- Email: [email protected]
Lysosomal Enzyme Levels in Corneal Storage Media
Lysosomal Enzyme Levels in Corneal Storage Media 1(-501 Versus McCarey-Kaufman KIRK M. MORGAN, MD, WILLIAM E. BRUNER, MD, RICHARD E. WYSZYNSKI, MD, MARIANNE S. DiMARCO, MD
Abstract: Evidence indicates that lysosomal enzymes can carry out comeal autolysis during corneal storage and that they are damaging to the comeal endothelium. The authors investigated the release of lysosomal enzymes into two corneal storage media (K-Sol and McCarey-Kaufman [M-K]) by paired human donor corneas during 4°C storage. The authors also studied the interaction of these media with lysosomal enzymes from human cornea. K-Sol and M-K stimulated (P < 0.01) both t3-glucuronidase and a-galactosidase about equally. 13N-Acetyl-glucosaminidase, a major catabolic enzyme of the cornea, was inhibited by the chondroitin sulfate in K-Sol by over 90% (P < 0.01). Corneas stored in M-K released more lysosomal enzymes than corneas stored in K-Sol. At 4 days, the values approached significance (P < 0.06) and by day 10 significantly higher values were found in the M-K media (P < 0.01). Both storage methods showed a linear release. Individual corneas were found to vary in their release rates. Whether corneas that release more enzyme will show higher endothelial cell loss or produce less successful penetrating keratoplasty grafts deserves further study. [Key words: t3-N-acetyl-glucosaminidase, t3-glucuronidase, corneal storage, lysosomal enzyme, penetrating keratoplasty.] Ophthalmology 95:14981503, 1988
A number of corneal preservation methods are currently in use and being investigated. Classically, whole globe storage was the first method used and is still being used in many parts of the world. The major disadvantage with this method is that the cornea should be used within the first 24 to 36 hours. I McCarey-Kaufman (M-K) medium allows corneas to be stored successfully for up to 4 days.2-4 Even with this extension in storage time, most Originally received: November 13, 1986. Revision accepted: July 26, 1988. From Case Westem Reserve University, VA Medical Center, Cleveland. Presented in part at the American Academy of Ophthalmology Annual Meeting, New Orleans, November 1986. Supported by a grant from Medical Research Service, VA Medical Center. Cleveland, Ohio. Reprint requests to William E. Bruner, MD, 1611 South Green Rd. Cleveland, OH44121.
1498
corneal surgeons still have to perform penetrating keratoplasty on a semi-emergent basis. It would be advantageous to both the patient and the surgeon if surgery could be electively scheduled. This has contributed to the current interest in inexpensive and simple methods to extend corneal storage past this 3- to 4-day limit. Studies by a number of Japanese authors in the late 1960s indicated that the addition of the macromolecular colloid chcndroirin sulfate was beneficial to the cornea during storage. 5- 8 Kaufman et al,9 in 1984, reported that the replacement of the 5% dextran in M-K with 2.5 to 10% chondroitin sulfate gave the best storage results for 14 days at 4°C. In 1985, Kaufman et al lO introduced KSol, which contains "purified" chondroitin sulfate and showed that clear grafts could be obtained using this method for up to 16 days of storage. Electron microscopic studies by Stein et alII and Stein and Laibson 12 confirmed K-Sol's superior endothelial preservation at 4°C for up to 2 weeks. Bourne l3 reported equal transplantation re-
MORGAN el al
•
ENZYME LEVELS: K-SOL VERSUS M-K
suIts with K-Sol, at a mean storage time of6.8 days, when compared with M-K storage with a mean time of39 hours (both at 4°C). Furthermore, Bourne l3 showed that the highest corneal cell losses occurred after 2 days of M-K storage and after 10 days of K-Sol storage. We became interested in studying the concentration of lysosomal enzymes in storage media because of reports that these enzymes may be a cause of corneal autolysis and endothelial damage during storage. Basu and Hasanyl4 reported that lysosomal enzymes can carry out corneal autolysis to varying degrees depending on the conditions of storage and that the loss of corneal endothelial cells was directly related to the concentration of the hydrolytic enzymes. 15 Spencer et al l6 showed that the addition of corticosteroids to the preservation media reduced the release oflysosomal enzymes. Basu et all? reported that the addition of steroid to M-K increased the number of viable endothelial cells remaining at 4 days of 4°C storage and believed that this was because of lysosomal membrane stabilization by the corticosteroid. Additionally, Arya et al l8 ,l9 showed that a lysosomal extract from polymorphonuclear leukocytes was injurious to endothelial cell in vivo and in vitro. Hydrolytic lysosomal enzyme levels in corneal storage media may therefore provide an index of corneal autolysis which might relate to the performance of the graft at the time of transplantation. As an initial step to investigate this possibility, we measured lysosomal acid hydrolase levels, to see what changes occur with storage time and to compare M-K with K-Sol. Additionally, the interactions of M-K and K-Sol with this class of enzyme has been studied.
METHODS SAMPLING METHODS
Matched pairs of corneas (2-mm corneoscleral rims) were obtained from the Cleveland Eye Bank which had contraindications for transplantation (most often age or infection at a site other than the eye). Postmortem times were no greater than 9 hours (average, 6 hours) and all corneas appeared normal. The average age of the donor was 69 years. One cornea of each pair was placed in 20 ml ofK-Sol (Cilco, Bellevue, WA) or M-K (Aurora, Buffalo, NY) storage solution. One cornea was transferred to K-Sol after 1 day of storage at 4°C in M-K medium. All corneas were stored at 4°C during sampling, and matched pairs were always handled in the same manner. Two sampling methods were used: daily or every other day, 0.50 ml of media was removed under aseptic conditions or 1.0 ml was removed with replacement of fresh media. At the end of the sampling period all media batches were cultured on blood-augerose culture medium at 37°C for 72 hours. Samples were immediately frozen at -20°C until the analyses could be performed.
MIXING EXPERIMENTS
Fresh human corneal tissue from three donors were separately homogenized in 1.0 ml of 0.25 M sucrose, 0.05 M TRIS-HCI buffer at pH 7.4 and 0.1% Triton X-lOO using a Potter Elvehjem-type ground glass homogenizer (Fisher, Pittsburgh, PA). The homogenates were centrifuged at 1500 X g for 10 minutes and the supernatants saved and frozen for enzyme assay. The analysis of each was carried out in triplicate as described below. ENZYME ASSAY
Solutions of the following three 4-methyl-umbelliferyl glycosides (Sigma, St Louis, MO) were prepared in distilled water and used as substrates: a-galactoside, 0.5 mM20; (3glucuronide, 5 mM21; and (3-glucosaminide, 5 mM.22 In conducting the storage experiments, a known volume (determined for each enzyme studied to bring the values on scale) ofK-Sol or M-K medium was added to distilled water to reach a total volume of 100 ~l. In conducting the mixing experiments, 10 ~l of corneal supernatant was added to 90 ~l of either distilled water, M-K, or K-Sol to reach this 100-~1 value. To these were added 100 ~l of citrate-phosphate buffer of appropriate pH (pH used was based on the data of Bruner and Stark23 ) and 100 ~l of appropriate umbelliferyl-glycoside substrate so that a total volume of 0.30 ml in a 7.5 X 1-cm Pyrex test tube was achieved. The citrate-phosphate buffers were prepared with 0.10M citric acid and 0.20-M sodium phosphate, as described by Gomori.24 We performed pH measurements on the assay solutions to assure that dilution did not affect the pH of the reaction mixture. All measurements were performed in duplicate or triplicate, and under conditions of linearity with incubation time and substrate concentration. 22 Incubation was carried out for 30 to 45 minutes in a water bath at 37°C, and the reaction was stopped by adding 1.0 ml of0.40-M glycine-sodium hydroxide buffer, pH 10.3. This method is based on that described by Leaback and Walker. 22 Enzymatic activity was quantitated by determining the amount of umbelliferone released by the enzyme, as measured using an American Instrument Company (AMINCO) fluorometer Model J4-7439 with a primary filter of Corning 7-60 (Corning, NY) and a secondary filter (AMINCO J4-7116). Umbelliferone standard curves using crystalline 4-methyl-umbelliferone (Eastman, Rochester, NY) were performed with each experiment. CALCULATIONS AND STATISTICS
For each sample, other than the first sample, the activity which was removed had to be added back so that the values would approximate that which would have been found had the media not been sampled previously. The activity (nmole/hour) was calculated for the amount removed and added back to that which was found for the total sample then converted to activity per milliliter based on the total volume in the storage vessel for that data
1499
OPHTHALMOLOGY
•
NOVEMBER 1988
•
VOLUME 95
•
NUMBER 11
Mixing Experiment: Effect of Media on Activity Fig I. Effect of mixing M-K or K-Sol on the average activities of three lysosomal enzymes from normal human cornea. Both /3-g1ucuronidase and a-galactosidase were stimulated (P < 0.0 I) to an equal extent (P < 0.05), whereas the activity of /3-N-acetyl-g1ucosaminidase was inhibited (P < 0.0 I). This experiment shows that i3-g1ucuronidase may be used to compare release rates in M-K and K-Sol media.
4
Average Activity (Ilmol.al mg prolelnl hr .)
3 •
2
Corneal Activity
II Activity
with MK
m Activity
with
K·Sol
o Bela·Glucuron
NAGase
Alpha·Gal
ENZYME
NAGase_Beta · N· Acetyl·Glucosaminidase
Beta-Glucuronidase: MK vs K-Sol
3.5 ..-..- ...-.-..-....-... Fig 2. The average activity (nmoles/ml/hour) of 13glucuronidase released by matched pairs of human donor corneas stored at 4°C in either M-K or K-Sol as a function of storage time is shown. Data points represent average values from at least three paired corneas with some data points representing data from as many as seven paired corneas. Lines were generated by linear regression.
3.0
2.5
'1.-...-_. . -..-... .-'...-..-.. -..-..-.... .-. ..r...-...-....-....-...-....-.,..---~--c-.---. .........................:........................ .
~
- --
! - i-- t - - - - i --.:...ti-----""----I
2 . 0~ --------+-------4--------t-----~,r-------~---~--- 1
r;MKI
~
0.5
0.0
+-......---i--....--+--..--t----..,....---t---.---+---.---t 15 30 10 20 25 o 5 Time(Oays)
point. Statistical analysis was performed using a two-tailed Student's t test based on paired differences.
RESULTS MIXING EXPERIMENTS
Figure 1 shows the effect of M-K and K-Sol on the average activities of three lysosomal enzymes from three 1500
normal human corneas. McCarey-Kaufman and K-Sol stimulated (P < 0.01) both ,8-glucuronidase and a-galactosidase about equally (P > 0.05). ,8-N-acetyl-glucosaminidase, however, was stimulated to some extent by MK media (P > 0.05) but was inhibited by K-Sol by over 90% (P < 0.01). Since there was equal stimulation of,8glucuronidase by both M-K and K-Sol, this enzyme was used to compare release by the two media.
MORGAN et al
•
ENZYME LEVELS: K-SOL VERSUS M-K
Beta-Glucuronidase: Activity In MK
5.0
0
4.5
I
4.0
Activity (nmolesl ml/hr)
A /·4
I
3.5 3.0
1
2.5
/'
A"1i.~7;6 ...-
~..__Ii;
.-
.-I ~!-
.1. ....---•••
2.0 1.5
A~---· ~4"1'
.~
y"
1.0
o
./1
d____ O
1
0-
'0-
1
0.0
o
JI ~." ..-
-~/
-.~
.. ~~~~t0.0-0- -0.0..-0::;0 0'0-0.0'0
,,.a.p ~.... O~
0.5
)'0
""'".
Fig 3. The release of ,B-glucuronidase by individual corneas stored at 4°C in M-K media as a function of storage time is shown. Corneas differ in their release rates, with some corneas showing a more rapid rise in activity than other corneas.
10
5
20
15
25
30
Days
B.ta-Glucuronid... : Activity In K-Sol 5.0 4.5 4.0 3.5 Activity
3.0
(nmole.' mllhr)
2.5
_.I.
2.0 1.5 1.0
~.l.J
I
. ;.•. ~
i
r
-r-.-. ,;.J
....
~
•I
• ..-.J----O·O-O-O-O· 9"-.~·_.·.··--~"'[j:d:0-1-0= 0=0=0 i
0.5
00
va.-
J
I ____........
.e::..I.-.l.--.1.-.1.-- _:I: . , .1..../ ~•.I..a.- I _.,....~or-. ~Ii.
• ~.,.~ ~._.-.-.-.,.IJ. ..-~ ~ 0 l"b ~•.• -•.•
.I.'
Fig 4. The release of ,8-glucuronidase from individual corneas stored at 4°C in KSol as a function of storage time is shown. Individual variation in the release rate is shown as was found with MK storage but with less overall activity released.
~o_o
o
5
10
LYSOSOMAL ENZYME ACTIVITY IN CORNEAL STORAGE MEDIA
Figure 2 shows the average activity (nmoles/ml/hour) of ,8-glucuronidase released by matched pairs of human donor corneas in the two storage media measured as a function of storage time. Data points represent average values from at least three paired corneas with some data points representing data from as many as seven paired corneas. Lines were obtained by linear regression. It is noteworthy that the activities become disparate by 3 days of storage. At 4 days, the values approached significance (P
20
25
30
< 0.06) and by day 10 significantly higher levels of ,8glucuronidase were found in the M-K media (P < 0.01». This level of significance continued up to 30 days of storage when sampling was terminated. Figures 3 and 4 show the release of ,8-glucuronidase by individual corneas for M-K and K-Sol, respectively. These figures demonstrate that individual corneas release activity at different rates. No correlation to age of donor, postmortem time, or enucleation to preservation time could be demonstrated. Some corneas stored in M-K released enzyme at rates comparable with K-Sol, whereas some released considerably more enzyme. 1501
OPHTHALMOLOGY
•
NOVEMBER 1988
CULTURE RESULTS
At 72 hours, all of the cultures showed no growth indicating that contamination by microorganisms did not contribute to enzyme activities.
DISCUSSION Basu et a1 14 •15 have shown that lysosomal enzymes are important to the endothelial viability of stored corneas. Arya et al 18.19 have shown lysosomal enzymes to be injurious to endothelial cells in vivo and in vitro. Spencer et al l6 showed that an inhibition of lysosomal enzyme release occurred when steroids were added to storage media, and Basu et al 17 found that steroids increased endothelial viability, suggesting that the inhibition of enzyme release accounted for this finding. We investigated the interactions of M-K and K-Sol with lysosomal enzymes and their release from stored corneas. Additionally, these studies represent preliminary experiments on the possibility for the use of lysosomal enzyme levels in corneal storage media as a screen for excessive autolysis as a predictor of excessive endothelial cell loss. In order to determine if enzyme levels could be used to legitimately compare M-K and K-Sol storage, we first studied the interaction of these media with activities of ground human corneal supernatants. Both M-K and KSol stimulated (P < 0.01) the activities of ,B-glucuronidase and a-galactosidase to an equal extent. This could be due to the presence of cations or protein in the preservation media. McCarey-Kaufman stimulated the activity of {3N-acetyl-glucosaminidase to some extent (P > 0.05), whereas K-Sol was found to inhibit this enzyme (P < 0.01). {3-N-Acetyl-glucosaminidase is an important lysosomal enzyme in the metabolism of proteoglycans since five of the seven species ofproteoglycans (i.e., chondroitin4-sulfate, chondroitin-6-sulfate, dermatin sulfate, hyaluronic acid, and keratin sulfate I and II), contain {3-hexosaminidic linkages that are degraded by this enzyme. 25 Keratin sulfate and chondroitin sulfate would be most important since they are the major proteoglycans of the cornea. 26 The chondroitin sulfate in K-Sol probably inhibits the activity of this enzyme by acting as a substrate inhibitor since it has been shown to degrade this proteoglycan.27 K-Sol might inhibit corneal autolysis by reducing the catabolism of corneal proteoglycans by this enzyme. Since {3-glucuronidase was stimulated to the same extent by both media, this enzyme was used to compare lysosomal enzyme release into the two media. Lysosomal enzyme release into M-K and K-Sol media during 4°C storage was investigated. As shown in Figure 2, we found that the rate of release in both media to be linear. Corneas stored in K-Sol released less lysosomal enzymes, with the difference approaching significance by day 4 (P < 0.06) and by day 10 and beyond significantly more activity was found in the M-K media (P < 0.01). It is interesting to note that average {3-glucuronidase activities in M-K media at day 3 (0.75 nmoles/ml/hour) 1502
•
VOLUME 95
•
NUMBER 11
were not equal in K-Sol until day 10 (0.78 nmoles/ml/ hour). This finding seems to correspond well with the observations of Bourne l3 that high endothelial cell loss was seen in corneas stored more than 2 days in M-K and 10 days in K-Sol. Two sampling methods were used since each method approximates activity values which would have been obtained for each day had the media not been sampled previously. With one method, we removed samples without adding back fresh media. This method might be expected to underestimate the true value since during the sampling period the total volume was decreased and therefore the concentration of enzyme would be increased, possibly leading to reduced release because of a smaller concentration gradient. The other method in which fresh media was added back would cause a dilution of activity and therefore an increased concentration gradient resulting in increased release. By using both methods and combining the data, we hoped to more accurately approximate the values which one might obtain for a particular day of storage. Data from the two methods were not significantly different. Since we are interested in the possible use of these enzyme values as a screen for unacceptable corneal autolysis, it was important to see if individual corneas differed in their release rates. Figures 3 and 4 show the enzyme release curves for five individual corneas stored in M-K and KSol, respectively. We can see that corneas vary in their release rates with some corneas showing slow release during the early storage period, whereas others show a more rapid rise in activity. No correlation to age of donor, postmortem time, or enucleation to preservation time could be demonstrated. Whether corneas that show more release would have also showed higher endothelial cell losses after penetrating keratoplasty cannot be commented on with the current data. This possible use of lysosomal enzyme analysis deserves further investigation. In this study, we did not show whether the enzymes were coming from the endothelium or epithelium. In a separate study of 12 cat corneas stored in K-Sol at 4°C for 14 days with and without their epithelium present (data not shown), significantly higher enzyme levels were observed from corneas with their epithelium intact when compared with those without epithelium, suggesting that in fact most of the activity is coming from epithelial cells. This does not, however, negate the usefulness of enzyme analysis because the literature suggests that these enzymes are potentially damaging to the endothelium regardless of their origin. One might also reasonably assume that enzyme release from epithelium would be paralleled by the same release from endothelium. In 20 ml of storage solution, one might question whether the enzyme concentrations observed could become high enough to show a toxic effect on the endothelium. Diffusion of these enzymes through the stroma could result in higher concentrations of enzyme affecting the endothelial cells at their attachment sites to the cornea in areas where certain acidic microenvironments might exist.
MORGAN et al
•
ENZYME LEVELS: K-SOL VERSUS M-K
Even before release of lysosomal enzymes into the media would occur one could envision intracellular release and damage to the individual cell concerned. However, in the current study, we have not demonstrated whether such a mechanism exists or whether the release of these enzymes indicates cellular damage or merely decreased metabolism at 4°C. Because these media are kept at 4°C and pH 7.4, one might not expect these acid hydrolases to be highly active even if certain acidic microenvironments existed. They may, however, serve as markers for the release of other lysosomal enzymes such as cathepsins which possess neutral pH optima and therefore may be active under these storage conditions. In summary, we have shown that K-Sol medium inhibits the release of lysosomal enzymes, which may be damaging to the corneal endothelium during corneal perservation, to a greater extent than does MK medium. KSol also inhibits the activity of ~-N-acetyl-glucosamini dase, an important enzyme in the catabolism of corneal proteoglycans. Corneas differ in their release rates of lysosomal enzyme in both K-Sol and M-K. Whether these enzyme levels will predict endothelial cell performance after penetrating keratoplasty remains to be seen.
REFERENCES 1. Binder PS. Eye banking and corneal preservation. In: Symposium on Medical and Surgical Diseases of the Comea: New Orleans Academy of Ophthalmology. St Louis: CV Mosby Co, 1980; 320-54. 2. McCarey BE, Sakimoto T, Bigar F. Ultrastructure of M-K and refrigerated moist chamber-stored corneas. Invest Ophthalmol Vis Sci 1974; 13: 859-63. 3. Van Horn DL, Shultz RO, DeBruin J. Endothelial survival in corneal tissue stored in M-K medium. Am J Ophthalmol1975; 80:642-7. 4. Aquavella JV, Van Horn DL, Haggerty, CJ. Corneal preservation using M-K medium. Am J Ophthalmol1975; 80:791-9. 5. Kuwahara Y, Sakanoue M, Hayashi M. et al. Studies on the long-term preservation of the cornea for penetrating keratoplasty. Nippon Ganka Gakkaizasshi 1965; 69:1751-840. 6. Nakano T. Studies on the viability of the cornea stored for the penetrating keratoplasty under various experimental conditions. Report III. Oxygen consumption of rabbit corneas stored in newly devised preservation medium containing various macromolecular colloids. Nippon Ganka Gakkaizasshi 1966; 70:662-74. 7. Mizukawa T. Corneal transplantation. J Jpn Med Assoc 1967; 58: 957-64. 8. Mizukawa T, Manabe R. Recent advances in keratoplasty with special
9.
10. 11.
12.
13.
14. 15.
16. 17.
18.
19. 20.
21.
22.
23. 24. 25.
26. 27.
reference to the advantages of liquid preservation. Folia Ophthalmol Jpn 1968; 19:1310-8. Kaufman HE, Varnell ED, Kaufman S. Chondroitin sulfate in a new cornea preservation medium [Letter]. Am J Ophthalmol 1984; 98: 112-4. Kaufman HE, Varnell ED, Kaufman S. et al. K-Sol corneal preservation. Am J Ophthalmol1985; 100:299-304. Stein RM, Bourne WM, Campbell, RJ. Chondroitin sulfate for corneal preservation at 4°C: evaluation by electron microscopy. Arch Ophthalmol1986; 104:1358-61. Stein RM, Laibson PRo Comparison of chondroitin sulfate to McCareyKaufman medium for corneal storage. Am J Ophthalmol 1987; 104: 490-3. Bourne WM. Endothelial cell survival on transplanted human corneas preserved at 4 C in 2.5% chondroitin sulfate for one to 13 days. Am J Ophthalmol 1986; 102:382-6. Basu PK, Hasany S. Autolysis of the cornea of stored human donor eyes. Can J Ophthalmol 1974; 9:229-35. Basu PK, Hasany SM, Ranadive, NS, Chapman M. Damage to corneal endothelial cells by lysosomal enzymes in stored human eyes. Can J Ophthalmol1980; 15:137-40. Spencer JA, Dixon WS, Ranadive NS, Basu PK. Factors in the survival of stored corneas. Can J OphthalmoI1977; 12:123-7. Basu PK, Hasany SM, Doane FW, Schultes K. Can steroid reduce endothelial damage in stored corneas? Effect on cell viability and ultrastructure. Can J Ophthalmol1978; 13:31-8. Arya DV, Mannagh J, Irvine AR Jr., Yuhasz Z. Effect of Iysosomes on corneal endothelium: an in vivo study. Invest Ophthalmol 1972; 11:655-61. Arya DV, Mannagh J, Irvine AR Jr. Effects of Iysosomes on cultured rabbit corneal endothelial cells. Invest Ophthalmol1972; 11 :662-7. Desnick RJ, Allen KY, Desnick SJ, et al. Fabry's disease: enzymatic diagnosis of hemizygotes and heterozygotes. J Lab Clin Med 1973; 81:157-71. Mead JAR, Smith IN, Williams RT. The biosynthesis of the glucuronides of umbelliferone and 4-methylumbelliferone and their use in fluorimetric determination of J3-glucuronidase. Biochem J 1955; 61 :569-74. Leaback DH, Walker PG. Studies on glucosaminidase 4. The fluorimetric assay of N-acetyl-J3-glucosaminidase. Biochem J 1961; 78: 151-6. Bruner WE, Stark WJ. Glycosidases of normal human donor cornea. Curr Eye Res 1982; 1:705-10. Gomori, G. Preparation of buffers for use in enzyme studies. Methods Enzymol 1955; 1: 138-46. Roden, L. Structure and metabolism of connective tissue proteoglycans. In: Lennarz WJ, ed. The Biochemistry of Glycoproteins and Proteoglycans. New York: Plenum Press, 1980; 285-6. Maurice DM, Riley MV. The cornea. In: Graymore CN, ed. Biochemistry of the Eye. New York: Academic Press, 1970; 30. Mahadevan S, Dillard CT, Tappel AL. Degradation of polysaccharides, mucopolysaccharides, and glycoproteins by lysosomal glycosidases. Arch Biochem Biophys 1969; 129:525-33.
1503
Abstract: Evidence indicates that lysosomal enzymes can carry out comeal autolysis during corneal storage and that they are damaging to the comeal endothelium. The authors investigated the release of lysosomal enzymes into two corneal storage media (K-Sol and McCarey-Kaufman [M-K]) by paired human donor corneas during 4°C storage. The authors also studied the interaction of these media with lysosomal enzymes from human cornea. K-Sol and M-K stimulated (P < 0.01) both t3-glucuronidase and a-galactosidase about equally. 13N-Acetyl-glucosaminidase, a major catabolic enzyme of the cornea, was inhibited by the chondroitin sulfate in K-Sol by over 90% (P < 0.01). Corneas stored in M-K released more lysosomal enzymes than corneas stored in K-Sol. At 4 days, the values approached significance (P < 0.06) and by day 10 significantly higher values were found in the M-K media (P < 0.01). Both storage methods showed a linear release. Individual corneas were found to vary in their release rates. Whether corneas that release more enzyme will show higher endothelial cell loss or produce less successful penetrating keratoplasty grafts deserves further study. [Key words: t3-N-acetyl-glucosaminidase, t3-glucuronidase, corneal storage, lysosomal enzyme, penetrating keratoplasty.] Ophthalmology 95:14981503, 1988
A number of corneal preservation methods are currently in use and being investigated. Classically, whole globe storage was the first method used and is still being used in many parts of the world. The major disadvantage with this method is that the cornea should be used within the first 24 to 36 hours. I McCarey-Kaufman (M-K) medium allows corneas to be stored successfully for up to 4 days.2-4 Even with this extension in storage time, most Originally received: November 13, 1986. Revision accepted: July 26, 1988. From Case Westem Reserve University, VA Medical Center, Cleveland. Presented in part at the American Academy of Ophthalmology Annual Meeting, New Orleans, November 1986. Supported by a grant from Medical Research Service, VA Medical Center. Cleveland, Ohio. Reprint requests to William E. Bruner, MD, 1611 South Green Rd. Cleveland, OH44121.
1498
corneal surgeons still have to perform penetrating keratoplasty on a semi-emergent basis. It would be advantageous to both the patient and the surgeon if surgery could be electively scheduled. This has contributed to the current interest in inexpensive and simple methods to extend corneal storage past this 3- to 4-day limit. Studies by a number of Japanese authors in the late 1960s indicated that the addition of the macromolecular colloid chcndroirin sulfate was beneficial to the cornea during storage. 5- 8 Kaufman et al,9 in 1984, reported that the replacement of the 5% dextran in M-K with 2.5 to 10% chondroitin sulfate gave the best storage results for 14 days at 4°C. In 1985, Kaufman et al lO introduced KSol, which contains "purified" chondroitin sulfate and showed that clear grafts could be obtained using this method for up to 16 days of storage. Electron microscopic studies by Stein et alII and Stein and Laibson 12 confirmed K-Sol's superior endothelial preservation at 4°C for up to 2 weeks. Bourne l3 reported equal transplantation re-
MORGAN el al
•
ENZYME LEVELS: K-SOL VERSUS M-K
suIts with K-Sol, at a mean storage time of6.8 days, when compared with M-K storage with a mean time of39 hours (both at 4°C). Furthermore, Bourne l3 showed that the highest corneal cell losses occurred after 2 days of M-K storage and after 10 days of K-Sol storage. We became interested in studying the concentration of lysosomal enzymes in storage media because of reports that these enzymes may be a cause of corneal autolysis and endothelial damage during storage. Basu and Hasanyl4 reported that lysosomal enzymes can carry out corneal autolysis to varying degrees depending on the conditions of storage and that the loss of corneal endothelial cells was directly related to the concentration of the hydrolytic enzymes. 15 Spencer et al l6 showed that the addition of corticosteroids to the preservation media reduced the release oflysosomal enzymes. Basu et all? reported that the addition of steroid to M-K increased the number of viable endothelial cells remaining at 4 days of 4°C storage and believed that this was because of lysosomal membrane stabilization by the corticosteroid. Additionally, Arya et al l8 ,l9 showed that a lysosomal extract from polymorphonuclear leukocytes was injurious to endothelial cell in vivo and in vitro. Hydrolytic lysosomal enzyme levels in corneal storage media may therefore provide an index of corneal autolysis which might relate to the performance of the graft at the time of transplantation. As an initial step to investigate this possibility, we measured lysosomal acid hydrolase levels, to see what changes occur with storage time and to compare M-K with K-Sol. Additionally, the interactions of M-K and K-Sol with this class of enzyme has been studied.
METHODS SAMPLING METHODS
Matched pairs of corneas (2-mm corneoscleral rims) were obtained from the Cleveland Eye Bank which had contraindications for transplantation (most often age or infection at a site other than the eye). Postmortem times were no greater than 9 hours (average, 6 hours) and all corneas appeared normal. The average age of the donor was 69 years. One cornea of each pair was placed in 20 ml ofK-Sol (Cilco, Bellevue, WA) or M-K (Aurora, Buffalo, NY) storage solution. One cornea was transferred to K-Sol after 1 day of storage at 4°C in M-K medium. All corneas were stored at 4°C during sampling, and matched pairs were always handled in the same manner. Two sampling methods were used: daily or every other day, 0.50 ml of media was removed under aseptic conditions or 1.0 ml was removed with replacement of fresh media. At the end of the sampling period all media batches were cultured on blood-augerose culture medium at 37°C for 72 hours. Samples were immediately frozen at -20°C until the analyses could be performed.
MIXING EXPERIMENTS
Fresh human corneal tissue from three donors were separately homogenized in 1.0 ml of 0.25 M sucrose, 0.05 M TRIS-HCI buffer at pH 7.4 and 0.1% Triton X-lOO using a Potter Elvehjem-type ground glass homogenizer (Fisher, Pittsburgh, PA). The homogenates were centrifuged at 1500 X g for 10 minutes and the supernatants saved and frozen for enzyme assay. The analysis of each was carried out in triplicate as described below. ENZYME ASSAY
Solutions of the following three 4-methyl-umbelliferyl glycosides (Sigma, St Louis, MO) were prepared in distilled water and used as substrates: a-galactoside, 0.5 mM20; (3glucuronide, 5 mM21; and (3-glucosaminide, 5 mM.22 In conducting the storage experiments, a known volume (determined for each enzyme studied to bring the values on scale) ofK-Sol or M-K medium was added to distilled water to reach a total volume of 100 ~l. In conducting the mixing experiments, 10 ~l of corneal supernatant was added to 90 ~l of either distilled water, M-K, or K-Sol to reach this 100-~1 value. To these were added 100 ~l of citrate-phosphate buffer of appropriate pH (pH used was based on the data of Bruner and Stark23 ) and 100 ~l of appropriate umbelliferyl-glycoside substrate so that a total volume of 0.30 ml in a 7.5 X 1-cm Pyrex test tube was achieved. The citrate-phosphate buffers were prepared with 0.10M citric acid and 0.20-M sodium phosphate, as described by Gomori.24 We performed pH measurements on the assay solutions to assure that dilution did not affect the pH of the reaction mixture. All measurements were performed in duplicate or triplicate, and under conditions of linearity with incubation time and substrate concentration. 22 Incubation was carried out for 30 to 45 minutes in a water bath at 37°C, and the reaction was stopped by adding 1.0 ml of0.40-M glycine-sodium hydroxide buffer, pH 10.3. This method is based on that described by Leaback and Walker. 22 Enzymatic activity was quantitated by determining the amount of umbelliferone released by the enzyme, as measured using an American Instrument Company (AMINCO) fluorometer Model J4-7439 with a primary filter of Corning 7-60 (Corning, NY) and a secondary filter (AMINCO J4-7116). Umbelliferone standard curves using crystalline 4-methyl-umbelliferone (Eastman, Rochester, NY) were performed with each experiment. CALCULATIONS AND STATISTICS
For each sample, other than the first sample, the activity which was removed had to be added back so that the values would approximate that which would have been found had the media not been sampled previously. The activity (nmole/hour) was calculated for the amount removed and added back to that which was found for the total sample then converted to activity per milliliter based on the total volume in the storage vessel for that data
1499
OPHTHALMOLOGY
•
NOVEMBER 1988
•
VOLUME 95
•
NUMBER 11
Mixing Experiment: Effect of Media on Activity Fig I. Effect of mixing M-K or K-Sol on the average activities of three lysosomal enzymes from normal human cornea. Both /3-g1ucuronidase and a-galactosidase were stimulated (P < 0.0 I) to an equal extent (P < 0.05), whereas the activity of /3-N-acetyl-g1ucosaminidase was inhibited (P < 0.0 I). This experiment shows that i3-g1ucuronidase may be used to compare release rates in M-K and K-Sol media.
4
Average Activity (Ilmol.al mg prolelnl hr .)
3 •
2
Corneal Activity
II Activity
with MK
m Activity
with
K·Sol
o Bela·Glucuron
NAGase
Alpha·Gal
ENZYME
NAGase_Beta · N· Acetyl·Glucosaminidase
Beta-Glucuronidase: MK vs K-Sol
3.5 ..-..- ...-.-..-....-... Fig 2. The average activity (nmoles/ml/hour) of 13glucuronidase released by matched pairs of human donor corneas stored at 4°C in either M-K or K-Sol as a function of storage time is shown. Data points represent average values from at least three paired corneas with some data points representing data from as many as seven paired corneas. Lines were generated by linear regression.
3.0
2.5
'1.-...-_. . -..-... .-'...-..-.. -..-..-.... .-. ..r...-...-....-....-...-....-.,..---~--c-.---. .........................:........................ .
~
- --
! - i-- t - - - - i --.:...ti-----""----I
2 . 0~ --------+-------4--------t-----~,r-------~---~--- 1
r;MKI
~
0.5
0.0
+-......---i--....--+--..--t----..,....---t---.---+---.---t 15 30 10 20 25 o 5 Time(Oays)
point. Statistical analysis was performed using a two-tailed Student's t test based on paired differences.
RESULTS MIXING EXPERIMENTS
Figure 1 shows the effect of M-K and K-Sol on the average activities of three lysosomal enzymes from three 1500
normal human corneas. McCarey-Kaufman and K-Sol stimulated (P < 0.01) both ,8-glucuronidase and a-galactosidase about equally (P > 0.05). ,8-N-acetyl-glucosaminidase, however, was stimulated to some extent by MK media (P > 0.05) but was inhibited by K-Sol by over 90% (P < 0.01). Since there was equal stimulation of,8glucuronidase by both M-K and K-Sol, this enzyme was used to compare release by the two media.
MORGAN et al
•
ENZYME LEVELS: K-SOL VERSUS M-K
Beta-Glucuronidase: Activity In MK
5.0
0
4.5
I
4.0
Activity (nmolesl ml/hr)
A /·4
I
3.5 3.0
1
2.5
/'
A"1i.~7;6 ...-
~..__Ii;
.-
.-I ~!-
.1. ....---•••
2.0 1.5
A~---· ~4"1'
.~
y"
1.0
o
./1
d____ O
1
0-
'0-
1
0.0
o
JI ~." ..-
-~/
-.~
.. ~~~~t0.0-0- -0.0..-0::;0 0'0-0.0'0
,,.a.p ~.... O~
0.5
)'0
""'".
Fig 3. The release of ,B-glucuronidase by individual corneas stored at 4°C in M-K media as a function of storage time is shown. Corneas differ in their release rates, with some corneas showing a more rapid rise in activity than other corneas.
10
5
20
15
25
30
Days
B.ta-Glucuronid... : Activity In K-Sol 5.0 4.5 4.0 3.5 Activity
3.0
(nmole.' mllhr)
2.5
_.I.
2.0 1.5 1.0
~.l.J
I
. ;.•. ~
i
r
-r-.-. ,;.J
....
~
•I
• ..-.J----O·O-O-O-O· 9"-.~·_.·.··--~"'[j:d:0-1-0= 0=0=0 i
0.5
00
va.-
J
I ____........
.e::..I.-.l.--.1.-.1.-- _:I: . , .1..../ ~•.I..a.- I _.,....~or-. ~Ii.
• ~.,.~ ~._.-.-.-.,.IJ. ..-~ ~ 0 l"b ~•.• -•.•
.I.'
Fig 4. The release of ,8-glucuronidase from individual corneas stored at 4°C in KSol as a function of storage time is shown. Individual variation in the release rate is shown as was found with MK storage but with less overall activity released.
~o_o
o
5
10
LYSOSOMAL ENZYME ACTIVITY IN CORNEAL STORAGE MEDIA
Figure 2 shows the average activity (nmoles/ml/hour) of ,8-glucuronidase released by matched pairs of human donor corneas in the two storage media measured as a function of storage time. Data points represent average values from at least three paired corneas with some data points representing data from as many as seven paired corneas. Lines were obtained by linear regression. It is noteworthy that the activities become disparate by 3 days of storage. At 4 days, the values approached significance (P
20
25
30
< 0.06) and by day 10 significantly higher levels of ,8glucuronidase were found in the M-K media (P < 0.01». This level of significance continued up to 30 days of storage when sampling was terminated. Figures 3 and 4 show the release of ,8-glucuronidase by individual corneas for M-K and K-Sol, respectively. These figures demonstrate that individual corneas release activity at different rates. No correlation to age of donor, postmortem time, or enucleation to preservation time could be demonstrated. Some corneas stored in M-K released enzyme at rates comparable with K-Sol, whereas some released considerably more enzyme. 1501
OPHTHALMOLOGY
•
NOVEMBER 1988
CULTURE RESULTS
At 72 hours, all of the cultures showed no growth indicating that contamination by microorganisms did not contribute to enzyme activities.
DISCUSSION Basu et a1 14 •15 have shown that lysosomal enzymes are important to the endothelial viability of stored corneas. Arya et al 18.19 have shown lysosomal enzymes to be injurious to endothelial cells in vivo and in vitro. Spencer et al l6 showed that an inhibition of lysosomal enzyme release occurred when steroids were added to storage media, and Basu et al 17 found that steroids increased endothelial viability, suggesting that the inhibition of enzyme release accounted for this finding. We investigated the interactions of M-K and K-Sol with lysosomal enzymes and their release from stored corneas. Additionally, these studies represent preliminary experiments on the possibility for the use of lysosomal enzyme levels in corneal storage media as a screen for excessive autolysis as a predictor of excessive endothelial cell loss. In order to determine if enzyme levels could be used to legitimately compare M-K and K-Sol storage, we first studied the interaction of these media with activities of ground human corneal supernatants. Both M-K and KSol stimulated (P < 0.01) the activities of ,B-glucuronidase and a-galactosidase to an equal extent. This could be due to the presence of cations or protein in the preservation media. McCarey-Kaufman stimulated the activity of {3N-acetyl-glucosaminidase to some extent (P > 0.05), whereas K-Sol was found to inhibit this enzyme (P < 0.01). {3-N-Acetyl-glucosaminidase is an important lysosomal enzyme in the metabolism of proteoglycans since five of the seven species ofproteoglycans (i.e., chondroitin4-sulfate, chondroitin-6-sulfate, dermatin sulfate, hyaluronic acid, and keratin sulfate I and II), contain {3-hexosaminidic linkages that are degraded by this enzyme. 25 Keratin sulfate and chondroitin sulfate would be most important since they are the major proteoglycans of the cornea. 26 The chondroitin sulfate in K-Sol probably inhibits the activity of this enzyme by acting as a substrate inhibitor since it has been shown to degrade this proteoglycan.27 K-Sol might inhibit corneal autolysis by reducing the catabolism of corneal proteoglycans by this enzyme. Since {3-glucuronidase was stimulated to the same extent by both media, this enzyme was used to compare lysosomal enzyme release into the two media. Lysosomal enzyme release into M-K and K-Sol media during 4°C storage was investigated. As shown in Figure 2, we found that the rate of release in both media to be linear. Corneas stored in K-Sol released less lysosomal enzymes, with the difference approaching significance by day 4 (P < 0.06) and by day 10 and beyond significantly more activity was found in the M-K media (P < 0.01). It is interesting to note that average {3-glucuronidase activities in M-K media at day 3 (0.75 nmoles/ml/hour) 1502
•
VOLUME 95
•
NUMBER 11
were not equal in K-Sol until day 10 (0.78 nmoles/ml/ hour). This finding seems to correspond well with the observations of Bourne l3 that high endothelial cell loss was seen in corneas stored more than 2 days in M-K and 10 days in K-Sol. Two sampling methods were used since each method approximates activity values which would have been obtained for each day had the media not been sampled previously. With one method, we removed samples without adding back fresh media. This method might be expected to underestimate the true value since during the sampling period the total volume was decreased and therefore the concentration of enzyme would be increased, possibly leading to reduced release because of a smaller concentration gradient. The other method in which fresh media was added back would cause a dilution of activity and therefore an increased concentration gradient resulting in increased release. By using both methods and combining the data, we hoped to more accurately approximate the values which one might obtain for a particular day of storage. Data from the two methods were not significantly different. Since we are interested in the possible use of these enzyme values as a screen for unacceptable corneal autolysis, it was important to see if individual corneas differed in their release rates. Figures 3 and 4 show the enzyme release curves for five individual corneas stored in M-K and KSol, respectively. We can see that corneas vary in their release rates with some corneas showing slow release during the early storage period, whereas others show a more rapid rise in activity. No correlation to age of donor, postmortem time, or enucleation to preservation time could be demonstrated. Whether corneas that show more release would have also showed higher endothelial cell losses after penetrating keratoplasty cannot be commented on with the current data. This possible use of lysosomal enzyme analysis deserves further investigation. In this study, we did not show whether the enzymes were coming from the endothelium or epithelium. In a separate study of 12 cat corneas stored in K-Sol at 4°C for 14 days with and without their epithelium present (data not shown), significantly higher enzyme levels were observed from corneas with their epithelium intact when compared with those without epithelium, suggesting that in fact most of the activity is coming from epithelial cells. This does not, however, negate the usefulness of enzyme analysis because the literature suggests that these enzymes are potentially damaging to the endothelium regardless of their origin. One might also reasonably assume that enzyme release from epithelium would be paralleled by the same release from endothelium. In 20 ml of storage solution, one might question whether the enzyme concentrations observed could become high enough to show a toxic effect on the endothelium. Diffusion of these enzymes through the stroma could result in higher concentrations of enzyme affecting the endothelial cells at their attachment sites to the cornea in areas where certain acidic microenvironments might exist.
MORGAN et al
•
ENZYME LEVELS: K-SOL VERSUS M-K
Even before release of lysosomal enzymes into the media would occur one could envision intracellular release and damage to the individual cell concerned. However, in the current study, we have not demonstrated whether such a mechanism exists or whether the release of these enzymes indicates cellular damage or merely decreased metabolism at 4°C. Because these media are kept at 4°C and pH 7.4, one might not expect these acid hydrolases to be highly active even if certain acidic microenvironments existed. They may, however, serve as markers for the release of other lysosomal enzymes such as cathepsins which possess neutral pH optima and therefore may be active under these storage conditions. In summary, we have shown that K-Sol medium inhibits the release of lysosomal enzymes, which may be damaging to the corneal endothelium during corneal perservation, to a greater extent than does MK medium. KSol also inhibits the activity of ~-N-acetyl-glucosamini dase, an important enzyme in the catabolism of corneal proteoglycans. Corneas differ in their release rates of lysosomal enzyme in both K-Sol and M-K. Whether these enzyme levels will predict endothelial cell performance after penetrating keratoplasty remains to be seen.
REFERENCES 1. Binder PS. Eye banking and corneal preservation. In: Symposium on Medical and Surgical Diseases of the Comea: New Orleans Academy of Ophthalmology. St Louis: CV Mosby Co, 1980; 320-54. 2. McCarey BE, Sakimoto T, Bigar F. Ultrastructure of M-K and refrigerated moist chamber-stored corneas. Invest Ophthalmol Vis Sci 1974; 13: 859-63. 3. Van Horn DL, Shultz RO, DeBruin J. Endothelial survival in corneal tissue stored in M-K medium. Am J Ophthalmol1975; 80:642-7. 4. Aquavella JV, Van Horn DL, Haggerty, CJ. Corneal preservation using M-K medium. Am J Ophthalmol1975; 80:791-9. 5. Kuwahara Y, Sakanoue M, Hayashi M. et al. Studies on the long-term preservation of the cornea for penetrating keratoplasty. Nippon Ganka Gakkaizasshi 1965; 69:1751-840. 6. Nakano T. Studies on the viability of the cornea stored for the penetrating keratoplasty under various experimental conditions. Report III. Oxygen consumption of rabbit corneas stored in newly devised preservation medium containing various macromolecular colloids. Nippon Ganka Gakkaizasshi 1966; 70:662-74. 7. Mizukawa T. Corneal transplantation. J Jpn Med Assoc 1967; 58: 957-64. 8. Mizukawa T, Manabe R. Recent advances in keratoplasty with special
9.
10. 11.
12.
13.
14. 15.
16. 17.
18.
19. 20.
21.
22.
23. 24. 25.
26. 27.
reference to the advantages of liquid preservation. Folia Ophthalmol Jpn 1968; 19:1310-8. Kaufman HE, Varnell ED, Kaufman S. Chondroitin sulfate in a new cornea preservation medium [Letter]. Am J Ophthalmol 1984; 98: 112-4. Kaufman HE, Varnell ED, Kaufman S. et al. K-Sol corneal preservation. Am J Ophthalmol1985; 100:299-304. Stein RM, Bourne WM, Campbell, RJ. Chondroitin sulfate for corneal preservation at 4°C: evaluation by electron microscopy. Arch Ophthalmol1986; 104:1358-61. Stein RM, Laibson PRo Comparison of chondroitin sulfate to McCareyKaufman medium for corneal storage. Am J Ophthalmol 1987; 104: 490-3. Bourne WM. Endothelial cell survival on transplanted human corneas preserved at 4 C in 2.5% chondroitin sulfate for one to 13 days. Am J Ophthalmol 1986; 102:382-6. Basu PK, Hasany S. Autolysis of the cornea of stored human donor eyes. Can J Ophthalmol 1974; 9:229-35. Basu PK, Hasany SM, Ranadive, NS, Chapman M. Damage to corneal endothelial cells by lysosomal enzymes in stored human eyes. Can J Ophthalmol1980; 15:137-40. Spencer JA, Dixon WS, Ranadive NS, Basu PK. Factors in the survival of stored corneas. Can J OphthalmoI1977; 12:123-7. Basu PK, Hasany SM, Doane FW, Schultes K. Can steroid reduce endothelial damage in stored corneas? Effect on cell viability and ultrastructure. Can J Ophthalmol1978; 13:31-8. Arya DV, Mannagh J, Irvine AR Jr., Yuhasz Z. Effect of Iysosomes on corneal endothelium: an in vivo study. Invest Ophthalmol 1972; 11:655-61. Arya DV, Mannagh J, Irvine AR Jr. Effects of Iysosomes on cultured rabbit corneal endothelial cells. Invest Ophthalmol1972; 11 :662-7. Desnick RJ, Allen KY, Desnick SJ, et al. Fabry's disease: enzymatic diagnosis of hemizygotes and heterozygotes. J Lab Clin Med 1973; 81:157-71. Mead JAR, Smith IN, Williams RT. The biosynthesis of the glucuronides of umbelliferone and 4-methylumbelliferone and their use in fluorimetric determination of J3-glucuronidase. Biochem J 1955; 61 :569-74. Leaback DH, Walker PG. Studies on glucosaminidase 4. The fluorimetric assay of N-acetyl-J3-glucosaminidase. Biochem J 1961; 78: 151-6. Bruner WE, Stark WJ. Glycosidases of normal human donor cornea. Curr Eye Res 1982; 1:705-10. Gomori, G. Preparation of buffers for use in enzyme studies. Methods Enzymol 1955; 1: 138-46. Roden, L. Structure and metabolism of connective tissue proteoglycans. In: Lennarz WJ, ed. The Biochemistry of Glycoproteins and Proteoglycans. New York: Plenum Press, 1980; 285-6. Maurice DM, Riley MV. The cornea. In: Graymore CN, ed. Biochemistry of the Eye. New York: Academic Press, 1970; 30. Mahadevan S, Dillard CT, Tappel AL. Degradation of polysaccharides, mucopolysaccharides, and glycoproteins by lysosomal glycosidases. Arch Biochem Biophys 1969; 129:525-33.
1503