than osmotic stress. Although 25% v/v propane-1,2-dial does not vitrify at cooling rates that are practicable for corneas, it could at this concentration form a major component of a vitrification solution comprising a mixture of cryoprolectants. 0 1990 Academic Press, Inc.
Conventional methods of cryopreservation involving ice formation have not, as yet, been satisfactorily applied to the cornea (23). Freezing injury in an organized tissue such as the cornea is caused by two factors: namely, ice formation per se, which causes structural damage and hence compromises overall function; and increasing solute concentration, as a result of ice crystallization, which damages individual cells (16, 24). Ice formation in corneas particularly undermines the integrity of the endothelium, which is a monolayer of cells that maintain cornea1 transparency (14). Vitrification is a method of cryopreservation that should avoid damage related to ice formation (6). To achieve vitrification at practicable cooling rates requires the pres-
ence of very high concentrations of cryoprotectants (12). In 1985, Rail and Fahy (19) vitrified mouse embryos using a solution containing dimethyl sulfoxide (20.5% w/v), acetamide (15.5% w/v), propane-1,2-diol (10% w/v), and polyethylene glycol (6% w/v). Human monocytes (22) and islets of Langerhans (9) have since been shown to retain functional integrity after vitrification using the same cryoprotectant mixture, thus suggesting that this approach to cryopreservation is’applicable to a range of cell types. Preliminary experiments on vitrification of the cornea showed that cornea1 endothelium tolerated brief exposure to this solution (1). Propane- I ,2-diol has good glass-forming properties (4), an apparently low toxicity, and high cell permeability (18). The aim of the present study, therefore was to investiReceived December 12,1988; accepted May 2, 1989. ’ To whom reprint requests should be addressed at gate the effects of propane-1,2-diol on the structure and function of rabbit comeal enU.K. Transplant Service, Southmead Road, Bristol, BSlO 5ND, UK. dothelium. 42 0011-2240190$3.00 Copyrigbr Q 1990 by Academic F’rcss, Inc. All rights oE reprduction in any form reserved.
43
CORNEAL TOLERANCE OF PROPANE-l ,2-DIOL TABLE 1 Composition of Solutions”
NaCl KCI CaCl, * 6H,O MgC12 . 6Hz0 NaHC03 NaH,PO . tH,O Glucose Glutathione Adenosine Hepes buffer Bovine serum albumin Newborn calf serum
GBRb (mmol/liter)
Bicarbonate-free Ringe? (mmol/liter)
GBR,= (mmoliliter)
128.34 4.83 1.04 0.79 29.16 0.66 5.00 0.30 0.50 1.o% v/v
128.34 4.83 1.04 0.79 0.66 5.00 0.30 0.50 1.0% v/v
128.34 4.83 1.04 0.79 4.99 0.66 5.00 0.30 0.50 20.00 6.0% w/v -
a GBR and bicarbonate-free Ringer were gassed with 5% CO2 in oxygen. b Solutions for normothermic perfusion. c Cryoprotectant vehicle.
Solutions
METHODS
Preparation of Corneas New Zealand White rabbits (2-3 kg) were killed by an intravenous overdose of pentobarbitone sodium, The eyes were enucleated and the corneas removed by the method of Dikstein and Maurice (5). The corneas were secured on plastic support rings to minimize damage through wrinkling and distortion.
The composition of glutathione-bicarbonate Ringer (GBR), which is adequate for maintaining endothelial function for several hours during normothermic perfusion, is given in Table 1. This solution was supplemented with 1% v/v newborn calf serum (17). Two other versions of the solution were prepared: namely, without sodium bicarbonate, and with only 5 mmol/liter sodium bicarbonate (GBR,) (Table 1). In the former solution, the lack of sodium bicarbonate was not compensated for by another solute; TABLE 2 Protocol A: Stepwise Addition and Dilution of 10% v/v Propane-l ,t-diol
Perfusion
Time
(hours)
FIG. 1. Change in mean stromal thickness of four corneas (?SEM) during perfusion immediately after comeal isolation. Perfusion with isosmotic GBR was interrupted by 2 hr of bicarbonate-free perfusion (open circles). The rate of swelling between 2.5 and 4 hr was 34 + 5 p.m/hr and the rate of thinning between 4.5 and 6 hr was 21 + 1 pm/hr.
Step
Propane- 1,2-diol (% v/v)
Duration (n-W
1
0
10
2 3 4 5 6 7 8 9
5 10 8 6 4 2 1 0
10 20 5 5 5 5 5 10
Temperature ec,
44
RICH AND
ARMITAGE
TABLE 3 Protocol 8: Stepwise Addition and Sucrose Dilution of 10% v/v Propane-L,Zdiol Duration (tin)
Temperature ec,
0
10
0 0 1 0
10 20 15 15
4 4 4 25 25
Step
Propane-l ,2-diol (% v/v)
Sucrose W
1
0
2 3 4 5
5 10 0 0
thus its osmolality was less than that of GBR. The latter solution, which contained 6% w/v bovine serum albumin but no newborn calf serum, was used as the cryoprotectant vehicle. Chondroitin sulfate (2.5% w/v; Sigma, grade III, 99% mixed isomers, sodium salt from whale or shark cartilage) or 2.5% w/v dextran (Sigma, av mol wt 40,000 Da) and 1 mol/liter sucrose were also included in this solution as indicated. The molar concentration of electrolytes in GBR5 was maintained constant in all solutions containing propane- 1,2-diol or sucrose .
from the eye: these corneas served as a control group. Each experimental group also comprised four corneas. Differences between groups both in mean stromal thickness at various times during perfusion and in rates of change in thickness were com-
Assessment of Endothelial Function
After exposure to propane-l ,Zdiol, the comeal endothelium was perfused (1.25 ml/ hr, 13 cm H,O) for 6 hr at 34°C. The, epithelial surface was covered with silicone oil (Dow Corning 200/20 cs) to prevent evaporation. Endothelial function was assessed by measuring stromal thickness using a specular microscope with a calibrated fine focus. Measurements were made by focusing in turn on the endothelium and the epithelium, and mean stromal thickness was calculated from duplicate measurements taken at three random sites. Measurements were made at 30-min intervals throughout the perfusion period. The initial 2 hr of perfusion was with GBR. The perfusate was then changed to the bicarbonate-free Ringer solution for a further 2 hr before returning to GBR perfusion for the final 2 hr. Four untreated corneas were perfused, as described, immediately after their isolation
300! 0
, . ,
1
2 Pwfuskm
, . , , , , , 3 Time
4
6
B
hxml
FIG. 2. Each graph represents change in mean corneal thickness of four corneas (*SEM) during perfusion following exposure to 10% v/v propane-l,Z-diol. The 6&r GBR perfusion was intermpted by 2 hr of bicarbonate-free perfusion (open circles). (A) Stepwise dilution (protocol A). The rate of swelling between 2.5 and 4 hr was 37 2 3 cLm/hrand the rate of thinning between 4.5 and 6 hr was 31 + 4 qn/hr. (B) Sucrose dilution (protocol B). The rate of swelling between 2.5 and 4 br was 28 k 2 pm/br and the rate of thinning between 4.5 and 6 hr was 18 + 2 &hr.
45
CORNEAL TOLERANCE OF PROPANE-1,2-DIOL
FIG. 3. Scanning electron micrograph of the endothelial surface of a cornea following exposure to 10% v/v propane-1,2-diol (protocol B) and a 4hr perfusion. The endothelial cells appear normal, i.e., uniform in size and shape with clearly defined intercellular cell borders. Bar = 25 pm.
pared by one-way analysis of variance, with the level of significance set at 5%, and the Welsch step-up procedure (20). Means are quoted with the standard error of the mean.
buffer (pH 7.3) containing 3 mmol/liter CaCl,. The corneas were subsequently washed in the same buffer prior to criticalpoint drying and gold sputtering. The corneas were examined in a Philips 505 scanning electron microscope.
Scanning Electron Microscopy
After the 6-hr perfusion the corneas were Exposure of Corneas to Propane-l ,2-diol fixed overnight at 4°C with 2.5% v/v gluImmediately after excision, corneas were taraldehyde in 0.1 mol/liter cacodylate immersed in 50 ml GBRS in polystyrene TABLE 4 Protocol C: Stepwise Addition and Three-Step Sucrose Dilution of 15% v/v Propane-1,2-dial Propane-l ,2-diol (% v/v)
Sucrose WI
Duration (min)
Temperature (“Cl
cl 5 10 15 5 0 0
cl 0 0 0 1 1 0
10 10 10 20 15 15 15
4 4 4 4 25 25 25
46
RICH AND ARMITAGE RESULTS
Perfusion of Freshly Isolated Corneas
At the start of perfusion, these corneas were 420 + 10 pm thick. During the initial 2 hr of perfusion with GBR, they thinned slightly to 410 2 6 pm (Fig. I). During the 30 min immediately following the changeover to the bicarbonate-free perfusate, the 2 O 3 5 6 corneas swelled to 475 + 10 pm: they then Perfuslon Tkne (hours) FIG. 4. Change in mean comeal thickness of four continued to swell at a lower rate of 34 f 5 When perfusion with GBR was recorneas (?SEM) during perfusion following sucrose p&r. dilution of 15% v/v propane-l ,2-diol (protocol C). The sumed, there was a drop in thickness of 47 6-hr GBR perfusion was interrupted by 2 hr of bicar+ 4 pm during the first 30 min after changebonate-free perfusion (open circles). The rate of swell- over: this was followed by a period of tbining between 2.5 and 4 hr was 37 + 7 CLm/hrand the rate ning at a lower rate of 21 * 1 pm/hr. The of thinning between 4.5 and 6 hr was 19 2 3 km/hr. initial rapid changes in thickness that folpots. The corneas were then exposed to fi- lowed the changeover of perfusate were in nal concentrations of propane-l ,2-diol of part due to the difference in osmolality beIO, 15, 25, or 30% v/v. To reduce osmotic tween the two perfusates. The subsequent stress during the removal of cryopro- lower rates of swelling and thinning were tectant, the concentration of propane- similar to those reported by Hodson and 1,Zdiol was reduced in steps OF by a “su- Miller (8) for corneas in which the endothecrose dilution” technique (15). The proto- lial bicarbonate pump was first inhibited cols for the addition and subsequent and then restored, respectively. removal of propane-l ,2-diol are given in Tables 2-8. Corneas were immersed in 50 Stepwise Dilution of 10% vlv (I.4 M) Propane-i ,2-dial ml of solution for each step, and exposure temperatures of 0, -5, and - 10°C were Corneas were exposed to 10% v/v proattained with ice/water slush, an alcohol pane-1,2-diol (PD) for 20 min at 4°C accordbath (FTS Multicool), and a controlled-rate ing to protocol A (Table 2). Figure 2A freezer (Planer Biomed Kryo lo), respec- shows that during the initial 2 hr of GBR tively. perfusion, these corneas thinned slightly to
J.-v-T-
TABLE 5 Protocols D and E: Exposure of Corneas to 25% v/v Propane-l ,2-dial at WC (D) or at - 5°C (E)”
Step
Propane- 1,Zdiol (or0v/v) 0 5 10 15 25 LO 0 0
Sucrose (Ml
Temperature (X’)
Duration bin)
D
E
10 IO 10 20 20 15 15 15
4 4 4 4 0 0 25 25
4 4 4 4 -5 0 25 25
n The propane-1,2-diol was removed by a three-step sucrose dilution. These protocols were repeated with all solutions containing 2.5% w/v chondroitin sulfate or 2.5% w/v dextran.
CORNEAL TOLERANCE OF PROPANE-1,ZDIOL
47
FIG. 5. Scanning electron micrograph of corneal endothelium following exposure to 25% v/v propane-l ,2-diol (protocol E) and 6 hr perfusion. A central large hexagonal cell (a normal occurrence in rabbit endothelium) contrasts with the large cell to the right which is irregular in shape and appears to have resulted from intercellular membrane breakdown or cell loss. Bar = 25 km.
compensate for the swelling that occurred during the removal of PD. The corneas then demonstrated a period of swelling during perfusion with bicarbonate-free Ringer, followed by a period of thinning when perfusion with GBR was resumed. This behavior was indicative of a functioning endotheliurn. Cornea1 thickness and the rates of swelling and thinning were similar to the control values throughout the perfusion. All four corneas showed signs of endothelial disruption at the beginning of the perfusion in that dark areas were present where cells had apparently fallen away from Descemet’s membrane. This soon reorganized and a confluent endothelial mosaic was maintained for the remainder of the perfusion period. Sucrose Dilution of 10% vlv (1.4 M) Propane-i,2-dial
Corneas were exposed to 10% v/v PD for
20 min at 4°C according to protocol B (Table 3); but, unlike protocol A, the dilution of PD was carried out with a solution containing 1 mol/liter sucrose and in fewer steps. At the start of perfusion, these corneas were thinner than the controls (P < O.OS),presumably because they were partially dehydrated during the sucrose dilution, and Fig. 2B shows that they swelled rather than shrank (cf. Fig. 2A) during the first 2 hr of perfusion. During the ensuing perfusion, however, the mean thickness and rates of swelling and thinning were not significantly different from the control values. The reversal of swelling following cessation of the bicarbonate-free perfusion again demonstrated a functioning endothelial pump. The rate of thinning during the last 1.5 hr of perfusion was, however, lower (P < 0.05) than that of corneas exposed to 10% PD by protocol A (i.e., PD removed without sucrose dilution). All four corneas
48
RICH AND ARMITAGE
FIG. 6. Scanning electron micrograph of comeal endothelium exposed to 25% v/v propane-1,Zdiol (protocol E) in the presence of 2.5% w/v dextran. The endothelial cells are hexagonal and uniform in size with no visible sign of damage. Bar = 25 pm.
had a visibly intact endothelial mosaic throughout the 6-hr perfusion and scanning electron microscopy performed after the perfusion showed a normal endothelial mosaic with regularly sized hexagonal cells and intact intercellular cell borders (Fig. 3). Sucrose Dilution of 15% vlv (2.0 M) Propane-l ,2-diol
Corneas were exposed to 15%v/v PD for 20 min at 4°C according to protocol C (Table 4). Three of the four corneas in this group had a visibly normal endothelial mosaic throughout the perfusion period. AlI four corneas had a functioning endothelial pump (Fig. 4) despite the fourth cornea showing slight areas of endothelial damage during perfusion. These corneas were thinner than the controls at the start of perfusion (P < 0.05) but after 2 hr, the thickness and rates of swelling and thinning were not significantly different from the control values.
Sucrose Dilution of 25% vlv (3.4 M) Propane-l ,2-diol
Corneas were exposed to 25% v/v PD for 20 min at 0°C (protocol D) or at - 5°C (protocol E) (see Table 5). The four corneas exposed to 25% PD at 0°C were initially hazy but soon cleared to reveal a normal endothelial mosaic. By the end of the dhr perfusion, however, all four corneas were losing cells from the endothelial mosaic. One cornea showed no decrease in thickness on changeover from bicarbonate-free Ringer and went on to swell at 1.8 +n/hr during the ensuing GBR perfusion, which suggesteda lack of bicarbonate pump activity or a loss of barrier function. The mean rate of thinning for the three remaining corneas was 16.6 pm/hr (range: 30.2-22.8 prnlhr) . When corneas were exposed to 25% PD at - 5”C, three of the corneas, although iuitially hazy, soon cleared to show an intact endothelial mosaic that persisted for the re-
CORNEAL TOLERANCE OF PROPANE-1,2-DIOL
FIG. 7. Each graph represents change in mean corneal thickness of four corneas (-+SEM) following exposure to 25% v/v propane-1,2-diol with 2.5% w/v chondroitin sulfate. The 6-hr GBR perfusion was interrupted by 2 hr of bicarbonate-free perfusion (open circles). (A) Exposure to 25% v/v propane-1,2-dial plus chondroitin sulfate at 0°C. The rate of swelling between 2.5 and 4 hr was 37 & 2 Frn/hr and the rate of thinning between 4.5 and 6 hr was 11 * 2 prn/hr. (B) Exposure to 25% v/v propane-l ,2-diol plus chondroitin sulfate at - 5°C. The rate of swelling between 2.5 and 4 hr was 35 + 1 prn!hr and the rate of thinning between 4.5 and 6 hr was 15 + 4 cl.m/hr.
mainder of the perfusion. These three corneas demonstrated endothelial pump activity during the final 1.5 hr of perfusion by thinning at a mean rate of 17.8 cl,m/hr (range: 14.4-22.2 &hr). The fourth cornea, however, swelled during the final GBR perfusion, and had a persistently hazy appearance. Scanning electron microscopy of the corneas following perfusion showed some large irregular cells resulting from the breakdown of intercellular ceil membranes or cell loss (Fig. 5). Protocol E was repeated with the addition of 2.5% w/v dextran to all vehicle so-
49
lutions. Three corneas were subjected to this treatment, two of which showed a functioning bicarbonate pump by thinning during the final 1.5 hr of perfusion at 18.6 and 19.2 kmlhr, respectively. The third cornea swelled slightly during the final GBR petfusion (2.4 pm/hr) although subsequent microscopy showed a good endothelial mosaic with uniformly sized hexagonal cells (Fig. 6). Consistently good results were obtained with protocols D and E, however, when 2.5% w/v chondroitin sulfate was added to all vehicle solutions. In each case the initially hazy endothelium soon cleared and stayed clear, retaining a normal mosaic for the remainder of the perfusion. AU of these corneas demonstrated a functioning endothelial pump (Fig. 7). Apart from their initial thickness at the start of perfusion, which was less than the control values (P < O.OS),the thickness of these corneas and their rates of swelling and thinning were not significantly different from the control values. Scanning electron microscopy of these corneas following the 6-hr perfusion showed a good endothelial mosaic with uniformly sized hexagonal cells and intact intercellular cell borders (Fig. 8). Sucrose Dilution of 30% v/v (4.1 M) Propane-l ,2-diol Corneas were exposed to 30% v/v PD plus 2.5% w/v chondroitin sulfate for 10 min at either - 5 or - 10°C according to protocols F, G, and H (Tables 6 and 7). Protocols H and I (Tables 7 and S), involving exposure to 30% v/v PD at - lO”C, were performed with 2.5% w/v dextran in place of chondroitin sulfate. In all experiments involving exposure to 30% v/v PD the corneas were initially thin but showed some signs of endothelial damage in that cells were missing from the mosaic. Despite variations in the temperature of exposure to 30% v/v PD and in the rate of dilution, after about 1 hr of perfusion all of these corneas became so cloudy that stromal thickness
50
RICH AND ARMITAGE
FIG. 8. Scanning electron micrograph of the endothelial surface of a cornea exposed to 25% v/v propane-1,2-diol (protocol D) in the presence of 2.5% w/v chondroitin sulfate. The endothelial mosaic appears undamaged; the hexagonal cells are regular in size with intact intercellular borders. Bar = 25 urn.
measurements were no longer possible. The corneas remained opaque for the rest of the perfusion period. Scanning electron microscopy performed after the 6-hr perfu-
sion revealed extensive damage to these corneas as exemplified by Figs. 9 and 10. Large irregular cells resulting from cell loss or membrane breakdown (Fig. 9) and se-
TABLE 6 Protocols F and G: Exposure of Corneas to 30% v/v Propane-1,2-diol for 10 min at -5°C” Duration (min)
Step
Propane-l ,Zdiol (% v/v)
Temperature (“C)
F
G
1 2 3 4 5 6 7 8 9 10
0 5 10 15 25 30 25 10 0 0
4 4 4 4 -5 -5 -5 0 25 25
10 10 10 10 i0 10 15 15 15 1.5
10 10 10 10 IO 20 20 20 20
10
QThe propane-l ,2-diol was removed by sucrose dilution in 4 x 15-min steps (protocol F) or 4 x 20-min steps (protocol G). All solutions contained 2.5% w/v chondroitin sulfate.
CORNEAL
51
OF PROPANE-1,2-DIOL
TOLERANCE
TABLE 7 Protocol II: Exposure of Corneas to 30% v/v Propane-l ,2-diol at - 10°C for 10 mine Step
Propane- 1,2-diol (% v/v)
Sucrose (Ml
Temperature 03
Duration bin)
1
0
cl
4
10
2 3 4 5 6 7 8 9 10
5 10 15 25 30 25 10 0 0
0 0 0 0 0 1 1 1 0
4 4 4 -5 -10 -5 0 25 25
10 10 10 10 10 20 20 20 20
a The orouane-1 .2-diol was removed bv sucrose dilution in 4 X 20-min steps. All solutions contained 2.5% w/v _ chondrokn sulfate or 2.5% w/v dextran.
age than swelling (2); hence any damage due to osmotic stress is more likely to occur during the dilution of cryoprotectant. Osmotic stress during removal of the cryoprotectant can be mitigated by performing dilutions in steps or by balancing the decrease in concentration of extracellular permeating cryoprotectant by the simultaneous addition of a nonpermeating solute such as sucrose (so-called sucrose dilution). The advantage of sucrose dilution is that there can be larger decrements in the concentration of permeating cryoprotectant at each dilution step, thereby reducing the time that cells are exposed to potentially
verely contorted intercellular borders (Fig. 10) were typical in corneas exposed to 30% v/v PD under these conditions. DISCUSSION
High concentrations of permeating cryoprotectants can cause cellular damage by toxic effects or by osmotic stress created during addition and removal of cryoprotectant. Osmotic imbalances cause shrinkage of cells during addition of cryoprotectant and swelling during cryoprotectant removal. Cornea1 endothelial cells have been shown to be more tolerant of shrink-
TABLE 8 Protocol I: Exposure of Corneas to 30% v/v Propane-l ,2-diol at - 10°C for 10 mm* Step 1
2 3 4 5 6 7 8 9 10
Propane-l ,2-dial (% v/v)
Sucrose @fJ
0 5
0 0 0 0 0 0
10 15 25 30 15 5 0 0
1 1 1 0
Temperature eo 4 4 4 4 -5 - 10 0 25 25 25
Duration (mid 10 10 10 10 10 10 15 15 15 15
n The propane-1,2-diol was removed by sucrose dilution in 4 x 15min steps. All solutions contained 2.5% w/v dextran .
52
RICH
AND
ARMITAGE
FIG. 9. Scanning electron micrograph of corneal endothelium exposed to 30% v/v propane-1,2-diol (protocol F) with 2.5% w/v chondroitin sulfate. Irregularity in cell size and shape indicates endothelial disruption. Bar = 25 pm.
toxic concentrations of cryoprotectant (6, 19). Twenty minutes of exposure to 10% (1.4 M’) or 15% (2.0 AI) v/v propane-1,Zdiol at 4°C was well tolerated by corneal endothelium. Twenty minutes of exposure to 25% v/v (3.4 M) propane-1,2-diol, on the other hand, did cause some endothelial damage at both 0 and - 5°C. The addition of 2.5% w/v chondroitin sulfate to the cryoprotectant vehicle solution ameliorated the damage caused by exposure to 25% v/v propaneI ,2-diol. The substitution of chondroitin sulfate with 2.5% w/v dextran (which is similar in molecular weight to chondroitin sulfate) also had a beneficial effect; but dextran was not as effective as chondroitin sulfate in reducing damage caused by propane-l ,Zdiol. Chondroitin sulfate occurs naturally in comeal stroma (14) and is included in short-term storage solutions for human corneas because of an apparently
beneficial effect on the endothelium (IO, 11, 21), The corneas exposed to 10, 15, and 25% v/v propane-1,2-diol (the latter in the presence of 2.5% w/v chondroitin sulfate) showed no differences at the 5% level from the freshly isolated corneas in their rates of swelling and thinning when bicarbonate was restored to the perfusate. The rate of thinning of edematous corneas is dependent on the balance between the rate of pumping of bicarbonate ions by the endothelium and the passive leak of water and solute across the endothelium into the stroma. The leak is governed by the passive permeability of the endothelial monolayer and by the stromal swelling pressure, which is inversely related to corneal thickness (14). The similarity of swelling rates in these corneas exposed to propane- 1,2-diol suggested that the passive endothelial permeabilities, and hence the endotheIial barrier properties,
CORNEAL TOLERANCE OF PROPANE-1,2-DIOL
53
FIG. 10. Scanning electron micrograph of cornea endothelial surface following exposure to 30% v/v propane-1,2-diol (protocol H) with 2.5% w/v chondroitin sulfate. There is evidence of extensive damage to the endothelial mosaic. The intercellular borders are contorted resulting in apparent “holes” between some cells. Bar = 25 pm.
were similar to control values. Since these corneas were also of similar thickness to the controls at the beginning of the second period of GBR perfusion, the subsequent thinning (after the initial rapid fall in thickness) was presumed to be the result of endothelial pump activity. Exposure of 30% v/v (4.1 M) propane1,2-diol for 10 min at -5°C was not tolerated by corneal endothelium. Neither increasing the duration of the dilution steps from 15 to 20 min to reduce osmotic stress nor decreasing the propane-l ,2-diol concentration in larger steps to reduce the time of exposure to potentially toxic levels of the cryoprotectant improved endothelial function. In a further attempt to reduce toxicity of propane-l,Zdiol, exposure to 30% v/v was performed at a lower temperature ( - 10°C). There was, however, still no improvement in endothelial function, It is also possible that the use of higher concentrations of chondroitin sulfate may improve the function of these corneas.
Two molar propane-1,Zdiol has been reported to be damaging to comeal endothelium after only 5 min exposure at 20°C (13); however, a serial dilution technique was used. Limited tolerance of other tissues to propane- 1,2-diol has recently been reported, Reduced survival of human blood platelets occurs following exposure to greater than 2 mol/liter propane-1,2-diol at 20°C (3). Kidneys have been reported to tolerate only 2.5-3 M (7,18). Although corneas tolerated exposure to 25% v/v propane-l ,Zdiol, they were damaged by higher concentrations under the conditions studied in these experiments. Apparent vitrification of 35% v/v (4.8 M) propane-l,2-diol in the presence of 2.5% w/v chondroitin sulfate plus 6% w/v bovine serum albumin was achieved by plunging a glass vial containing 5 ml of the solution into liquid nitrogen (unpublished observation). Although no ice crystals were visible, nucleation would almost certainly have occurred upon cooling, necessitating ex-
54
RJCH AND ARMITAGE
tremely high warming rates to avoid devitrifrcation and recrystallization. Such high cooling and warming rates are impracticable for the cornea. Thus, a comeal vitrification solution based solely on this cryoprotectant is currently not feasible. However, since 25% v/v propane-1,Zdiol is well tolerated, this compound could form a major component of a vitrification solution containing a mixture of cryoprotectants. ACKNOWLEDGMENTS S.J.R. is supported by a B&ton Research Studentship from the National Eye Research Centre. W.J.A. is supported by a Royal National Institute for the Blind Research Fellowship. The work was funded by a grant from the T. F. C. Frost Charitable Trust. We thank Professor David Easty for his continued support of this work. REFERENCES 1. Armitage, W. J. Survival of comeal endothelium following exposure to a vitrification solution. Cryobiology26, 318-327 (1989). 2. Armitage, W. J., Moss, S. J., and Easty, D. L.
Effects of osmotic stress on rabbit comeal endothelium. Cryobiology 25, 425439 (1988). 3. Amaud, F., and Pegg, D. E. Attempts to cryopreserve platelets with propylene glycol. CryobioC Qgy 24, 587(1987). 4. Boutron, P., and Kaufmann, A. Stability of the
amorphous state in the system water-1,2propanediol. Cryobiology 16, 557-568 (1979). 5. Dikstein, S., and Maurice, D. M. The metabolic basis to the fluid pump in the cornea. J. fhysiol 221, 29-41 (1972).
6. Fahy, G. M., MacFarlane, D. R., Angell, C. A., and Meryman, H. T. Vitrification as an approach to cryopreservation. Cryobiology 21, 407-426(1984). 7. Halasz, N. A., and Collins, G. M. Studies in cryo-
preservation, II. Propylene glycol and glycerol. Cryobiology 21, 14147 (1984). 8. Hodson, S., and Miller, F. The bicarbonate ion pump in the endotbelium which regulates the hydration of rabbit cornea. J. Physiol. 263,563577 (1976). 9. Jutte, N. H. P. M., Heyse, P., Jansen, H. G.,
Bruining, G. J., and Zeilmaker, G. H. Vitriftcation of human islets of Langerhans. Cryobiology 24,403-411 (1987). 10. Kaufman, H. E., Vamell, E. D., Kaufman, S.,
Beuermau, R. W., and Barron, Et. A. K-Sol corneal preservation. Amer. J. Ophthaimol. 100, 299-304(1985).
11. Lindstrom, R. L., Skelnik, D. L., Mindrup. E. A., and Doughman, D. J. Organ culture corneal preservation with chondroitin sulphate. In “ARVO Abstracts.” Invest. Ophthalmol. Vis. Sci. Suppl., p. 266 (1984). 12. MacFarlane, D. R. Physical aspects of vitrification in aqueous solutions. Cryobiology 24, 181195 (1987). 13. Madden, P. W. The assessment of endothelial integrity by scanning electron microscopy and fluorescein diacetate staining following treatment with cryoprotective additives. Curr. Eye Res. 8, 17-36 (1989). 14. Maurice, D. M. The cornea and sclera. ZR “The Eye” (H. Davson, Ed.), 3rd ed., Vol. lb, Chap. 1, pp. L-158. Academic Press, Orlando, FW London, 1984. 15. Mazur. P. Fundamental cryobiology and the preservation of organs by freezing. In “Organ Preservation for Transplantation” (A. M. Karow and D. E. Pegg, Eds.), 2nd ed., pp. 143-175. Marcel Dekker, New York, 1981. 16. Mazur, P. Freezing of living cells: Mechanisms and implications. Amer. J. Physiol. 247, 125142 (1984). 17. Neuwirth Lux, O., and Dikstein, S. Survival of isolated rabbit cornea and free radical scavengers. Curr. Eye Res. 4, 15S-154 (1985). 18. Pegg, D. E., Jacobsen, I. A., Diaper, M. P., and Foreman, J. Perfusion of rabbit kidneys with solutions containing propane-l ,2-dial. Crjlobiology 24, 42B-428 (1987). 19. Rail, W. F., and Fahy, G. M. Ice-free cryopreservation of mouse embryos at - 196°Cby vitrllication. Nature (London) 313, 573-575 (1985). 20. Sokal, R. R., and Rohlf, F. J. “Biometry,” 2nd ed. Freeman, New York, 1981. 21. Stein, R. M., Boume, W. M., andcampbell, R. J. Chondroitin sulfate for comeal preservation at 4°C. Evaluation by electron microscopy. Arch. Ophrhalmol. 104, 13%1361(1986). 22. Takahashi, T., Hirsch, A., Erbe, E. F., Brass, J. B., Steere, R. L., and Williams, R. J. Vitrification of human monocytes. Cryobiology 23, 103-115 (1986). 23. Taylor, M. J. Clinical cryobiology of tissues: Preservation of corneas. Cryobiology 23, 323-353 (1986). 24. Taylor M. J., and Pegg, D. E. The effect of ice formation on the function of smooth muscle tissue stored at - 21 or -60°C. Cryobiology 20, 36-N-I (1983).
