Porcine Corneal Epithelial Cells Consist of High- and Low-Integrin β1–Expressing Populations

Abstract

purpose. Immunohistochemically, all cells in the porcine corneal epithelium, from the superficial to the basal layer, express integrin β1. Flow cytometric study has shown that they comprise integrin β1 high-expressing (β1²⁺) and low-expressing (β1⁺) populations. This study was undertaken to determine their proliferation characteristics.

methods. Epithelial cells from porcine corneas were sorted and labeled with anti-integrin β1 antibody and a fluorescent-dye–conjugated secondary antibody. The fluorescent intensity of labeled cells was analyzed and β1²⁺ and β1⁺ cells were cultured in an adhesive-coated culture plate.

results. Flow cytometry demonstrated that the epithelial cells comprised two distinct populations with a similar ratio throughout the cornea. Whereas β1²⁺ cells attached and grew to confluence in the plate, β1⁺ attached only transiently to the plate and exhibited minimal growth.

conclusions. The data indicate that the porcine cornea contains two distinct populations of epithelial cells, one exhibiting high and the other low integrin β1 expression. The observation that β1²⁺ cells had greater growth potential suggests that they may represent an enriched population of transit-amplifying cells.

Anonkeratinized, stratified corneal epithelium is essential for proper vision. Corneal epithelial cells are thought to derive from stem cells located in the limbus, which is the transitional zone between the cornea and conjunctiva. ^{1 2} Corneal epithelial transit-amplifying (TA) cells arising from the stem cells rapidly divide to produce daughters during their migration from the peripheral to the central cornea. ³ In patients with corneal stem cell deficiency after chemical burns or due to Stevens-Johnson syndrome, cultivated cells from corneal and limbal epithelia have been used for ocular surface reconstruction. ^{4 5 6} The quality of the cultivated graft is the key to success. We contend that the selection of a large number of highly proliferating cells (i.e., stem cells and TA cells) in the epithelia enhances the reproducibility, quality, and longevity of these grafts. 

Integrins, a large family of heterodimeric glycoproteins that attach cells to extracellular matrix proteins or to ligands on other cells, consist of noncovalently bound transmembrane α- and β-subunits. To date, 16 α- and 8 β-subunits have been identified. One of the β-subunits, integrin β1, associates with 10 different α-subunits (α1-9 and αv) to form receptors that bind several extracellular matrices, including fibronectin, collagens, and laminins. Integrin β1 is an adhesion molecule; high levels of integrin β1 expression reportedly are a surface marker of stem cells in the skin epidermis, ⁷ hair follicles, ⁸ and limbus (Yiu SC, et al. IOVS 2001;42:ARVO Abstract 2575 and Yiu SC, et al. IOVS 2003;44:ARVO E-Abstract 1351). Although whether the cells with high levels of integrin β1 are appropriate as stem cells of these tissues needs further investigation, there seems to be no doubt that the cells have a high proliferative potential. 

Our studies of the expression of integrin β1 by porcine corneal epithelial cells revealed that they comprised populations of integrin β1 high-expressing (β1²⁺)- and low-expressing (β1⁺) cells. We successfully isolated viable integrin β1–expressing cells and examined the proliferation characteristics of each of these populations. 

Materials and Methods

Animal tissues were obtained and handled in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Normal pig eyes from an abattoir were transported to the laboratory on ice and used immediately. 

Immunofluorescence Labeling and Flow Cytometry

Corneas were carefully dissected from the limbus and abraded with a cotton bud to remove the endothelium. To collect corneal epithelium, five corneas were cut into approximately 5-mm-wide strips and incubated for 5 hours at 37°C with 0.15 mg/mL thermolysin (Sigma-Aldrich, St. Louis, MO) in Dulbecco’s modified Eagle’s medium (DMEM; Invitrogen Corp., Carlsbad, CA) and then gently scraped with a cell scraper. ^{9 10} To collect central and peripheral epithelial cells, whole corneas were divided into central and peripheral regions with an 8-mm trephine before enzymatic digestion. Epithelia were pooled and treated with versene 0.02% EDTA solution (Invitrogen Corp.) to stop the digestion and prepare single-cell suspensions. Isolated cells were resuspended in phosphate-buffered saline (PBS) containing 0.5% bovine serum albumin and 2 mM EDTA and incubated for 10 minutes on ice with a diluted (1:500) mouse monoclonal anti-integrin β1 antibody (Chemicon International, Temecula, CA) that recognizes porcine integrin β1 and does not affect integrin-mediated cell adhesion. The cells were then washed in PBS and incubated with Alexa Fluor 488-conjugated anti-mouse IgG (Molecular Probes, Inc., Eugene, OR). After the cells were washed with PBS, they were filtrated through a 35-μm mesh, immediately analyzed, and sorted (FACS Vantage SE; BD Biosciences, San Jose, CA). Just before inoculation into a culture plate, the viability of sorted cells was determined to be >85% by the trypan blue exclusion test. 

Cell Culture

A 24-well culture plate was coated with 5 μg/mL tissue adhesive (Cell-Tak; BD Biosciences) according to the manufacturer’s instructions. Cell-Tak, an adhesive material extracted from marine mussel Mytilus edulis, is used to stick cells and tissues without cellular adhesion molecules, such as integrins, to glass, plastics, and metals. After flow cytometry, 1 × 10⁴ cells/cm² were inoculated into the adhesive-coated plate with serum-free medium (RCGM2; Kurabo Industries, Ltd., Osaka, Japan) that contained epidermal growth factor (10 ng/mL), insulin (5 μg/mL), hydrocortisone (0.5 μg/mL), and bovine pituitary extract (0.4%) and cultured at 37°C in a 5% CO₂–95% air atmosphere. The medium was changed after the first day of incubation and every 2 days thereafter. 

Colony-Forming Efficiency

To determine colony-forming efficiency (CFE), cells were cultured on an inactivated 3T3 feeder layer, according to a modification of a previously described method. ^{11 12 13} Briefly, 3T3 fibroblasts were exposed for 2 hours at 37°C to mitomycin C (Roche Diagnostics GmbH, Mannheim, Germany), to inactivate their proliferative activity and then were inoculated at 2.5 × 10⁴ cells/cm² into 35-mm culture dishes. Aliquots of sorted cells suspended in culture medium (DMEM and Ham’s F12, 1:1; Invitrogen Corp.), 10% FBS, insulin (5 μg/mL, Sigma-Aldrich), cholera toxin (0.1 nM, Calbiochem, La Jolla, CA), and human recombinant epidermal growth factor (10 ng/mL; Invitrogen Corp.) were seeded (2 × 10³ cells/cm²) into dishes containing an inactivated 3T3 feeder layer. After a 9-day culture, they were fixed with buffered formalin for 2 hours and then stained with 0.01% toluidine blue to visualize cell colonies. ^{7 11} Colonies containing >10 cells were carefully counted using a microscope. 

Statistical Analysis

The number of colonies in these populations were expressed as the mean ± SEM. Probability was evaluated by Student’s t-test. 

Results

Similar to the human cornea, the porcine corneal epithelium is highly stratified into seven to nine layers and the cells can be classified as basal-, wing-, and superficial cells. ¹⁴ In addition, the porcine cornea has a Bowman’s layer similar to that in human eyes (Fig. 1A) . ¹⁵ The fluorescence of an integral protein, integrin β1, was identified around each cell, from the superficial to the basal layer, and at the border between the epithelium and Bowman’s layer. Under a fluorescence microscope, isolated cells exhibited diverse fluorescence intensity for integrin β1 (Fig. 1B) . Flow cytometric analysis showed that they were all integrin β1 positive with a broad range of fluorescence intensity (Fig. 2A) , compared with negative control (Fig. 2C) . The fluorescence profile revealed a marked bimodal pattern (Figs. 2A 2B) , suggesting that there were two epithelial cell populations: One population had a high and the other a low level of integrin β1 expression (β1²⁺ and β1⁺) cells, respectively. In the whole porcine corneal epithelium, 56.0% of the cells were β1⁺ and 41.7% were β1²⁺ cells. These percentages were almost the same in the peripheral (50.1% and 45.1%) and central portions (56.4% and 39.7%; Figs. 2D 2E ) and in the peripheral cornea containing the limbus. 

The viability of sorted β1⁺ and β1²⁺ cells exceeded 85% before their inoculation into the adhesive-coated culture plate. As shown in Figure 3 , during the first day of culture, some β1⁺ cells became attached to the plate (Fig. 3A) ; by the seventh day of culture, almost no β1⁺ cells remained in the plate (Fig. 3C) . On the other hand, some β1²⁺ cells readily attached during the first day of culture (Fig. 3B) and grew to confluence by day 7 (Fig. 3D) . 

For CFE assay we used cells grown on an inactivated 3T3 feeder layer, because the growth of corneal epithelial cells requires seeding at high density without feeder cells. On day 9 of culture, we counted 554.7 ± 14.1 β1²⁺ and 56.3 ± 7.3 β1⁺ colonies (Fig. 4) , indicating that the CFE of β1²⁺ cells was significantly higher (P < 0.001). 

Discussion

Our data show that the porcine corneal epithelium comprises two cell populations, β1⁺ and β1²⁺, and that there is little difference between the central and peripheral portions with respect to the distribution of these populations. The β1²⁺ cells proliferated well, whereas β1⁺ cells did not. We posit that the poor proliferation of β1⁺ cells was independent of their adhesion to the culture substrate because some of β1⁺ cells forcibly attached with adhesive (Cell-Tak; BD Biosciences) during the first day of culture. 

Regenerative therapy, including tissue reconstruction with stem cells or proliferative precursor cells, has attracted attention. These procedures involve cell-sorting to facilitate the harvest of targeted viable cells. Flow cytometry cell-sorting, with combined antibodies specific for cell-surface antigens, has become the standard. However, there are few reports of the use of this technique for the selection of corneal epithelial cells. Flow cytometry cell-sorting requires viable single-cell suspensions; however, it is difficult to prepare these from epithelium. 

We tested some dissociation enzymes and reagents including dispase and trypsin and found that a combination of thermolysin and EDTA allowed us to prepare single-cell suspensions of viable corneal epithelial cells. Thermolysin is a kind of protease, and it has been reported that thermolysin-exposed keratinocytes retain good proliferative potential. ⁹ Exposure to thermolysin and EDTA completely removed corneal epithelial cells from the corneal stroma and facilitated their separation into viable single cells. When we used dispase, the presence of sticky, digested materials prevented the preparation of single-cell suspensions. 

Regenerative stratified epithelia are thought to contain two types of proliferative cells: stem cells and TA cells. ¹⁶ Under normal conditions, stem cells are a minor subpopulation of relatively quiescent or slow-cycling cells, characterized by their great proliferative potential and unlimited capacity for self-renewal. TA cells are the progeny of stem cells and terminally differentiate after a few cell cycles. Based on the in vitro proliferative potential of these cells, Barrandon and Green ¹⁷ classified proliferative cells as holoclones, paraclones, and meroclones. Holoclones have the highest proliferative capacity and are considered stem cells. Paraclones are TA cells with low proliferative capacity, and meroclones are an intermediate type of cell and are considered “young” TA cells endowed with greater proliferative capacity than paraclones. 

Based on in vitro clonal analysis and the analysis of labeled cells, stem cells of the corneal epithelium are thought to be located exclusively in the limbus. ^{18 19 20} Because our study showed that β1²⁺ cells were distributed in almost identical numbers throughout the corneal epithelium, even in the central portion, they may represent an enriched subpopulation of TA cells and/or meroclones in the porcine corneal epithelium. 

It remains to be determined whether central and peripheral β1²⁺ cells exhibit the same characteristics. We posit that their proliferative potential is similar under serum-free culture conditions, because there was no difference between peripheral and central corneal epithelial cells in such conditions. ^{21 22}  

The CFE of the β1²⁺ population was 2.8% (555 colonies of >10 cells from 2 × 10³ seeded cells). Although the value we obtained is small, we cannot conclude that it is insignificant. Our β1²⁺ cells may have been exposed to stress because they were sorted before seeding. ²³  

We found that porcine corneal epithelial cells comprise two distinct populations with different levels of integrin β1 expression and successfully sorted them. Only the β1²⁺ population exhibited strong proliferative activities. In porcine corneal epithelium, they are an enriched subpopulation of TA cells rather than stem cells. Studies are under way in our laboratory to strengthen our contention that for successful human ocular surface reconstruction and for the reproducibility, quality, and longevity of cultured grafts, the transplants must contain a large number of highly proliferative cells. 

 Submitted for publication April 5, 2004; revised June 3 and June 30, 2004; accepted July 10, 2004.

 Disclosure: K. Endo, None; T. Nakamura, None; S. Kawasaki, None; S. Kinoshita, None

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

 

Figure 1.

 

Integrin β1 immunofluorescence on porcine corneal epithelium and isolated porcine corneal epithelial cells. (A) The epithelial section was incubated with mouse anti-integrin β1 antibody (1:100) and then stained with Alexa Fluor 488 (green)–conjugated anti-mouse IgG. Cell nuclei in these tissues were simultaneously stained with propidium iodide (red). (B) Isolated cells with a combination of thermolysin and EDTA were labeled with the same antibody set. Bars, 100 μm.

Figure 2.

 

Flow cytometric analysis of integrin β1 expression by porcine corneal epithelial cells. (A) Histogram of cells obtained from whole corneas. The bimodal fluorescence pattern for integrin β1 was consistently observed in all experiments. (B) Two cell populations are apparent in the density plot of integrin β1 versus side scatter. (C–E) Histograms of cells exposed to mouse IgG instead of anti-integrin β1 antibody (C, negative control). Cells were derived from the central (D) and peripheral (E) portion of the cornea. Percentages of β1⁺ and β1²⁺ cells are indicated in (A), (D), and (E).

Figure 3.

 

Phase-contrast micrographs of cultured porcine corneal epithelial cells. β1⁺ (A, C) and β1²⁺ (B, D) cells were cultured in serum-free medium on a plate coated with tissue adhesive. On day 1 (A, B), attachment to the plate of β1⁺ and β1²⁺ cells was almost equal. (C) On day 7, almost no β1⁺ cells were attached to the plate. (D) In contrast, β1²⁺ cells had grown to confluence. Scale bar, 200 μm.

Figure 4.

 

Clonal growth of β1⁺ (top row) and β1²⁺ (bottom row) cells on an inactivated 3T3 feeder layer. The cells were seeded at a density of 2 × 10³ cells/cm² and stained with toluidine blue on day 9.