Reconstruction of damaged cornea by autologous transplantation of epidermal adult stem cells

Isolation and identification of epidermal stem cells from the goat ear skin

The isolation and identification of epidermal stem cells were performed as previously described [28]. Briefly, pieces of skin approximately 0.5 cm × 0.6 cm were cut from the ear of Guanzhong dairy goats (three to four years old). The tissues were rinsed three times with Medium 199 (Invitrogen-Gibco BRL, Grand Island, NY), which contained 50 μg/ml gentamicin and 1.25 μg/ml amphotericin B. After the careful removal of cartilage and conjunctive tissue, the remaining tissue was moved to a culture dish and cut into pieces approximately 1 mm × 2 mm × 3 mm using a scalpel. We then placed three to four pieces of the explant into each well of a six-well TC-Plate (Hyclone, Logan, UT) and added Medium 199 (Invitrogen-Gibco BRL, Grand Island, NY) supplemented with 15% neonatal bovine serum (Baian Biotechnology Corporation, Zhengzhou City, China), 5 μg/ml insulin (Sigma-Aldrich Corporation, St. Louis, MO), 0.5 μg/ml hydrocortisone, 40 μg/ml gentamicin, and 0.25 μg/ml amphotericin B. The explants were cultured at 37 °C in a 5% CO₂ /95% air incubator. The medium was changed every two to three days while the extent of each outgrowth was monitored with a phase contrast microscope. We selected the holoclones under dissecting microscope and treated them with 0.2% trypsin and 0.02% EDTA for approximately 10 min to disperse each clone to single cells. The trypsin was then inactivated by adding a medium containing 20% serum, and cells were gently centrifuged and resuspended in a serum free medium. The selected cells were seeded into culture dishes coated with gelatin type B (Sigma) in a medium containing 80% Ham F12 (Invitrogen-Gibco BRL, Grand Island, NY), 20% GSFCM (goat skin fibroblast-conditioned medium), 1 - 2% BSA (Sigma-Aldrich, St. Louis, MO), 20 ng/ml EGF (Sigma), 20 ng/ml IGF-1 (Sigma), 5 μg/ml insulin, 0.5 μg/ml hydrocortisone, and 50 μg/ml penicillin and streptomycin.

Primary putative epidermal stem cells were left to grow until they reached about 70% confluence in the culture plate. Adherent cells were then harvested with 0.2% trypsin and 0.02% EDTA to dislodge the cells followed by rapid trypsin neutralization to prevent cell damage. Isolated single cell suspensions were subcultured on culture dishes coated with gelatin at final plating density of 0.5–2 × 10⁵ viable cells per well. At each passage, cells from one well out of a set of triplicate cultures were detached, pooled, centrifuged, resuspended in fresh growth medium, seeded into three new wells, and expanded until they reached the same level of confluence. The growth medium was changed completely every other day. The cultures were monitored under a Leica Microsystems Wetzier microscope, LEICA MPS 60 (Leitz, Wetzlar, Germany).

To confirm the EpiASC derived from interfollicular epidermis, we dissected epidermis from dermis as previously described [30] with the following modifications. Skin samples were treated with 0.5% dispase II (Boehringer, Ingelheim, Germany) for 14–16 h at 4 °C to separate the interfollicular epidermis as a sheet from the underlying dermal tissue, leaving the hair follicular from the infundibulum to the dermal papillae within the dermis. The dermal surface was brushed gently with curved forceps to release any loosely attached basal keratinocytes, and then both interfollicular epidermis and dermis were cultured in the same medium. The cultures were monitored under a Leica Microsystems Wetzier microscope LEICA MPS 60 (Leitz Wetzlar).

Preparation of human amniotic membrane and cultivation of autologous epidermal stem cells on human amniotic membrane

Human amniotic membranes (HAMs) were obtained at the time of Cesarean section with proper informed consent in accordance with the tenets of the Declaration of Helsinki for research involving human subjects and upon approval by the Institutional Review Board of the Northwest A&F University (Shaanxi, Yangling, China). Under sterile conditions, the membranes were washed with sterile phosphate-buffered saline (PBS) containing antibiotics (200 IU/ml penicillin and 200 IU/ml streptomycin) then were incubated with 2.5% dispase at 4 °C for 12–20 h to loosen cellular adhesion followed by gentle scraping with a cell scraper and washing three times with PBS. The denuded HAM was frozen and stored at −80 °C in pure glycerol. Immediately before use, the HAM was thawed, washed three times with PBS containing antibiotics, cut into approximately 4 cm x 4 cm pieces, and flattened with the stromal side on the surface of a nitrocellulose paper (NCP) framework.

We used passage three to four putative EpiASC as seed cells. The EpiASC (1 × 10⁵ cells/ml) were seeded on the denuded HAM spread upon the NCP frame in a culture dish. The culture was submerged in a serum-free medium as described above for one week and then exposed to air by lowering the medium level over a two week period. The medium contained 80% F12, 20% GSFCM, 1%–2% BSA, 20 ng/ml EGF, 20 ng/ml IGF-1, 5 μg/ml insulin, 0.5 μg/ml hydrocortisone, 50 μg/ml Vitamin C (Sigma), and penicillin and streptomycin (50 μg/ml each). Cultures were incubated at 37 °C in a 5% CO₂/95% air incubator for up to 21 days, and the medium was changed every day.

Creation of total limbal stem cell deficiency model

The model of total LSCD was created in the left eye of 30 goats using a method previously described [28] with slight modifications. Briefly, after anesthesia, an ocular surface injury was created in one eye of each of the 26 healthy Guanzhong dairy goats by excising all the conjunctival tissue with 5–7 mm of the limbus and performing a superficial keratectomy of the central corneal surface including limbal epithelial cells, and then the limbus was burned by mechanical erasure with 1 N NaOH. The goats received chloramphenicol and dexamethasone eye drops twice a day during the first week and the necessary care for 28 days after surgery. LSCD was verified using clinical signs such as corneal haze, vascularization, epithelial irregularity, and the presence of conjunctival goblet cells on the corneal surface.

Transplantation of autologous cultivated epidermal stem cells on human amniotic membrane sheets

Four weeks after the ocular surface injury, the conjunctivalized ocular surfaces of 10 goats were surgically reconstructed by receiving transplants of EpiASC-HAM sheets (EHS), eight animals received HAM only transplantation (first control group), and the remaining eight underwent removal of the fibrovascular pannus without any transplantation (second control group) [28]. After surgery, the eye was covered with sterile gauze. From the first postoperative day, the goats were treated with topical preservative-free 1% methyl prednisolone acetate and 0.3% ofloxacin four times a day and with intramuscular penicillin sulfate and streptomycin sulfate every day for a week.

Histology and immunohistochemistry

To identify whether epidermal stem cells expressed stem cell markers and to compare the protein expression profile of normal goat cornea with damaged and reconstructed corneal epithelium, the resultant cells and corneas were collected, fixed in 4% paraformaldehyde, and embedded in tissue Tek OCT compound (Electron Microscopy Sciences, Washington, PA) for sectioning. For hematoxylin and eosin (H&E), PAS, and AB-PAS staining, slides were re-fixed, washed, and stained by their respective methods. For immunofluorescent staining, slides were re-fixed and rinsed three times (15 min each time) in PBS that contained 0.5% Triton-X 100 and then incubated with 2% BSA in PBS for 60 min at 37 °C. Primary antibodies (for sources and working dilutions, see Table 1) were applied to tissues for 2 h at 37 °C. After washing, fluorescein-conjugated secondary antibodies were applied for 30 min at 37 °C. Immunofluorescence microscopy was performed with a Nikon microscope (Nikon, Tokyo, Japan).

Immunocytochemistry staining was done by the PV-6002 Histostain™–Plus method, using kits from Beijing Zhongshan Biological and Technological Co. (Beijing, China). Sections were blocked with normal goat serum for 15 min at room temperature and incubated with primary antibodies (Table 1) overnight at 4 °C. Cells were rinsed with 0.1 M PBS without calcium and magesium and incubated with the secondary goat anti-mouse IgG for 15 min at 37 °C. The samples were washed and then incubated with the streptavidin/peroxidase work medium (S-A/HRP) for 15 min at 37 °C. Cells were rinsed in 0.1 M PBS without calcium and magesium, and the samples were finally incubated with the 3, 3′-diaminobenzidine (DAB) peroxidase substrate to give a brown stain and counterstained with hematoxylin or with chromogen solution, BCIP/NBT, to give a blue stain. After washing with PBS, the mounted sections were examined and photographed with an epifluorescent microscope, (Nikon, Tokyo, Japan).

The characteristics and origin of epidermal adult stem cells

We first cultured skin explants in a serum-based medium. After five to six days in culture, we found epithelial-like cells growing in a mosaic pattern surrounding the explant. Next, small, blast-like cells with a high nuclear/cytoplasmic ratio appeared on the surface of the epithelial-like cells. After 12–15 days, these cells formed clones with diameters of approximately 50–100 μm (Figure 1A). Immunohistochemistry showed that these primary EpiASCs either in clusters or single were Oct4 positive (Figure 1B,C). These clones were picked under the dissecting microscope, trypsinized into single-cell suspensions, and cultured in a serum-free medium on gelatin-coated dishes. Subsequently, the adherent cells formed holoclones (Figure 1D). These clones were expanded in vitro. Immunohistochemistry showed that the holoclone cells strongly expressed CK19, P63, β1-integrin, and CD34, a group of previously reported EpiASC markers (Figure 1E-H). Surprisingly, we found ~50%of subcultured EpiASC still strongly expressing Oct4 (Figure 1). To confirm the origin of isolated EpiASC, we separated the epidermis from the dermis by dispase II treatment and cultured both of them in the same medium. After three to four days, we found that cells that were derived from the epidermis expand in a mosaic pattern (Figure 2A), which is different from the fibroblast cells derived from dermis (Figure 2B). After 8–10 days, cell clones were formed in the surface of epithelial cells, which exhibited the same properties as stem-like cell clones from total skin explant culture (Figure 2C). In contrast, the fibroblast cells became confluent, and no stem-like cell clones were observed (Figure 2D). These results showed that EpiASC are derived from the epidermis of the skin.

Human amniotic membrane serves as a niche for epidermal adult stem cells

Preserved HAM transplantation alone was previously shown to be sufficient to reconstruct the corneal surface in cases of partial LSCD [31,32], suggesting that HAM may help expand the residual limbal epithelial stem cells. Clinical and experimental studies have also shown that ex vivo-expanded limbal epithelial cells on HAM are capable of restoring the corneal surface in LSCD [32]. Meller et al. [33] provided evidence that HAM cultures preferentially preserve and expand limbal epithelial stem cells that maintain their in vivo properties of slow cycling, label retention, and undifferentiation. We proposed that HAM might be an ideal matrix for ex vivo preservation and expansion of EpiASC. To investigate this, we seeded EpiASC that were isolated from a single cell (Figure 3A) on denuded HAM (Figure 3B). The culture was submerged in a serum-free medium for one week and then exposed to air by lowering the medium level over two weeks. We found that EpiASC adhered to HAM within 20–30 min, forming EHS. EpiASC remained quiescent during the first two days and then proliferated quickly during a further three days culture. We also observed that EpiASC formed clones in a similar way when co-cultured with 3T3 cells [data not shown]. To promote EpiASC proliferation, we exposed the surface of EHS to the air a week after culture and added 50 μg/ml Vitamin C to the medium. We found that EpiASC proliferation was arrested and differentiation had begun. The surface of EHS was covered with flattened cells, which expressed CK1/10 (Figure 3C). When the EHS was digested with 0.25% trypsin for 20–25 min, we found that a single cell layer was firmly attached to the HAM. Staining with anti-P63 antibody revealed that P63-positive cells were surrounded by groups of P63-negative cells (Figure 3D). This suggested that in the serum-free culture conditions, EpiASC reconstructed their niche on the surface of HAM. Immunostaining also showed that the P63 expression pattern of the EHS (Figure 3E) was similar to the pattern observed within the basal membrane layer of corneal epithelium (Figure 3F), further evidence that HAM can serve as a niche for EpiASC and can support EpiASC proliferation in vitro.

Clinical results of transplantation of EpiASC-HAM sheet to the corneal surface

We have previously reported the phenotypic outcome of transplantation of EHS to the corneal surface of total LSCD with a follow-up period of 24 months [28]. During this study, with an average follow-up period of 30 months, approximately 30% (3/10) of the eyes of treated animals regained four quadrants of clear cornea. This was assessed as a success according to the criterion of Ti et al. [29]. However, part of the limbus of these three eyes was unclear, which is distinct from normal eyes (Figure 4A-D). Half (5/10) of the eyes of the treated animals regained three quadrants of clear cornea, which was deemed a partial success [29], but the further limbus of these eyes was opaque (Figure 4E-G). The remaining 20% (2/10) failed to regain two quadrants of clear cornea (data not shown). The corneas of the control eyes were covered with pannus (Figure 4H,I), which manifested as vascularization and chronic inflammation of the cornea, ingrowth of fibrous tissue, and corneal opacification.

Epidermal adult stem cells can be converted into corneal epithelium by corneal stroma

We have previously reported that autologous EpiASC transplantation reconstructs the ocular surface of total LSCD to functionality [28], which only involves the morphological and histological features of the reconstructed ocular surface. However, the fate of grafted EpiASC after transplantation was not studied. In the present study, we compared the features of normal corneal epithelium (NCE), reconstructed cornea epithelium (RCE), and the pannus (P) of control groups by H&E, PAS, and immunocytochemistry staining. The results showed that reconstructed cornea shares certain features with normal cornea such as the corneal surface was smooth and the epithelium was integral without goblet cells (Figure 5A,B). In contrast, the corneal surface in the control group was covered with fibers, blood vessels, and goblet cells (Figure 5C,D). Interestingly, AB-PAS staining showed that both normal and reconstructed corneal epithelium were red (Figure 5E,F), which is characteristic of neutral mucus. In contrast, the goblet cells in the control group were purple (Figure 5G), characteristic of mixed mucus. These findings suggest that the corneal epithelium of reconstructed cornea has a function of secretion similar to normal corneal epithelium and distinct from invading goblet cells.

To confirm whether EpiASC may serve as LSCs to functionally reconstruct the damaged cornea of LSCD, a series of markers of corneal epithelium were compared among NCE, RCE, and P. The results showed that the RCE and NCE exhibit similar expression profiles of CK3, CK12, and PAX-6 (Figure 5H-M). The pannus expressed PAX-6 (Figure 5N) but not CK3 (Figure 5O). These data provide good evidence that EpiASC regenerated the function of damaged corneal surface caused by total LSCD.

Although we could not provide direct evidence that the reconstructed corneal epithelium were derived from grafted EpiASC (because of the lack of a valid marker to trace the fate of EpiASC during host wound repair), there is indirect evidence suggesting that the reconstructed corneal epithelium derives from grafted EpiASC. To show this, the expression profiles both of CK1/10 and PAX-6 were compared in normal skin (NS) and RCE six months after the operation and 12 months after the operation. We found that the expression of CK1/10 in NS is much greater than in RCE six months after operation (Figure 6A-B); however, PAX-6 was strongly expressed in the nucleus of epithelial cells in the RCE (Figure 6C). Interestingly, CK1/10 was also expressed in the cell cytoplasm of limbal and central corneal epithelial basal layer, which is different from the normal location of CK1/10 in NS (Figure 6D). NS is typically expressed in the superficial epithelial cells. Furthermore, CK1/10 was not expressed by the RCE 12 months after the operation (Figure 6E) while PAX-6 was strongly expressed in the nucleus of epithelial cells of the RCE (Figure 6F). This suggests that the expression of CK1/10 was down-regulated by the corneal stroma of the host animal during the process of corneal epithelium regeneration and that PAX-6 expression was upregulated.