Human Corneal Epithelial Equivalents for Ocular Surface Reconstruction in a Complete Serum-Free Culture System without Unknown Factors

Abstract

purpose. To establish a culture technique for human corneal epithelial equivalents that do not require fetal bovine serum (FBS), feeder cells, or bovine pituitary extracts and compare this system with conventional culture medium with FBS and mouse 3T3 fibroblasts.

methods. Human corneal limbal tissue from donor corneas was dissociated on denuded amniotic membranes and then cultured for 3 weeks in feeder-cell– and serum-free medium containing epidermal growth factor and B-27. Then, the cell sheet was evaluated by light microscopy, immunohistochemistry, and electron microscopy. The epithelial proliferative capacity was compared between serum- and feeder-cell–free medium and conventional medium. The cultured cell sheets were transplanted onto the denuded rabbit ocular surface to cover the resected area.

results. A stratified cell sheet expressing cytokeratin-3 and -12 was grown in serum- and feeder-cell–free medium without unknown growth factors. The epithelial proliferative capacity in feeder-cell– and serum-free medium determined by WST-1 and colony-forming efficiency was significantly higher than that in conventional medium. Scanning and transmission electron microscopy showed well-formed stratified epithelium with clear cell boundaries, microvilli, and hemidesmosomal/desmosomal junctions. The transplanted cell sheets remained transparent without epithelial defects during the follow-up period.

conclusions. This method using serum- and feeder-cell–free medium not containing unknown growth factors allows the highly proliferative culture of human corneal epithelium. It avoids exposure of the corneal epithelial equivalent to FBS and animal feeder cells, thus minimizing the risk of contamination by pathogens that could transmit diseases to recipients.

Adult stem cells that supply new epithelial cells for the cornea throughout life reside in the basal cell layer of the corneal limbal epithelium. ^{1 2 3 4 5 6} Ocular surface disorders, such as Stevens-Johnson syndrome, ocular pemphigoid, and chemical burns, are caused by a deficiency of limbal adult stem cells and are a contraindication for full-thickness corneal transplantation. To treat such severe ocular disorders, transplantation of a cultured corneal epithelial cell equivalent that contains potential adult stem cells would allow us to restore the ocular surface more quickly with stratified squamous epithelium growing on amniotic membrane as a scaffold. ^{7 8 9 10} However, this method currently requires unknown factors, such as fetal bovine serum (FBS), xenogeneic mouse 3T3-fibroblasts, or bovine pituitary extracts (BPEs) for successful growth of corneal epithelial equivalents. Coculture of human limbal stem cells with animal cells and FBS raises concern about infection with recognized or unknown agents. ^{11 12}  

Therefore, we tried to avoid the use of FBS, feeder cells (including 3T3 fibroblasts), and/or BPE during the growth of human corneal epithelial (HCE) cell equivalents, to minimize possible contamination during culture. We focused our attention on the growth factor mix, B-27. ¹³ B27 was originally optimized for culture of hippocampal neurons and used for the growth of neurons from embryonic rat striatum, substantia nigra, subiculum, cerebral cortex, postnatal dentate granule, cerebral cortex, cerebellum, and dentate gyrus in a serum-free condition. ¹⁴ Moreover, corneal endothelial precursors proliferate actively in B-27-containing medium with no FBS and feeder cells. ¹⁵ Accordingly, we investigated whether serum- and feeder-cell–free culture with B-27 was feasible, and eventually established a new culture method without unknown growth factors for HCE cell equivalents. 

Materials and Methods

Human Amniotic Membranes

Amniotic membrane tissues were obtained with the approval of the Institutional Review Board of the University of Tokyo Graduate School of Medicine and in accordance with the Helsinki Declaration of 1975 and its 1983 revision. After written consent was obtained from each subject, human placentas were harvested under sterile conditions during elective cesarean section and were the source of amniotic membranes. 

Culture of Human Corneal Limbal Epithelium

Corneas were obtained from the Rocky Mountain Lions’ Eye Bank at 3 to 5 days after harvesting. The ages of the donors ranged from 28 to 58 years. The limbal tissues, including the subconjunctival Tenon’s capsule, were cut into small pieces approximately 2 mm in diameter from clear cornea, which were incubated overnight at 37°C in basal culture medium (DMEM-F12, 1:1; Sigma-Aldrich, St. Louis, MO) with 20 ng/mL of epidermal growth factor (EGF; Wako Pure Chemical Industries, Ltd. Osaka, Japan), 0.02% type IA collagenase (Sigma-Aldrich), B-27 (Table 1) ¹³ (Invitrogen, Grand Island, NY), 100 U/mL penicillin, 100 μg/mL streptomycin, and 250 ng/mL amphotericin B. Cells were collected from the incubated tissues in tubes (Sumilon Stem Full; Sumitomo Bakelite Co., Ltd., Tokyo, Japan), allowed to stand in 0.05% trypsin/EDTA (trypsin-EDTA) for 10 minutes at 37°C, and then dissociated into single cells by pipetting. After addition of trypsin inhibitor (Invitrogen), the cells were resuspended in medium with 20 ng/mL of EGF, B-27, 100 U/mL penicillin, 100 μg/mL of streptomycin, and 250 ng/mL amphotericin B. Next, epithelial cells (1 × 10⁵ cells/well) were seeded onto pieces of denuded amniotic membrane spread over the culture inserts of six-well culture dishes. The cells were covered with medium for 3 weeks in the culture insert at 37°C in an atmosphere of 5% CO₂/95% air, and the medium was changed daily. In the conventional culture system ^{10 16} using mouse 3T3-fibroblasts and FBS for HCE equivalent, confluent 3T3 fibroblasts were incubated with 4 μg/mL of mitomycin C for 2 hours at 37°C under 5% CO₂, after which the cells were trypsinized and plated onto plastic dishes of outer wells at a density of 1 × 10⁵ cells per well. Corneal limbal epithelial cells (1 × 10⁵ cells/well) were seeded onto denuded amniotic membrane spread on the culture inserts, and coculture was performed with 3T3 fibroblasts. The culture medium was DMEM plus Ham’s F12 medium (Sigma-Aldrich) at a 1:1 ratio supplemented with 10% FBS, 0.5% dimethyl sulfoxide, 2 ng/mL EGF, 1 μg/mL recombinant human insulin (Wako Pure Chemical Industries, Ltd.), and 0.1 μg/mL cholera toxin (Sigma-Aldrich). The cell sheets were observed by phase-contrast microscopy. 

Histologic Examination

Corneal limbal epithelial cells cultured on amniotic membranes for 3 weeks were examined histologically. After they were fixed in 4% buffered paraformaldehyde, the epithelial samples were cut and embedded in paraffin. Then, 5-μm sections of each paraffin block were cut and placed on Silane-coated microscope slides (Matsunami, Osaka, Japan). These sections were stained with hematoxylin and eosin (HE) and observed under a light microscope. 

Immunohistochemistry

Corneal limbal epithelial cells cultured on amniotic membranes for 3 weeks were embedded in OCT compound (Tissue-Tek; Sakura Finetek, Torrance, CA) and 10-μm-thick cryosections were cut for immunostaining. The sections were incubated for 2 hours at room temperature with the primary antibodies, which were a mouse monoclonal antibody (mAb) for human cytokeratin-3 (AE-5, 1:10,000; Progen Biotechnik GMBH, Heidelberg, Germany), a mouse polyclonal antibody for human cytokeratin-12 (N-16, 1:500; Santa Cruz Biotech, Santa Cruz, CA), a mouse mAb for human cytokeratin-4 (M6B10, 1:10; ICN Pharmaceuticals, Aurora, OH), a mouse mAb for human cytokeratin-13 (1:5; American Research Products, Belmont, MA), a mouse mAb for human cytokeratin-1 (1:30), a mouse mAb for human cytokeratin-10 (1:80), and a mouse mAb for human nuclei (1:30; Chemicon, Temecula, CA). Each of these antibodies was added at a concentration of 0.2 μg/mL, after which the fluorescein isothiocyanate-conjugated secondary antibody was applied for 30 minutes. The cell nuclei were then stained with Hoechst 33342. As a negative control, the primary antibody was replaced with nonspecific mouse IgG1 or IgG2a (1.0 mg/mL) and normal rabbit serum diluted with PBS containing 1% bovine serum albumin and 0.03% Triton X-100. After four washes in PBS, the sections were covered with mounting medium (Vector Laboratories, Burlingame, CA) and were examined under a fluorescent microscope (model BH2-RFL-T3 or BX50; Olympus, Tokyo, Japan). 

Colony-Forming Assay

The clonal growth ability of cultured corneal limbal epithelium was evaluated by determining colony-forming efficiency (CFE). The cells were plated at a clonal density of 1000 cells onto six-well culture dishes. For the conventional system, 1 × 10⁴ of 3T3 fibroblasts were plated on the dishes. A colony was defined as a group of eight or more contiguous cells as described elsewhere. ^{17 18} The colonies were fixed on day 10 and stained with rhodamine B. The number of colonies was counted and the average number (n = 5 to 6) was recorded. The CFE was defined as follows: CFE (%) = (colonies formed at end of growth period ÷ total number of viable cells seeded) × 100%. 

WST-1 Cell-Proliferation Assay System

In serum- and feeder-cell–free and conventional system, the cell proliferative capacity was evaluated by a nonradioactive colorimetric assay (WST-1; Dojindo Laboratories, Kumamoto, Japan), based on the cleavage of a tetrazolium salt, as recommended by the manufacturer. Both the epithelial cells in insert culture dishes were transferred to DMEM/F12 basal medium and evaluated 1, 2, and 3 weeks after cultivation for comparison of the cell proliferative capacity. WST-1 reagent in DMEM/F-12 was applied for 20 minutes, and the dye was measured at 450 nm by a plate reader (Victor3 V Multilabel Counter model 1420; Perkin Elmer, Waltham, MA). DMEM/F12 medium was the negative control. The number of cells in each well was counted and the average number was calculated at each week. 

Scanning and Transmission Electron Microscopy

The samples were fixed in 2% glutaraldehyde in PBS, washed in PBS for 15 minutes, and postfixed in 2% osmium tetroxide for 2 hours. The samples were washed in PBS again before being passed through an alcohol series. After immersion in 100% ethanol twice for 20 minutes each, the samples were transferred to hexamethyldilisane for 10 minutes and air dried. The samples were mounted on aluminum specimen stubs and sputter-coated with gold before being examined under a scanning electron microscope (model JSM-6320F; JEOL, London, UK). Ultrathin sections were cut on a microtome (Ultracut E; Reichert Jung, Vienna, Austria), collected on bare copper grids, and stained with aqueous uranyl acetate, phosphotungstic acid, and lead citrate before examination by transmission electron microscopy (model JEM-2000EX; JEOL). 

Human Corneal Limbal Epithelium Transplantation onto Rabbit Corneas

The rabbits were treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. HCE sheet transplantation was performed according to a published technique. ¹⁹ Briefly, the entire corneal epithelium and surrounding conjunctival tissue in New Zealand White rabbits weighing 2.0 to 2.4 kg (Saitama Jikkenn, Saitama, Japan) were removed. The cultured HCE sheets were transplanted onto the denuded ocular surface, followed by six to eight sutures with 10-0 nylon (Mani, Tochigi, Japan). The graft was covered with a soft contact lens, and tarsorrhaphy was performed. After surgery, the rabbits were treated with 0.3% ofloxacin ointment (Santen Pharmaceutical Co., Ltd, Osaka, Japan), subconjunctival injection of 0.2 mL triamcinolone acetonide (Bristol-Myers Squibb Co., Tokyo, Japan), and intramuscular injection of 10 mg gentamicin (Nacalai Tesque Inc. Kyoto, Japan). FK-506 (Tacrolimus; 0.2 mg/kg, Astellas Co., Ltd., Tokyo, Japan) was injected intramuscularly daily during the observation period. Eyeballs with conjunctivae were removed 14 days after transplantation and were divided into two portions for immunohistochemical study and HE staining. 

Statistical Analysis

The Mann-Whitney U test was used to compare mean values for the groups. The level of significance was set at P < 0.05. Analyses were performed with a commercial statistical software package (Stat View, ver. 5; Abacus Concepts, Berkeley, CA). 

Results

Findings on HE Staining

A representative HE-stained cell equivalent after 3 weeks of culture is shown in Figure 1 . The cells cultured in the serum- and feeder-cell–free (Fig. 1A)and conventional (Fig. 1B)systems formed six to seven and four to five layers on the amniotic membrane scaffold, respectively. 

Cytokeratin Expression

After 3 weeks of culture, the cells showed no staining with nonspecific mouse IgG1 or IgG2a, or with normal rabbit serum (not shown). Cultured cells in both the serum- and feeder-cell–free and conventional systems expressed cytokeratin-3 (Figs. 2A 2B)and cytokeratin-12 (Figs. 2C 2D) , which are detected in human corneal epithelium, but not cytokeratin-4 or -13, which are detected in conjunctival epithelium (data not shown). Antibodies for cytokeratin-1 or -10, which are associated with physiological keratinization in the epidermis, did not react with the cultured cells (data not shown). 

Comparison of Clonal Growth in Serum- and Feeder-Cell–Free and Conventional Culture Systems

The CFEs were evaluated in corneal limbal epithelial cells using serum- and feeder-cell–free medium and conventional medium containing 10% FBS cocultured with 3T3 fibroblast feeder cells. Cells cultured in serum- and feeder-cell–free medium (Fig. 3 , left) showed a higher colony-forming ability in comparison with those in conventional medium (Fig. 3 , right). Significantly more CFEs were detected in serum- and feeder-cell–free medium than in conventional medium (P < 0.01; Fig. 3 , bottom). 

Comparison of Proliferative Capacity in Serum- and Feeder-Cell–Free and Conventional Culture Systems

Proliferative capacity in serum- and feeder-cell–free medium containing B-27 and EGF and conventional medium containing 10% FBS with 3T3 fibroblast feeder cells ^{16 18} were compared 1, 2, and 3 weeks after culture. Limbal epithelium cultured in serum- and feeder-free medium showed significantly higher WST-1 signal than those cultured in medium containing 10% FBS with feeder cells in both 1 and 2 weeks after culture (Fig. 4) . No significant differences were detected in both groups at 3 weeks after culture (Fig. 4 , right). Averaged cell number in serum- and feeder-cell–free and conventional medium was 80.2 ± 0.6 × 10⁴ and 23.2 ± 2.6 × 10⁴ in 1 week, 148.8± 4.1 × 10⁴ and 61.5 ± 4.5 × 10⁴ in 2 weeks, and 96.7 ± 1.5 × 10⁴ and (44.6 ± 1.5) × 10⁴ in 3 weeks, respectively. 

Scanning and Transmission Electron Microscopy Findings

To evaluate further the corneal epithelial cell equivalents cultured in serum-free medium without any feeder cells and BPEs, we examined the cell sheets after 3 weeks by scanning and transmission electron microscopy. Scanning electron microscopy showed a continuous layer of flat squamous polygonal epithelial cells in the cultures of corneal epithelium. These cells were closely attached to each other with tightly opposed cell junctions and had distinct boundaries (Fig. 5A) . On the surface of the superficial cell layer, numerous microvilli were observed (Fig. 5B) . Transmission electron microscopy of the cultured corneal epithelial equivalents showed that the cells had differentiated into basal columnar cells, suprabasal wing cells, and flat squamous superficial cells (Fig. 5C) . The basal cells were adherent to the amniotic membrane substrate by hemidesmosomal junctions (Fig. 5D)and the suprabasal wing cells were attached by desmosomal junctions and actin filaments (Fig. 5E) . Epithelial cells including wing cells and superficial cells were closely attached to their neighboring cells by numerous desmosomal junctions (Fig. 5C) . 

Transplantation of Sheets in the Serum- and Feeder-Cell–Free System

Corneal epithelial cell equivalents were transplanted onto the corneas of rabbits (n = 3). The transplanted epithelium covered all corneas and remained transparent and devoid of epithelial defects during the 2-week postoperative observation period (Fig. 6A) . No fluorescein-positive stainings were observed on the recipients’ corneal surface (Fig. 6B) . The epithelium showed four to five layers of stratified squamous epithelium, consistent with the findings of normal corneal epithelium (Figs. 6C) . Because the corneal epithelium was positive for the anti-human nuclei mAb, which reacts with human tissue, ¹⁹ the epithelia on the rabbit corneas were derived from humans (Fig. 6D) . Rabbit corneal keratocytes stained with Hoechst 33342 were negative with the anti-human nuclei mAb (Fig. 6D 6E) . Normal human cornea and rabbit cornea served as positive and negative control specimens, respectively (data not shown). Cytokeratin-3- and -12-positive stainings were obtained on the ocular surface of the rabbit eyes (data not shown). 

Discussion

Incubation in FBS, feeder cells, or BPE during the growth of corneal epithelial sheets increases the risk of contamination by known pathogens ^{20 21} and unknown agents. Because these agents may transmit various diseases to recipient patients, we should minimize this risk of cell culture when developing an ideal corneal epithelial equivalent for ocular surface reconstruction. HCE has been cultured in medium containing autologous serum and feeder cells, ¹⁸ bovine serum without feeder cells, ⁸ or serum-free medium with feeder cells. ²² Human conjunctival epithelium has also been cultured without animal serum or feeder cells, although the medium was supplemented with BPE, ²³ as well as in human serum without feeder cells. ²⁴ Recently, culture medium (EpiLife; Cascade Biologics, Inc., Portland, OR) enabled corneal epithelial cell culture without serum and feeder cells, but components of medium have been undocumented and the quality of corneal epithelial equivalent, including histologic findings and proliferative capacity has yet been reported. ²⁵ Although B-27-supplemented medium ¹³ contains purified bovine albumin derived from prion-free New Zealand cows, we cannot completely exclude the possibility of contamination by animal-derived factors. However, serum-, feeder-cell–, and BPE-free human corneal limbal epithelium equivalents can contribute to minimizing one of the major problems associated with the use of cultured human limbal stem cells. 

The cultured cell sheets obtained in this study had several layers, including amniotic membrane stroma, columnar basal cells, suprabasal wing cells, and elongated surface cells, on HE staining. Distinct cell boundaries, microvilli, basal columnar cells, suprabasal wing cells, flat squamous superficial cells, and numerous desmosomal–hemidesmosomal junctions were seen on electron microscopy. These findings were similar to the features of human corneal epithelium. Cytokeratins are epithelial intermediate filament proteins that have a crucial role in maintaining the structural integrity of the epithelium. Cytokeratin-3 and -12 are regarded as markers of corneal differentiation. Cells showing stratified growth on amniotic membranes in serum- and feeder-cell–free culture expressed cytokeratin-3 and -12, but not the conjunctiva-specific cytokeratin-4 and -13, demonstrating that the cell sheet maintained the characteristics of differentiated corneal epithelium. Limbal epithelium cultured in serum- and feeder-cell–free medium showed significantly higher WST-1 signal and CFEs than those cultured in conventional medium containing FBS cocultured with feeder cells, indicating that epithelium cultured in serum- and feeder-free medium has a stronger proliferative capacity than that cultured in conventional medium. Moreover, in vivo transplantation experiments on rabbit eyes suggest the feasibility of epithelia cultured in serum- and feeder-free medium in a clinical setting. We chose culture of a cell suspension and exposure of the new epithelium to produce a stratified corneal epithelial cell sheet. A well-stratified cell sheet was created by using a cell suspension, rather than limbal tissue explants in our preliminary study (data not shown). The limbal tissue explant technique is an alternative choice for cell sheet culture. This result was consistent with the findings that cultured epithelium obtained from cell suspensions was morphologically superior to explant culture epithelium. ¹⁶ The number of intermediate layers in the cultured stratified corneal epithelium we produced exceeded that of actual corneal epithelium. This finding is not a disadvantage for the clinical use of corneal epithelial sheets and conversely suggests that the culture period for construction of a stratified corneal epithelial sheet can be shortened. 

Regenerative medicine is not only a field of science; it is also an industry. Because treatment using allogenic limbal stem cells (with or without culture) requires immunosuppressive therapy, ^{26 27} the use of allogenic donor corneas is limited to severe cases and large-scale employment of corneal epithelial equivalents is unlikely to occur. Serum- and feeder-cell–free culture systems should have an advantage for the supply of high-quality cell equivalents to patients around the world and may contribute to progress in individualized medicine (i.e., small-scale regenerative medicine). Also, serum-, feeder-cell–, and BPE-free culture has marked advantages for studying the effects of specific growth factors, cytokines, and toxins on cell proliferation and differentiation, because it provides more defined conditions for investigating the effect of various factors on growth and proliferation by eliminating unknown growth factors. 

In summary, serum-, feeder cell-, and BPE-free culture medium containing B-27 and EGF made it possible to grow human corneal limbal epithelial cell equivalents. This method allows the highly proliferative culture of HCE in serum- and feeder-cell–free medium, in comparison with conventional medium containing FBS cocultured with mouse 3T3 fibroblast feeder cells. The culture system thus minimizes the risk of contaminating corneal epithelial cell equivalents during culture and provides more defined conditions for investigation of cell growth and proliferation by elimination of unknown factors contained in serum, feeder cells, and BPE. 

 Supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.

 Submitted for publication December 7, 2006; revised May 21 and August 10, 2007, and February 1, 2008; accepted April 24, 2008.

 Disclosure: S. Yokoo, None; S. Yamagami, None; T. Usui, None; S. Amano, None; M. Araie, 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.

 

Table 1.

 

Composition of B27 Medium Supplement

Insulin	Retinyl acetate
Selenium	dl-α-Tocopherol (vitamin E)
Transferrin	dl-α-Tocopherol acetate
Ethanolamine	l-Carnitine
Bovine serum albumin	Corticosterone
T3 (triodo-l-thyronine)	Progesterone
d(+)-galactose	Putrescine
Biotin	Glutathione (reduced)
Linoleic acid	Superoxide dismutase
Linolenic acid	Catalase

Figure 1.

 

HE staining of human corneal epithelium cultured under serum- and feeder-cell–free conditions and the conventional system for 3 weeks. (A) The cultured sheet in the serum- and feeder-cell–free medium formed six to seven cell layers on the amniotic membrane scaffold. (B) The cultured corneal epithelium in conventional medium formed four to five cell layers on the amniotic membrane. Original magnification, ×50; scale bar, 100 μm.

Figure 2.

 

Cytokeratin expression by human corneal epithelium cultured in serum- and feeder-cell–free conditions and in conventional medium. Nuclei were stained with Hoechst 33342. Cornea-specific cytokeratin-3 (A) and cytokeraitn-12 (C) were expressed in cellular cytoplasm of the epithelial cells in serum- and feeder-cell–free conditions. Similar expression pattern of cytokeratin-3 (B) and cytokeraitn-12 (D) was detected in the conventional system.

Figure 3.

 

Colony-forming assay. The CFEs were evaluated. Cells cultured in serum- and feeder-cell–free medium (top left) showed a higher colony-forming ability in comparison with those in conventional medium (top right). Significantly high CFEs in serum- and feeder-cell–free medium were detected compared with those in conventional medium (P < 0.001) (bottom). Similar findings were obtained in repeated experiments. (A) Feeder cell- and serum-free medium containing B-27 and EGF. (B) Conventional medium containing 10% FBS with 3T3 fibroblast feeder cells.

Figure 4.

 

Comparison of proliferative capacity by WST-1 assay. Proliferative capacity in feeder-cell– and serum-free medium containing B-27 and EGF (A) and conventional medium containing 10% FBS with 3T3 fibroblast feeder cells (B) were compared at 1 and 2 weeks after culture. Limbal epithelium cultured in serum- and feeder-free medium (n = 5) showed significantly higher WST-1 signal than those cultured in conventional medium containing 10% FBS with feeder cells (n = 5) in both 1 and 2 weeks after culture (*P < 0.05). No significant differences are detected in both groups at 3 weeks after culture. n.s., nonsignificant difference. Similar findings were obtained in repeated experiments.

Figure 5.

 

Scanning (A, B) and transmission (C, D, E) electron micrographs of HCE cultured for 3 weeks on an amniotic membrane scaffold in serum- and feeder-cell–free conditions. (A) Scanning electron microscopy showed a continuous layer of flat squamous polygonal epithelial cells. These cells were closely attached to each other by tightly opposed junctions and had distinct boundaries. (B) On the surface of the superficial cell layer, numerous microvilli were observed. (C) Transmission electron microscopy of the corneal epithelial sheet showed that the cells differentiated into basal columnar cells, suprabasal wing cells, and flat squamous superficial cells. The epithelial cells were closely attached to the adjacent cells by numerous desmosomal junctions. (D) Basal cells adhered to the amniotic membrane substrate by hemidesmosomal junctions (D, arrows). (E) Desmosomal junctions and actin filaments were observed on the membranes of suprabasal wing cells. Original magnification: (A, C) ×1000; (B) ×20,000; (D, E) ×10,000. Scale bars: (A) 10 μm; (B) 1 μm; (C) 5 μm; (D, E) 500 nm.

Figure 6.

 

HCE equivalents grafted onto the denuded rabbit corneas. The engrafted corneal epithelial equivalents were observed 14 days after transplantation. Representative photographs are shown. (A) Anterior segment photograph shows transparent and smooth surface under an operating microscope. (B) No fluorescein-positive stainings were detected on the engrafted corneal epithelial equivalents. (C) HE staining showed four to five stratified squamous cells, consistent with the histologic findings of normal corneal epithelium by light microscopy at 2 weeks. (D) The engrafted cornea was stained positively with anti-human nuclei monoclonal antibody. (E) The nuclei were counterstained with Hoechst 33342. Rabbit corneal keratocytes were stained by Hoechst 33342, but not anti-human nuclei monoclonal antibody.

The authors thank Kayo Aoyama and Toshiya Osawa for excellent technical support.