Increased endothelial survival of organ-cultured corneas stored in FGF-2-supplemented serum-free medium

(1)

Article

Reference

Increased endothelial survival of organ-cultured corneas stored in FGF-2-supplemented serum-free medium

RIECK, Peter W, et al.

Abstract

To investigate the effect of FGF-2 on corneal endothelial cell survival in porcine and human corneas during corneal storage in a serum-free medium.

RIECK, Peter W, et al . Increased endothelial survival of organ-cultured corneas stored in

FGF-2-supplemented serum-free medium. Investigative Ophthalmology & Visual Science , 2003, vol. 44, no. 9, p. 3826-32

PMID : 12939298

Available at:

http://archive-ouverte.unige.ch/unige:93217

Disclaimer: layout of this document may differ from the published version.

(2)

Corneas Stored in FGF-2–Supplemented Serum-Free Medium

Peter W. Rieck,

¹

Marcel Gigon,

²

Jan Jaroszewski,

¹

Uwe Pleyer,

¹

and Christian Hartmann

¹

PURPOSE.To investigate the effect of FGF-2 on corneal endothelial cell survival in porcine and human corneas during corneal storage in a serum-free medium.

M^ETHODS.Porcine and paired human corneas were stored at 32°C for 9 and 22 days, respectively. One cornea of each pair was stored in a serum-free culture medium, and the mate was preserved in the same medium supplemented with 10 ng/mL FGF-2. Quantitative analysis of corneal damage after storage was determined by the Janus green photometry technique.

5-Bromo-2-deoxyuridine (BrdU) labeling of the endothelium determined the effect of FGF-2 on endothelial proliferation during storage. Additional cell culture studies were performed to elucidate the role of FGF-2 on the incidence of endothelial apoptosis after serum deprivation.

RESULTS.When FGF-2 was added to the serum-free medium, the damage rates of porcine endothelia were reduced from 15.1%

⫾8.7% (control) to 6.4%⫾2.0% after 9 days and from 25.3%

⫾ 10.2% to 13.6% ⫾ 4.2% after 22 days of storage. In the human corneas stored during 22 days in FGF-2–supplemented medium, the amount of endothelial damage was 11.8%⫾3.2%, which was significantly less damage than in the control fellow corneas stored in unsupplemented serum-free medium (19.3%

⫾6.3%;P⬍0.01). DNA synthesis was not enhanced in corneas stored in serum-free medium, serum-free medium⫹FGF-2, or medium containing 10% FCS. Only a few (3.8%) TUNEL-positive endothelial cells were detected in cultures maintained in FGF-2–supplemented serum-free medium compared with a high number (48%) of apoptotic cells in control cultures.

C^ONCLUSIONS.FGF-2 efficiently reduces human corneal endothelial damage that occurs during organ culture storage in a serum- free medium. This effect is truly protective, because no proliferative activity and a decreased rate of apoptosis were determined. FGF-2 emerges as an important component of a future serum-free corneal organ-culture medium established to replace fetal calf serum (FCS) as a potential source of recipient infection. (Invest Ophthalmol Vis Sci. 2003;44:3826 –3832) DOI:10.1167/iovs.02-0601

O

rgan culture is the preferred storage method in Europe, mainly because of the benefits of an extended storage time that has advantages such as standardized quality and microbiologic testing, optimal matching of donor and recipient (e.g., age, HLA), convenient scheduling of the surgery, and constant availability of corneas for emergency procedures.^1,2 Excellent poststorage morphologic features³and postoperative clinical results⁴ have been achieved using this preservation procedure, with corneas stored for up to 48 days.⁵

The most important issue regarding donor cornea selection for transplantation refers to the cell number and vitality of the endothelium.⁴Although the in vitro viability of the endothelium stored in organ culture is longer than with other storage techniques, endothelial cell loss occurs with increasing length of storage time, probably due to metabolic changes in the culture medium (glucose depletion and lactate accumulation), nutrient deprivation, mechanical stress (Descemet’s folds as a consequence of stromal swelling), endotoxins, and loss of survival factors.^4,6,7The mean endothelial cell loss during organ culture is approximately 10% of the initial cell number.⁶

Efforts to reduce endothelial cell loss are directed toward agents that primarily protect the endothelium during the preservation period. Fetal (FCS) or newborn calf serum (NCS) as an additive for corneal organ culture media has a protective effect for endothelial cells⁸and its addition to standardized media has been a recommended procedure performed in most European cornea banks. However, due to variations in the quality of the different commercially available sera preparations, the viability of the preserved endothelium may sometimes be severely com- promised. In addition, FCS bears an at least theoretical risk as a potential infectious source, which has recently become evi- dent in view of the appearance of bovine spongiform enceph- alopathy (BSE) in cattle of several European countries during 2000. These detrimental features of calf serum emphasize the need for the development of a serum-free organ-culture medium.

FGFs are polypeptides that form a family of heparin-binding growth factors, actually consisting of 13 members. The two major forms, acidic and basic FGF (FGF-1 and -2, respectively), share a 55% absolute homology in their primary structure⁹and have been isolated from various tissues.¹⁰In vitro, both FGF-1 and -2 act on a wide variety of cells from neuroectodermal and mesodermal origin.¹⁰FGFs interact with specific cell surface receptors, and the genes encoding these receptors have been isolated.^11,12 FGF-1 and -2 have been shown to be mitogenic agents for cells and tissues and to stimulate wound repair of these cells, including those of the corneal endothelium, in vitro and in vivo.^13–19In addition, further studies have shown that the growth factor possesses the properties of a true survival factor in tissues, including ocular tissues, in that it promotes the survival of retinal ganglion and photoreceptor cells.^20,21

In the present study, we therefore investigated the ability of a recombinant-derived human FGF-2 to protect the corneal endothelium of porcine and human corneas during corneal storage in a serum-free organ culture medium.

From the¹Department of Ophthalmology, Charite´ Medical Fac- ulty, Humboldt University Berlin, Germany; and the²Department of Ophthalmology, Geneva University Hospital, Geneva, Switzerland.

Supported by Grant Ri 568/3-2 from the German Research Foun- dation, Bonn, Germany (PWR). MG is supported by the A. R. and J.

Leenard Foundation, Lausanne, Switzerland, and the Department of Ophthalmology, University Hospital of Geneva, Geneva, Switzerland.

Submitted for publication June 18, 2002; revised October 24, 2002; accepted January 31, 2003.

Disclosure:P.W. Rieck, None;M. Gigon, None;J. Jaroszewski, None;U. Pleyer, None;C. Hartmann, None

The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertise- ment” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Corresponding author: Peter W. Rieck, Department of Ophthal- mology, Charite´ Medical Faculty, Campus Virchow Hospital, Humbolt University Berlin, Augustenburger Platz 1, D-13353 Berlin, Germany;

peter.rieck@charite.de.

Investigative Ophthalmology & Visual Science, September 2003, Vol. 44, No. 9

(3)

M

ATERIAL AND

M

ETHODS

Materials

Minimum essential medium (MEM), Dulbecco’s modified Eagle’s medium (DMEM),L-glutamine, gentamicin, trypsin-EDTA (0.02%), and FCS were purchased from GibcoBRL (Paisley, Scotland, UK). Penicillin- streptomycin, human recombinant FGF-2, and trypan blue were pro- cured from Sigma-Aldrich (St. Louis, MO). HEPES was from Serva (Heidelberg, Germany). Culture dishes were from Falcon-BD Bio- sciences (Bedford, MA). The 5-bromo-2⬘-deoxyuridine (BrdU) labeling and detection kit and the in situ apoptosis detection kit were purchased from Roche Molecular Biochemicals (Mannheim, Germany).

Organ Culture

Porcine Eyes.Fresh eyes of 6-month-old pigs were obtained from a local abattoir. Enucleation was performed immediately after death. The eyes were transported in a saline solution (BSS; Alcon, Freiburg, Germany) at 4°C with 40 mg/L nystatin and 20 mg/L gentamicin sulfate and processed within 60 to 120 minutes after enucleation. In total, 96 porcine corneas were used for these experiments.

Human Corneas.Human donor corneas that were discarded by our cornea bank because of corneal scars, low endothelial cell density (ECD⬍2200 cells/mm²), or medical contraindications to transplantation were used in the study. All corneas were obtained in accordance with the guidelines set forth in the Declaration of Helsinki. Paired corneas only were used, to overcome interindividual variation of mor- phometric aspects and physiological responses of the endothelium.

Corneas with extreme polymegethetic, pleomorphic, or diseased endothelium were excluded. Fifteen pairs of human corneas were used.

Donor ages ranged from 46 to 79 years (mean, 63⫾19 [SD] years). The time interval between death and corneal dissection averaged 15.6⫾ 6.2 hours. The corneas had an average ECD of 1845⫾421 cells/mm² (range, 1320 –2150).

Processing of the Eyes.Both porcine and human globes were decontaminated in a solution of polyvinylpyrolidoniodine (25 g/L) for 5 minutes. Subsequently, the eyes were transferred in a solution of sodium thiosulfate (9.328 g/L) for 3 minutes and finally were rinsed in a saline solution (BSS; Alcon). All corneas were then carefully excised with a 2-mm scleral rim and placed on a wax cup with the endothelium side up. The whole endothelium was stained for 60 seconds with trypan blue (0.4%). The cornea was then rinsed thoroughly in BSS to expose stained cells and thus to detect endothelial lesions. In any case of severe endothelial damage, the cornea was excluded. A 7-0 suture was placed in the scleral rim, and the corneoscleral segment was suspended in 100 mL storage medium in a glass bottle closed with a rubber stopper. All manipulations were performed under sterile conditions under a laminar-flow hood.

Corneal Storage Medium. Each corneoscleral segment of porcine and human corneas was stored in 100 mL serum-free MEM containingL-glutamine (2 mM) and HEPES buffer (25 mM), NaHCO3

(20 g/L), penicillin (100 IE/mL), streptomycin (0.1 mg/mL), and nystatin (50 E/mL) or in 100 mL complete serum-free medium to which 10 ng/mL FGF-2 was added.

Recombinant Human (rh)FGF-2.Rh-FGF-2 was provided as a lyophilized powder containing 50 ␮g. The reconstitution was achieved with 5 mL saline solution (BSS; Alcon) to obtain an FGF-2 concentration of 10 ␮g/mL. The in vitro biological activity of the solution was assessed by measuring tritiated thymidine incorporation into cellular DNA of cultured bovine epithelial lens cells.²²The activity was confirmed to be in the range of that reported in the literature and of our laboratory experience (EC₅₀⫽50 –200 pg/mL).

Experimental Groups.To determine the endothelial damage with increasing storage time, porcine corneas were stored for 0 (freshly excised), 4, 9, 14, 21, and 28 days (n⫽8 corneas at each time point) under serum-free conditions at 32°C. The comparative study was performed with porcine eyes stored for either 9 or 21 days in serum-

free or FGF-2–supplemented medium, respectively (n⫽8 corneas for each experimental group). The corneas were randomly assigned to one of the following three experimental groups: fresh corneas, corneas preserved in serum-free medium, and corneas preserved in serum-free medium plus FGF-2.

One cornea of each pair of human eyes was randomly assigned to a 21-day preservation in serum-free medium, and the fellow eye was stored for the same time in serum-free medium supplemented with FGF-2. The media were changed and the FGF-2 supplemented every 7 days. After this storage period, to exactly reproduce the procedure in a cornea bank, each cornea was transferred for an additional 24 hours to a serum-free medium containing 5% dextran T500 which is routinely used before transplantation to reverse stromal swelling that occurs during organ culture. The following examinations were performed in a blind fashion, in that the observer was masked to the assignment of each vial.

Janus Green Photometry Technique

This dye exclusion assay quantitates endothelial cellular membrane integrity as an indication of cell viability. The assay measures the amount of Janus green dye extracted from corneas after vital staining.

The photometric determination of the extracted dye corresponds to the percentage of altered endothelial cells and has been established as a reliable parameter of endothelial damage. The technique has been described in detail.²³Briefly, after a brief rinse of freshly excised or preserved corneas in a BSS, the corneoscleral segment was placed on a corneal trephining wax cup with the endothelial side up. The whole endothelium was stained for 90 seconds with a 1% Janus green solution (Merck, Darmstadt, Germany). After thorough rinsing, the central corneal area was punched out with an 8.0-mm diameter hand trephine.

The stain from each trephined button was eluted in a disposable test tube that was filled with 1 mL of absolute ethanol. Complete extraction of the stain was achieved after 90 seconds. Photometric measurement of the eluate was performed with a spectrophotometer (Anthos Labtec Instruments, Salzburg, Austria) at 650 nm, taking absolute ethyl alcohol as the blank value. The extinction values corresponding to “normal”

endothelial damage in fresh corneas and complete (100%) damage were determined by taking (1) freshly excised corneas for the 0%

lesion and (2) for the 100% lesion, freshly excised corneas treated with absolute alcohol, to alter endothelial viability of the whole endothelial layer. A linear relation was established so that with a given extinction value, the damaged area in the percentage of the whole trephined endothelial surface could be determined.

Scanning Electron Microscopy

Corneas scheduled for examination by scanning electron microscopy (n⫽3 pairs) were fixed in 2.5% glutaraldehyde in 0.1 M cacodylate buffer for 24 hours and then stored at 4°C in 0.2 M cacodylate buffer.

The corneas were then dehydrated through a graded ethanol series (30%–100%) terminating in anhydrous acetone. Subsequently, the corneas were critical point dried from carbon dioxide, sputter coated with gold-palladium alloy, and examined using a scanning electron microscope (JSM T200; JEOL, Tokyo, Japan).

BrdU Assay

BrdU labeling medium was added to the endothelial side of corneas that had been organ cultured for 22 days in serum-containing medium, serum-free medium, or serum-free medium supplemented with 10 ng/mL FGF-2. As a positive control, a mechanical scratch injury was performed on a cornea before preservation in FCS-containing medium.

The corneas were incubated with the labeling medium at 37°C in 5%

CO2for 60 minutes and then washed three times with washing buffer (PBS). Subsequently, they were fixed with 70% ethanol in glycerin buffer (pH 2.0). After corneas were washed with PBS, they were incubated with an anti-BrdU solution for 30 minutes at 37°C. Anti- mouse-Ig-alkaline phosphatase was applied to the endothelium and incubated for 30 minutes at 37°C. The corneas were washed with PBS IOVS,September 2003, Vol. 44, No. 9 FGF-2 and Corneal Endothelial Cell Survival 3827

(4)

again, and the cell monolayer was covered with a colored substrate solution (nitroblue tetrazolium [NBT], X-phosphate). After an incuba- tion time of 30 minutes at room temperature, photomicrographs of stained cells were obtained with a light microscope (Leica, Heidelberg, Germany).

Cell Culture

Bovine corneal endothelial cell (BCEC) cultures were established according to our standardized protocol.²⁴ Briefly, the harvested cells were centrifuged and the supernatant was taken up in 1 mL DMEM.

The number of cells in the suspension was determined by counting with a cell counter (CASY 1; Scha¨rfe System GmbH, Reutlingen, Ger- many). After centrifugation, cells were resuspended in culture medium with 10% FCS and seeded at a density of 10⁴cells per well in six-well dishes containing 2 mL culture medium supplemented with 10% FCS.

After 48 hours, the cells were washed three times with serum-free DMEM. Subsequently, BCECs were cultured for three more days in either culture medium containing 10% FCS, in serum-free medium, or in serum- free medium supplemented with 10 ng/mL FGF-2, respectively.

TUNEL Assay

Primary BCECs cultured in the three different media were fixed and mounted on slides coated with poly-L-lysine. Subsequently, the cells were permeabilized with 0.1% saponin in PBS for 40 minutes, washed four times for 5 minutes each in PBS, treated with acetone for 1 minute, washed four times for 3 minutes each in PBS, digested with 1 mg/mL proteinase K for 10 minutes at room temperature, and washed again four times for 3 minutes each in PBS. The protocol supplied with the in situ apoptosis detection kit (Roche Molecular Biochemicals) was adhered to for the remainder of the procedure. Endogenous peroxidase activity was quenched with 2% hydrogen peroxidase, and specimens were equilibrated in 1⫻labeling buffer and labeled with biotin- ylated nucleotide triphosphates using terminal deoxynucleotidyl transferase (TdT) for 1 hour at 37°C. Reaction was stopped with 1⫻ stop buffer for 5 minutes. The specimens were then rinsed with water, streptavidin– horseradish peroxidase conjugate was applied for 10 minutes, and the cells were washed four times in water for 3 minutes each time. Peroxidase substrate was applied for 3 minutes, and specimens were rinsed again with four changes of water. For a negative control, the TdT enzyme was omitted from labeling reaction in some specimens and processed together with experimental assays.

Quantitative Determination of Apoptotic Cells

To improve visualization, the cells were stained with a 5% Giemsa solution (Hollborn & So¨hne, Leipzig, Germany). Apoptotic cells were quantified by counting TUNEL-positive cells per 200⫻field under a light microscope (Leica). This number was divided by the total number of cells in the same field, yielding the percentage of apoptotic cells within each specified field.

Statistics

Statistical analyses of the data were performed on computer (Sigma Stat; SPSS Sciences, Chicago, IL).

Organ Culture Studies.The difference between the endothelial damage of corneas stored in serum-free medium and corneas preserved in rhFGF-2–supplemented medium was analyzed with the Mann-Whitney test (porcine corneas) or the Wilcoxon signed rank test (human corneas).

Cell Culture Studies.The unpaired Student’st-test was applied to determine the statistical significance between the number of apoptotic endothelial cells of corneas preserved in medium containing 10% FCS, serum-free medium, and serum-free medium supplemented with FGF-2. To determine the statistical significance between more than two groups, ANOVA with a Bonferroni type␣ correction for unpaired groups (Duncan multiple range test) was used. For all studies, P⬍0.05 was considered significant.

R

^ESULTS

Quantitative Analysis of Porcine Endothelial Damage

The porcine corneas were chosen to determine the time course of endothelial damage occurring during storage in a serum-free culture medium. The extinction values of the 10 freshly excised normal porcine corneas revealed an arithmetic mean of 0.04⫾0.02 corresponding to 4.8%⫾1.8% damaged cells. Apparently, small areas of endothelium may have been altered by the enucleation and corneal excision procedures.

Figure 1 shows the increase in endothelial cell damage with prolonged storage times. The amount of endothelial damage remained relatively low during the first week in organ culture.

From day 10 on, a significantly increased damage rate was observed. The curve steepened again after 21 days of storage.

At this time point, the damaged cells represented 21% of the entire area examined. After 4 weeks of preservation, approximately 50% of the cells were altered or had sloughed off.

The mean extinction value for the 10 control corneas after 9 days of storage was 0.14⫾0.08, corresponding to 15.1%⫾ 8.7% damaged cells. When 10 ng/mL FGF-2 was added, the extinction value at day 9 for the 10 stored corneas was 0.06⫾ 0.01, corresponding to 6.4%⫾2.0% damaged cells (Fig. 2). The lower amount of damaged cells of the corneas preserved in FGF-2–supplemented medium was significantly different (P⫽ 0.004) from the respective control group and was not significantly different from fresh corneas (P⫽0.17; Fig. 2).

When the storage time was extended to 22 days, the endothelia of the porcine corneas preserved in a serum-free medium revealed mean overall endothelial cell damage of 25.3% ⫾ 10.2%. The addition of FGF-2 to the medium reduced the damage rate to 13.6%⫾4.2% (Fig. 2).

Effect of FGF-2 on Endothelial Preservation of Human Corneas

For the human comparative study, we chose a fixed storage time of 21 days with an additional 24 hours in serum-free medium plus 5% dextran, according to standard eye bank protocols for deswelling of corneas before keratoplasty. The FIGURE1. Time course of increasing endothelial alterations with longer storage times. Porcine corneas were stored in a serum-free organ culture medium for 0 (fresh corneas), 4, 7, 10, 14, 21, 24, and 28 days (n⫽8 corneas/time point). The given percentage represents the endothelial cell damage measured with the Janus green photometry technique. Data are expressed as the mean⫾SD.

(5)

mean extinction value for the 10 control corneas (stored in serum-free MEM) after 21⫹1 days’ storage time was 0.20⫾ 0.06, corresponding to 19.3%⫾6.3% damaged cells. The mean extinction value for the 10 fellow eye human corneas stored in FGF-2–supplemented serum-free medium corresponded to 11.8%⫾3.2% damaged endothelial cells. The statistical analysis of these data revealed a significant difference between both groups (P⫽0.006).

Scanning Electron Microscopy

The ultrastructural appearance of the endothelium after storage in both groups is shown in Figure 3. The cells of control corneas showed a swollen cytoplasm, breakage of the cellular borders with loss of the typical ruffled aspect of the apical edges, and a decreased number of microvilli (Fig. 3A). Necrotic cells were also clearly visible. In contrast, the corneas stored in FGF-2–supplemented medium exhibited an almost physiological endothelium with slightly swollen nuclei and expression of microvilli on the apical cellular membrane as an indication of metabolic activity (Fig. 3B). No apparent morphologic alterations were detected in the endothelium of FGF-2–treated corneas.

BrdU Assay

The immunocytochemical assay for the detection of BrdU incorporation into cellular DNA was performed to evaluate a possible proliferative effect of FGF-2 during organ culture.

After a storage time of 22 days, no significant BrdU labeling of the cells was detected in both experimental groups, independent of the presence or absence of FGF-2 or FCS in the culture medium (Fig. 4A). As a positive control, we stored one donor cornea in a medium containing 10% FCS after a scratch injury was made to the central endothelium. BrdU labeling was found only in cells surrounding the remaining wound (i.e., denuded Descemet’s membrane; Fig. 4B). These experiments demonstrate that the reduced alteration of the endothelium of corneas stored in FGF-2–supplemented medium is not due to cellular

renewal during the storage time but represents a true protective effect on the cells.

Quantitative Determination of Apoptosis

The TUNEL assay was performed to detect apoptosis occurring during culture of BCECs in a serum-free medium, in serum-free medium supplemented with 10 ng/mL FGF-2, or in the basal medium supplemented with 10% FCS. Counting of the number of apoptotic cells after 3 days in culture showed 48%⫾8.5% of the total number of cells undergoing apoptosis in the serum- free medium (Fig. 5A). Only 3.8%⫾ 0.5% of the cells were apoptotic in the serum-free medium supplemented with FGF-2 (Fig. 5B) and no TUNEL-labeling was detected in the medium containing 10% FCS (Fig. 5C). The difference in the number of apoptotic cells was significant for both FGF-2- and FCS-containing medium compared with the cultures kept under serum-free conditions (P⬍0.001). There were no significant differences in the amount of BCEC apoptosis between cell cultures grown in FGF-2– or FCS-supplemented medium.

D

^ISCUSSION

The addition of fetal or newborn calf serum to corneal organ- culture medium is a routine procedure performed in most European cornea banks. However, mainly because of quality variations of the different sera used—sometimes even between different batches of the same serum—the composition and thus the quality of these media are not defined at the moment,

FIGURE3. Scanning electron photomicrographs of human endothelial layers after 22 days of storage in either serum-free control medium (A) or in medium supplemented with 10 ng/mL FGF-2 (B). (A, arrow) Disrupted cellular membrane. Scale bars, 25␮m.

FIGURE2. Effect of FGF-2 on endothelial preservation. Porcine corneas were stored for 9 or 22 days in serum-free medium or in serum supplemented with 10 ng/mL FGF-2. The rate of endothelial damage at both time points was assessed by the Janus green photometry technique. Data are expressed as the mean⫾SD.

IOVS,September 2003, Vol. 44, No. 9 FGF-2 and Corneal Endothelial Cell Survival 3829

(6)

which may lead to severe variations in the quality of corneal preservation, mainly concerning the viability of the endothelium. In addition, the need for the development of a serum-free organ-culture medium has recently gained further importance in view of the appearance of BSE in cattle in several European countries, which has raised questions of potential risks of contagion with the use of FCS.

The results of this study demonstrate a beneficial effect of exogenous human recombinant FGF-2 to protect both porcine and human corneal endothelium from damage after storage in a serum-free organ culture medium. Fewer damaged endothelial cells (12% of the total population) were detected after a storage time of 22 days in an FGF-2–supplemented medium compared with the fellow-eye human corneas stored in serum- free medium without supplements (19% damage).

Previous studies have provided evidence of the role of FGF-2 as a survival factor in ocular tissues. The growth factor has been found to promote survival of cholinergic neurons and retinal ganglion cells^20,25and to delay the phenotypic degeneration of rat retinal photoreceptors.^21,26 In addition, exogenous FGF-2 has been shown to protect retinal ganglion cells and other inner retinal elements from ischemic injury.²⁷ Renaud et al.²⁸have shown that if the intracellular content of FGF-1 is increased by transfection of the FGF-1 gene into PC12 cells, the survival of these cells in serum-free cultures is dra- matically enhanced compared with nontransfected control cultures.

Earlier cell culture studies have provided indirect evidence that FGF-2 is a survival factor for corneal endothelial cells, as well. Gospodarowicz et al.¹³ have observed that survival of corneal endothelial cells maintained in cell culture is enhanced either by addition of serum to the culture medium or by seeding a sufficient number of cells per well. Sato and Rifkin²⁹ explained the latter observation by speculating from their studies with vascular endothelial cells that the synthesis of sufficient endogenous FGF-2 is able to mimic the effect of serum to prolong cell survival. Indeed, our own investigations

FIGURE5. TUNEL assay for detection of apoptosis in serum-free, serum-containing, and FGF-2–supplemented serum-free cultures of BCECs. Cells cultured for 3 days in a serum-free medium (A). BCECs cultured in this medium supplemented with 10 ng/mL FGF-2 (B). Cells cultured in the basal medium with 10% FCS (C). (A, arrowheads) representative TUNEL-positive cells. Scale bars, 50␮m.

FIGURE4. BrdU assay. Corneas stored for 22 days in serum-free medium supplemented with FGF-2 (A), or in 10% FCS medium with a mechanical injury to the endothelium performed before preservation (B). The corneas were prepared for the BrdU assay. BrdU labeling (arrowheads) was found only in cells surrounding the area of the wounded endothelium (w). Scale bars, 100␮m.

(7)

have demonstrated that FGF-2 protein and the associated FGFR-1 receptor are produced by human endothelial cells in vitro.³⁰ Exogenous FGF-2 has been used in previous studies that have demonstrated that the growth factor is able to accel- erate healing of mechanically denuded areas of human endothelial cells in organ culture^14,18,19and in animal wound models.^15,16,31One may thus speculate that cell renewal induced by the growth factor’s well known mitogenic effect could have been responsible for the reduced overall cell damage in FGF- 2–supplemented corneas. No consistent BrdU labeling of the cells stored in basal serum-free medium, FGF-2-containing serum-free medium, or medium supplemented with 10% FCS was detected. However, significant labeling with BrdU was observed in cells surrounding a mechanically induced larger endothelial injury. This observation indicates that sufficient endothelial injury is necessary for the growth factor to exert a significant mitogenic effect on the in situ endothelium, because contact inhibition prevents FGF-2-induced proliferation.

For this reason, the beneficial effect of FGF-2 in prevention of pre- and postoperative endothelial cell loss is thus in pro- tecting the viability of the stored endothelial cells rather than in inducing mitosis. Furthermore, the results presented strongly suggest that the growth factor exerts a true protective effect on endothelial cells because in cell culture, serum-deprived BCECs underwent rapid apoptosis that was substantially prevented by adding FGF-2 to the culture medium. We did not obtain reproducible results when applying the TUNEL assay to the in situ endothelium after organ culture, although the fea- sibility of applying the technique to the whole cornea has recently been demonstrated.⁷

For both cell and organ culture, one has to be aware that the TUNEL assay does not exclusively detect apoptotic cells, because DNA fragmentation also occurs during necrosis. How- ever, whatever cellular mechanism may predominantly lead to endothelial cell death during corneal organ culture, a protective effect of FGF-2 prevents an elevated increase in damaged or dead cells which is at present inevitable when a serum- deprived standard organ-culture medium (mainly MEM) is used. In this context, the results are of interest for the current research on the composition of a suitable, defined serum-free organ culture medium.³²

Some cells maintained in FGF-2–supplemented medium showed intracytoplasmic vacuolization. This morphologic fea- ture appears in conditions of cellular stress but was detected in the cell culture experiments only. A formation of closely at- tached cells as in the organ-cultured cornea apparently in- creases the resistance of the individual cell to environmental stress factors.

The storage time of 22 days was chosen as an average preservation time of donor corneas in most eye banks. A concentration of 10 ng/mL recombinant human FGF-2 used in the comparative study with human corneas was found to be the most effective in previous cell culture experiments. Our studies with human corneal endothelial cells (HCECs) showed a dose-dependent, significant stimulation of HCEC proliferation with a peak at 10 ng/mL and a decrease at higher concentra- tions.³⁰

The half-life of FGF-2 activity at 32°C is estimated to be 24 hours,³³ but the fixation of FGF-2 to low-affinity receptors (heparan sulfate proteoglycans) present on the cell surface and in the basement membrane is very stable and has been shown in animal models to persist as long as 18 days after a single topical application on a cornea with denuded epithelium.³⁴ This binding probably reflects the main rationale for the use of FGF-2 in this context, because the bound growth factor re- mains biologically active³⁵and is protected against proteolytic degradation.³³This may explain why one application of FGF-2 per week was sufficient in our experiments to enhance the

survival of these cells. The reason for the absence of a protective effect when epidermal growth factor (EGF) was used in a commercial storage medium (Optisol; Chiron Vision, Irvine, CA) at 4°C (Briat B, David T, Serdarevic O, Gordon J, Renard G, Pouliquen Y, ARVO Abstract 1860, 1994)³⁶could thus be the different binding and kinetic properties of EGF and FGF-2.

At the light microscopical and ultrastructural level, we were not able to detect any morphologic cellular side effects related to the application of the growth factor. The ex vivo application of the growth factor is particularly advantageous in this context, avoiding problems of an intraocular application mainly related to the angiogenic potency of FGF-2.

In conclusion, this study demonstrates that FGF-2 as an additive to a serum-free culture medium protects the endothelium from damage that occurs during corneal storage. FGF-2 addition alone might not be sufficient to meet the demands of current cornea bank standards in providing sufficient endothelial protection comparable to serum-containing media. How- ever, the growth factor could be an important component of an advanced organ culture medium with a known, defined composition that replaces the present use of FCS and thus additionally avoids the potential transmission of contagious disease to the corneal transplant recipient.

Acknowledgments

The authors thank Christine Jaeckel and Sylvia Metzner for professional technical assistance throughout the project.

References

1. Pels E, Schuchard Y. Organ-culture preservation of human corneas.

Doc Ophthalmol.1983;56:147–153.

2. Maas-Reijs J, Pels E, Tullo AB. Eye banking in Europe 1991–1995.

Acta Ophthalmol Scand.1997;75:541–543.

3. Crewe JM, Armitage WJ. Integrity of epithelium and endothelium in organ-cultured human corneas. Invest Ophthalmol Vis Sci.

2001;42:1757–1761.

4. Armitage WJ, Easty DL. Factors influencing the suitability of organ- cultured corneas for transplantation.Invest Ophthalmol Vis Sci.

1997;38:16 –24.

5. Frueh BE, Bo¨hnke M. Corneal grafting of donor tissue preserved for longer than 4 weeks in organ culture medium.Cornea.1995;

14:463– 466.

6. Redbrake C, Salla S, Vonderhecken M, Frantz A, Reim M. Metabolic changes of the human donor cornea during organ culture.Acta Ophthalmol Scand.1999;77:266 –272.

7. Albon J, Tullo AB, Aktar S, Boulton ME. Apoptosis in the endothelium of human corneas for transplantation.Invest Ophthalmol Vis Sci.2000;41:2887–2893.

8. Ventura ACS, Engelmann K, Bo¨hnke M. Fetal calf serum protects cultured porcine corneal endothelial cells from endotoxin-medi- ated cell damage.Ophthalmic Res.1999;31:416 – 425.

9. Esch F, Ueno N, Baird A, et al. Primary structure of bovine brain acidic FGF.Biochem Biophys Res Commun.1985;133:554 –562.

10. Gospodarowicz D, Neufeld G, Schweigerer L. Molecular and biological characterization of FGF, an angiogenic factor which also controls the proliferation and differentiation.Cell Differ.1986;19:

1–17.

11. Lee PL, Johnson DE, Cousens LS, Fried VA, Williams LT. Purifica- tion and complementary DNA cloning of a receptor for basic fibroblast growth factor.Science.1989;245:57– 60.

12. Houssaint E, Blanquet P, Champion-Arnaud P, et al. Related fibroblast growth factor receptor genes exist in the human genome.

Proc Natl Acad Sci USA.1990;87:8180 – 8184.

13. Gospodarowicz D, Mescher A, Birdwell CR. Stimulation of corneal endothelial cell proliferation in vitro by fibroblast and epidermal growth factor.Exp Eye Res.1977;25:75– 89.

14. Gospodarowicz D, Greenburg G. The effect of epidermal and fibroblast growth factors on the repair of corneal endothelial IOVS,September 2003, Vol. 44, No. 9 FGF-2 and Corneal Endothelial Cell Survival 3831

(8)

wounds in bovine corneas maintained in organ culture.Exp Eye Res.1979;28:147–157.

15. Landshman N, Belkin M, Ben-Hanan I, Ben-Chaim O, Assia E, Savion N. Regeneration of cat corneal endothelium induced in vivo by fibroblast growth factor.Exp Eye Res.1987;45:805– 811.

16. Rieck P, Hartmann C, Pouliquen Y, Courtois Y. Recombinant human bFGF stimulates corneal endothelial wound healing in rabbits.Curr Eye Res.1992;19:153–160.

17. Rieck P, Cholidis S, Hartmann C. Intracellular signaling pathway of FGF-2-modulated corneal endothelial cell migration during wound healing in vitro.Exp Eye Res.2001;73:639 – 650.

18. Hoppenreijs VP, Pels E, Vrensen GF, Treffers WF. Basic fibroblast growth factor stimulates corneal endothelial cell growth and endothelial wound healing of human corneas.Invest Ophthalmol Vis Sci.1994;35:931–944.

19. Sabatier P, Rieck P, Daumer ML, Courtois Y, Pouliquen Y, Hart- mann C. Effects of human recombinant basic fibroblast growth factor on endothelial wound healing in organ culture of human cornea.J Fr Ophtalmol.1996;19:200 –207.

20. Sievers J, Hausmann B, Unsicker K, Berry M. Fibroblast growth factors promote the survival of adult rat retinal ganglion cells after transection of the optic nerve.Neurosci Lett.1987;76:157–162.

21. Factorovich EG, Steinberg RH, Yasumura D, Matthes MT, LaVail MM. Photoreceptor degeneration in inherited retinal dystrophy delayed by bFGF.Nature.1990;347:83– 86.

22. Plouet J, Courty J, Olivie´ M, Courtois Y, Barritault D. A highly reliable and sensitive assay for the purification of cellular growth factor.Cell Mol Biol.1983;30:105–110.

23. Hartmann C, Rieck P. A new test for endothelial viability: the Janus green photometry technique.Arch Ophthalmol.1989;107:1511–

1515.

24. Schilling-Scho¨n A, Pleyer U, Hartmann C, Rieck P. The role of endogenous growth factors to support corneal endothelial migration after wounding in vitro.Exp Eye Res.2000;71:583–589.

25. Anderson KJ, Dam D, Lee S, Cotman CW. Basic fibroblast growth factor prevents death of lesioned cholinergic neurons in vivo.

Nature.1988;332:360 –361.

26. Akimoto M, Miyatake S, Kogishi J, et al. Adenovirally expressed basic fibroblast growth factor rescues photoreceptor cells in RCS rats.Invest Ophthalmol Vis Sci.1999;40:273–279.

27. Zhang C, Takahashi K, Lam TT, Tso MOM. Effects of basic fibroblast growth factor in retinal ischemia.Invest Ophthalmol Vis Sci.

1994;35:3163–3168.

28. Renaud F, Desset S, Oliver L, et al. The neurotrophic activity of FGF-1 depends on FGF-1 expression and is independent of the MAP kinase cascade pathway.J Biol Chem.1996;271:2801–2811.

29. Sato Y, Rifkin DB. Autocrine activities of basic fibroblast growth factor: regulation of endothelial cell movement, plasminogen ac- tivator synthesis, and DNA synthesis.J Cell Biol.1988;107:1199 – 1205.

30. Rieck P, Oliver L, Engelmann K, Fuhrmann G, Hartmann C, Cour- tois Y. Role of exogenous/endogenous FGF-2 and TGF-␤on human corneal endothelial cells proliferation in vitro.Exp Cell Res.1995;

220:36 – 46.

31. Rich LF, Hatfield JM, Louiselle I. The influence of basic fibroblast growth factor on cat corneal endothelial wound healing in vivo.

Curr Eye Res.1991;11:719 –25.

32. Bednarz J, Doubilei V, Wollnik PC, Engelmann K. Effects of three different media on serum-free culture of donor corneas and isolated human corneal endothelial cells.Br J Ophthalmol.2001;85:

1416 –1420.

33. Saksela O, Moscatelli D, Sommer A, Rifkin DB. Endothelial cell- derived heparan sulfate binds basic fibroblast growth factor and protects it from proteolytic degradation.J Cell Biol.1988;107:743–

751.

34. Assouline M, Hutchinson C, Morton K. In vivo binding of topically applied human bFGF on rabbit corneal epithelial wound.Growth Factors.1989;1:251–261.

35. Rogelj S, Klagsbrun M, Atzmon R, et al. Basic fibroblast growth factor is an extracellular matrix component required for support- ing the proliferation of vascular endothelial cells and the differentiation of PC 12 cells.J Cell Biol.1989;109:823– 831.

36. Lass JH, Musch DC, Gordon JF, Laing RA. The Corneal Preservation Study Group. Epidermal growth factor and insulin in corneal preservation.Ophthalmology.1994;101:352–359.