Human Umbilical Cord Lining Cells as Novel Feeder Layer for Ex Vivo Cultivation of Limbal Epithelial Cells

Abstract

Purpose: To determine the effectiveness of human umbilical cord-derived mucin-expressing cord lining epithelial cells (CLEC-muc) as feeder cells in a coculture system for the cultivation of human limbal stem cells.

Methods: Human CLEC-muc were cultured in PTTe-1 medium and treated with mitomycin C to arrest their growth to make the feeder layer. Single-cell suspension of limbal cells was prepared from corneal rim collected from the Singapore Eye Bank. Limbal cells were cultured in a coculture system with CLEC-muc as well as 3T3 cells as feeder layer. We compared the colony-forming efficiency and cell morphology of the limbal cells cultured in the two different feeder layers. We also compared the expression level of several putative limbal stem cell markers, such as HES1, ABCG2, ΔNP63, and BMI1, in the cultured limbal cells by immunostaining and quantitative (q)RT-PCR. Expression of cytokeratins CK14, CK15, CK19, CK3, and CK4 was further compared.

Results: Human limbal epithelial cells cultured in both types of feeder layers showed comparable cell morphology and colony-forming efficiency. These cells exhibited a similar expression pattern of HES1, ABCG2, ΔNP63, BMI1, CK14, CK15, CK19, and CK3 as detected by immunostaining and PCR.

Conclusions: Human CLEC-muc may be a suitable alternative to conventional mouse 3T3 feeder cells, which may reduce the risk of zoonotic infection.

Stem/progenitor cells located at the basal layer of limbal epithelium play crucial role in maintenance of corneal transparency and repairing of damaged corneal surface.^1,2 Several ocular surface diseases are associated with significant loss of stem cells, which results in a decrease in turnover of the required number of corneal cells.³ As a consequence, the cornea lacks functional integrity and the person proceeds toward blindness. Ex vivo expansion of autologous limbal epithelial stem cells or oral mucosal cells can produce the necessary cell population, which has been used as a surgical graft to treat patients with total or partial limbal stem cell deficiency.^2,4,5 This strategy, however, involves the use of growth-arrested murine 3T3 fibroblast feeder layers in a coculture system to support the epithelial stem/progenitor growth.^6,7 As 3T3 feeder cells originate from a nonhuman source, their use in ex vivo expansion of human cells to be transplanted carries the risk of transmission of zoonotic diseases from animal feeders to human cells. It has been reported that human embryonic stem cells cultured on mouse feeder layers generate an immunogenic nonhuman sialic acid.⁸ Therefore there is a growing need to substitute 3T3 cells with a human equivalent to avoid xenotoxicity. 

Several studies demonstrate that amniotic epithelial cells, human fibroblast cells, human mesenchymal stem cells, limbal mesenchymal cells, and adipocytes can be challenged to replace 3T3 feeder cells in a coculture system; however, none of these candidates have been so far suggested to be an ideal replacement for 3T3 cells.^9–13 We hypothesized that mucin-expressing cord lining epithelial cells (CLEC-muc) could be a better candidate since this cell is of human origin, is nonimmunogenic, has high proliferative capacity, and can be cryopreserved and passaged according to the need.¹⁴ CLEC-muc was isolated as a novel cell type from human umbilical cord and is different from mesenchymal cells. This cell is unique with regard to p63 expression.¹⁴ 

In this study, we cultivated limbal stem cells in a coculture system using CLEC-muc as well as mouse 3T3 cells as feeders and compared relevant characteristics of the expanded limbal cells in order to assess whether CLEC-muc can be used as feeder cells, avoiding zoonotic hazards. 

Materials and Methods

Chemical Reagents and Cell Culture Media

Dulbecco's modified Eagle's medium (DMEM), Ham's F12, keratinocyte serum-free medium (KGM), bovine pituitary extract, human epidermal growth factor (EGF), penicillin, streptomycin, amphotericin B, dispase, and trypsin-EDTA were purchased from Invitrogen-Gibco (Grand Island, NY, USA); Medium-171 was purchased from Cascade Biologics (Portland, OR, USA). Insulin, hydrocortisone, cholera toxin, insulin-like growth factor-1 (IGF-1), and rhodamine B were purchased from Sigma-Aldrich Corp. (St. Louis, MO, USA); fetal bovine serum (FBS) was purchased from Hyclone (Logan, UT, USA); transforming growth factor-β1 (TGF-β1) and platelet-derived growth factor BB (PDGF-BB) were obtained from R&D Systems (Minneapolis, MN, USA) and PeproTech (Rocky Hill, NJ, USA), respectively. Mitomycin C was purchased from Roche Life Science (Indianapolis, IN, USA). Primary antibodies used in this study and their sources are listed in the Table. All secondary antibodies were obtained from Invitrogen (Carlsbad, CA, USA), and mounting medium was purchased from DakoCytomation (Carpinteria, CA, USA); optimal cutting temperature (OCT) freezing compound (Tissue-Tek) and 4′,6-diamidino-2-phenylindole (DAPI) contained in the mounting media (Vectashield) were purchased from Sakura Finetek (Torrance, CA, USA) and Vector Laboratories (Burlingame, CA, USA), respectively. 

Table

 

Primary Antibodies and Sources

 

CLEC-muc Culture

CLEC-muc were isolated from human cord lining tissues obtained from healthy women undergoing delivery after obtaining proper informed consent as described previously.¹³ Briefly, surgical dissection of the umbilical cord was first performed to separate the umbilical cord lining membrane from the enclosed Wharton's jelly and other internal structures. The isolated cord lining membrane was then divided into small squares (0.5 cm²) for cell isolation. For CLEC-muc separation/cultivation, explant tissue samples were placed on cell culture plastic surfaces and submerged in 5 mL Medium-171 (Cascade Biologics). Outgrowing cells were harvested by trypsinization (0.0125% trypsin/0.05% EDTA) in Medium-171 and cryopreserved. The cryopreserved CLEC-muc were thawed and cultured in PTTe-1 medium containing Medium-171 supplemented with 2.5% FBS, 50 μg/mL IGF-1, 50 μg/mL PDGF-BB, 5 μg/mL EGF, 2 μg/mL TGF-β1, and 5 μg/mL insulin. We used cells from different passages (P5–P15). Serum-free media included KGM supplemented with 5 ng/mL human EGF, 5 μg/mL insulin, 0.5 μg/mL hydrocortisone, 50 IU/mL penicillin, 50 μg/mL streptomycin, and 50 ng/mL amphotericin B. Serum containing 1:1 media was prepared by mixing DMEM and Ham's F12 at a ratio of 1:1 supplemented with 5% FBS, 10 ng/mL human EGF, 5 μg/mL insulin, 0.5 μg/mL hydrocortisone, 0.1 nM cholera toxin, 50 IU/mL penicillin, 50 μg/mL streptomycin, and 50 ng/mL amphotericin B, while 3:1 medium contained a 3:1 mixture of DMEM and Ham's F12 supplemented with 10% FBS, 10 ng/mL human EGF, 5 μg/mL insulin, 0.5 μg/mL hydrocortisone, 0.1 nM cholera toxin, 50 IU/mL penicillin, 50 μg/mL streptomycin, and 50 ng/mL amphotericin B. The cells were incubated at 37°C under 5% CO₂ and 95% air, with a medium change every 2 days. Cultures were monitored under an inverted phase-contrast microscope (Axiovert; Carl Zeiss Meditec, Inc., Oberkochen, Germany). The study protocol complied with the Declaration of Helsinki and was approved by the Institutional Review Board of the Singapore National Eye Center and Singapore General Hospital. 

Feeder Cell Preparation

The 3T3 cells and CLEC-muc were maintained in DMEM with 10% FBS and 1:1 medium, respectively. At 70% to 80% confluence, both types of cells were treated with 4 μg/mL mitomycin C for 2 hours at 37°C under 5% CO₂ and 95% air to arrest cell growth. After incubation, the cells were washed with PBS three times for 5 minutes each, then trypsinized using 0.25% trypsin and 0.02% EDTA for 5 minutes, and replated at a density of 2.4 × 10⁴ cells/cm². 

Cultivation of Limbal Epithelial Cells

Human limbal rims were obtained from the Singapore Eye Bank after the central corneal button was used for corneal transplantation. After washing with PBS, the limbal rims were exposed to 1.2 U/mL dispase and incubated at 37°C for 2 hours. The epithelial sheet was removed by gentle scraping and separated into single cells by 0.25% trypsin and 0.02% EDTA for 8 minutes. Limbal cells were plated at 3 to 4 × 10⁴ cells/cm² in cell culture dishes containing mitomycin C-treated 3T3 and CLEC-muc feeder cells. 

Bromodeoxyuridine (BrdU) ELISA Cell Proliferation Assay

We determined proliferative capacity of CLEC-muc by BrdU-ELISA cell proliferation assay (RPN250; Amersham Biosciences, Freiburg, Germany). Cells were seeded in 96-well plates at a density of 2000 cells/well with 100 μL culture medium. Cultured cells were incubated with 10 μM BrdU labeling solution for 20 hours at 37°C, followed by washing with 200 μL PBS containing 10% serum. Then the cells were fixed and incubated with 100 μL monoclonal antibody against BrdU for 2 hours, followed by 100 μL peroxidase substrate per well. The BrdU absorbance in each well was measured directly using a spectrophotometric microplate reader (Tecan spectrophotometer, Grodig, Austria) at a test wavelength of 450 nm. The respective plain basal medium was used as negative control. The optical density (OD) reading for the negative control was subtracted from the sample readings. This gave us a measure of the degree of proliferation of cells, termed the proliferation index. 

Number of Population Doublings

The cells were subcultured by enzymatic disaggregation using 0.25% trypsin/0.02% EDTA for a period of 10 minutes upon reaching 80% confluence. The single-cell suspensions were plated at a density of 3 × 10⁴ cells/cm². The number of population doublings, x, was calculated as follows: x = log₂(N/N₀), where N is the total number of cells harvested at subculture, and N₀ is the number of viable cells seeded. 

Colony Forming Efficiency (CFE)

Limbal epithelial cells were seeded on 6-cm culture plates (BD Biosciences, San Jose, CA, USA) in 1:1 medium at a density of 500 cells per culture dish. After 10 days, the cultures were fixed with 4% paraformaldehyde and stained with 1% rhodamine B. Colony formation was then assessed under a dissecting microscope (Olympus, Miami, FL, USA). Colonies having diameters larger than 2.5 mm were considered “large” while smaller ones were considered “small.” The CFE was determined as follows:  

 

 

Immunocytochemistry

Limbal cells cultivated with CLEC-muc and 3T3 feeders were fixed in cold methanol for 15 minutes at room temperature before blocking and permeabilizing with 2% BSA in PBS with 0.4% Triton X-100. Primary antibodies were incubated overnight at 4°C in blocking buffer. After washing with PBS, cells were further incubated with the appropriate fluorophore-conjugated secondary antibodies (Alexa Fluor; Invitrogen) for 1 hour at room temperature and counterstained with DAPI contained in the mounting medium. Control experiments were performed with blocking buffer without primary antibodies. Immunofluorescence images were observed on a Zeiss Axioplan 2 microscope (Carl Zeiss Meditec, Inc.). All experiments were carried out in triplicate. 

Quantitative Reverse Transcription–Polymerase Chain Reaction (qRT-PCR)

RNAs were obtained from cells using the RNeasy Mini kit (Qiagen, Valencia, CA, USA), and cDNAs were generated using SuperScript III First Strand Kits (Invitrogen). Quantitative PCR was conducted using LightCycler 480 SYBR Green I Master kit (Roche Life Science). All assays were run in duplicate for three or four individual samples. Relative expression levels were calculated using the ΔCt method, normalizing to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) transcription. The PCR products were verified by melting curve analysis and/or 2% agarose gel electrophoresis. The specific primers were designed as follows: ABCG2, For, 5′-TGGCTTAGACTCAAGCACAGC-3′, Rev, 5′-TCGTCCCTGCTTAGACATCC-3′; BMI-1, For, 5′-CTGGTTGCCCATTGACAGC-3′, Rev, 5′-CAGAAAATGAATGCGAGCCA-3′; CK19, For, 5′-CTGCGGGACAAGATTCTTGGT-3′, Rev, 5′-CAGAAAATGAATGCGAGCCA-3′; ΔNp63, For, 5′-CTGGAAAACAATGCCCAGAC-3′, Rev, 5′-GGGTGATGGAGAGAGAGCAT-3′; KRT14, For, 5′-TCCGCACCAAGTATGAGACA-3′, Rev, 5′-GGCTCTCAATCTGCATCTCC-3′; HES1, For, 5′-GCGGACATTCTGGAAATGACA-3′, Rev, 5′-AGCGCAGCCGTCATCTG-3′; GAPDH, For, 5′-GCCAAGGTCATCCATGACAAC-3′, Rev, 5′-GTCCACCACCCTGTTGCTGTA-3′; Tp63, Cat. no. PPH01032E (SABiosciences, Frederick, MD, USA). 

Results

Morphology of CLEC-muc Cultivated in Various Media

We cultivated the cells in three different serum-containing media (Figs. 1A, 1B, 1D). In PTTe-1 medium, umbilical CLEC-muc grew as adherent cells and showed excellent proliferative behavior. The cells mostly appeared with a cuboidal shape and they remained linked to each other as found earlier (Fig. 1A).¹⁴ In 1:1 medium (DMEM:glycerol), cells exhibited similar proliferative capacity while they took on an elongated shape giving rise to a fibroblastic pattern of expansion (Fig. 1B). In 3:1 medium, cells displayed a mixed morphology; some were cuboidal and some were elongated (Fig. 1C). In all serum-containing media, cultures became confluent shortly. We also observed the growth pattern of CLEC-muc in a serum-free medium; however, the cells did not grow well in serum-free medium. Most of the cells exhibited a thin and elongated morphology. These cells underwent limited mitosis (Fig. 1D). 

Figure 1

 

Representative phase-contrast images of CLEC-muc in different culture media (A–D). (A) Morphology of the CLEC-muc cultivated in PTTe-1, (B) 1:1 media, (C) 3:1 media, and (D) serum-free media (SFM) on day 3. Morphology and growth profile of freshly collected limbal epithelial cells (P0) cultivated with 3T3 (E) and CLEC-muc (F) on day 6. Passage 1 limbal cells were further grown with 3T3 (G) and CLEC-muc (H). Experiments were performed in triplicate. Scale bar: 100 μm.

Cell Proliferation and Cell Doubling Assay

We performed a cell proliferation assay to assess the proliferative capacity of the CLEC-muc collected from different donors. Results from BrdU incorporation assay with CLEC-muc from five donors indicated that the proliferative competence of this cell type across the donors was very similar (Fig. 2A). We further carried out cell doubling assay from P3 to P11 using CLEC-muc from four donors and observed no remarkable difference in cell doublings among the donors (12.217 ± 0.320, Figs. 2B, 2C). These results suggest that there is no significant batch-to-batch variation of CLEC-muc between different donors. 

Figure 2

 

Growth profiles of CLEC-muc from different donors. Graph in (A) shows the levels of BrdU incorporation determined for CLEC-muc of P5 collected from five donors. Graph in (B) shows the cell numbers in each passage, and graph in (C) shows the number of cell doublings. Data are represented as mean ± SD of three independent experiments performed in triplicate. n, donor; P, passage.

Cultivation of Primary Human Limbal Epithelial Cells With CLEC-muc and NIH/3T3

CLEC-muc and 3T3 cells were cultured and treated with mitomycin C to make feeder layers as described in Materials and Methods. Primary limbal cells (P0) were seeded on two types of freshly prepared feeder layers. The cultures were continued for 6 days and we observed the cell morphology. Limbal cells grew as colonies with distinct boundaries; some were large and some small, with a regular cuboidal shape for both types of feeder cells (Figs. 1E, 1F). These cells were subcultured after 6 days (P1) and again seeded on freshly prepared feeder layers, which further exhibited a substantial number of colonies that maintained small, compact, and uniform cell morphology (Figs. 1G, 1H). In both feeder layers, we found a similar growth pattern of P0 and P1 limbal cells. No remarkable change in expanded cell shape and size was observed, suggesting that CLEC-muc possess appreciable potential as feeder layer to support ex vivo culture of human limbal cells, which can be compared with that of the 3T3 feeder layer. 

Colony-Forming Efficiency

To compare the capacity of CLEC-muc with 3T3 cells as feeder layer to aid colony formation, CFEs were studied using P1 limbal cells. With both types of feeder layers, a good number of colonies with diameter larger than 2.5 mm were formed, while many were smaller (Figs. 3A–C). We continued the cultures for 7 to 8 days and observed some giant colonies (Figs. 3A, 3B, arrows). CLEC-muc raised from a frozen stock or later-passaged cells (P15) showed a similar outcome (data not shown). These results suggest that CLEC-muc or 3T3 retain comparable potential to promote clonal growth of limbal cells. 

Figure 3

 

Limbal epithelial cells formed colonies in two types of feeder cells. Representative images showing colony formation with CLEC-muc (A) and 3T3 (B) as feeder cells. Graph represents the levels of large- and small-sized colonies formed in two types of feeder cells. Data are represented as mean ± SD of three independent experiments performed in triplicate.

Expression of Putative Stem Cell Markers in Expanded Limbal Cells

In order to understand the cultured cells' retention of their traits, we performed immunocytochemistry (Fig. 4) and qRT-PCR (Fig. 5) to determine several putative limbal stem cell markers. Previous studies have shown that many limbal basal cells that are thought to be limbal stem cells express HES1, ABCG2, ΔNP63, and BMI1.^15–18 These marker genes are essentially downregulated in differentiated corneal cells. Immunofluorescence staining revealed that cells cultivated with CLEC-muc strongly expressed HES1, ABCG2, ΔNP63, and BMI1 (Fig. 4, first through third columns). Similar patterns of expression of these molecules were also observed in cells cultured with 3T3 feeder layer (Fig. 4, fourth through sixth columns). Polymerase chain reaction detection further confirmed the expression of HES1, ABCG2, ΔNP63, TP63, and BMI1 genes in cultured cells grown with both types of feeder layers (Fig. 5). 

Figure 4

 

Expression profile of adult stem cell markers in limbal cells cocultured with CLEC-muc (first through third columns) and 3T3 (fourth through sixth columns) as feeder cells. Limbal cells cultured with CLEC-muc express HES1 (A–C), ABCG2 (G–I), ΔNP63 (M–O), and BMI1 (S–U). A similar expression pattern was observed in cells cultured with 3T3 for HES1 (D–F), ABCG2 (J–L), ΔNP63 (P–R), and BMI1 (V–X). Merged images are shown in columns 3 and 6. DAPI depicts cell nuclei (second, third, fifth, and sixth columns). Experiments were performed in triplicate. Scale bar: 100 μm.

Figure 5

 

Comparison of gene expression of selected genes. ΔCt values of real-time RT-PCR from at least three independent experiments reveal similar levels of gene expression of ABCG2, BMI1, TP63, ΔNP63, HES1, KRT14, and CK19. Higher ΔCt value corresponds to lower gene expression and vice versa. GAPDH was used as endogenous internal control.

Expression of Cytokeratins in Cultivated Limbal Cells

The expression of several cytokeratins at the protein level was examined in cultivated limbal epithelial cells to determine their stemness and differentiation status. We first explored basal cytokeratins CK14, CK15, and CK19 as these cytokeratins have been previously reported to be expressed in basal epidermal and limbal/progenitor cells.^14,16,19,20 Immunostaining against CK14, CK15, and CK19 antibodies revealed a positive expression of these proteins in limbal cells cocultured with CLEC-muc (Figs. 6A–C, 6G–I, 6M–O). Quantitative RT-PCR also showed consistent findings at the RNA level (Fig. 5). This result indicates that cells maintained stemness during expansion with the CLEC-muc feeder layer. A similar observation was also noted with the 3T3 feeder layer (Figs. 6D–F, 6J–L, 6P–R). We then studied the presence of differentiation markers CK3 and CK4 to determine the extent of differentiation occurring in the cultured cells with both type of feeder layers, since this is an obvious phenomenon observed in a coculture system. Although we observed similar expression level of CK3 in the cultivated limbal cells with both CLEC-muc and 3T3 feeder layers (Figs. 6S–X), CK4 expression was notably less in cells cultured with CLEC-muc feeder layer as compared with 3T3 (Figs. 6Y–Z2, 6Z3–Z5, respectively), suggesting that CLEC-muc can be a better substitute of xenologic 3T3 cells for feeder layer. 

Figure 6

 

Expression profile of cytokeratins in limbal cells cocultured with CLEC-muc (first through third columns) and 3T3 (fourth through sixth columns) as feeder cells. Limbal cells cultured with CLEC-muc express CK14 (A–C), CK15 (G–I), CK19 (M–O), and CK3 (S–U). A similar expression pattern was observed in cells cultured with 3T3 for CK14 (D–F), CK15 (J–L), CK19 (P–R), and CK3 (V–X). Merged images are shown in columns 3 and 6. Boxed areas shown in the insets depict clear expression of CK15 in the cytoplasm. However, CK4 expression was relatively weak in cells cultured with CLEC-muc (Y–Z2) as compared to that in cells cultured with 3T3 (Z3–Z5). DAPI depicts cell nuclei (second, third, fifth, and sixth columns). Experiments were performed in triplicate. Scale bar: 100 μm.

Discussion

The use of feeder cells appears to be vital in maintaining larger amounts of stem/progenitor cells for transplantation. Murine-derived 3T3 cells are widely used as feeder layer in a coculture system.²¹ It is perceived that feeder cells probably provide an artificial niche in the culture system that inhibits differentiation while favoring stem cell proliferation. Since 3T3 originates from an animal source, some ethical and safety issues arise, in particular when the cultured epithelial cells are to be used clinically.^22–25 In order to eliminate any risk, researchers have made various attempts to find an effective alternative to 3T3 feeder cells. Human placenta-derived fibroblast-like cells (HPC) originating from chorionic villi of women and primary human amnion epithelial cells (hAEC) have been implicated in culture of mouse embryonic stem cells.⁹ Another study has revealed successful undifferentiated growth of primate ES cells on human amniotic epithelial (HAE) feeder cells.¹⁰ Autologous fibroblasts as feeder cells in a rabbit alkali burn model suppressed peripheral neovascularization but promoted regular epithelial cell proliferation.¹² A recent study has demonstrated the use of human-derived marrow adherent stem cells (MASCS) as feeder cells in making cell sheets for transplantation in the ocular surface.¹¹ 

We aimed to explore the possibility of using the human-derived CLEC-muc as feeder cells to support the expansion of limbal epithelial cells ex vivo. We compared the CFE between CLEC-muc and 3T3 as feeder cells by seeding equal amounts of limbal cells. We observed that the number and shape of the colonies developed in the two systems were comparable, which signifies the importance of CLEC-muc as feeder cells to support limbal stem/progenitor cell growth ex vivo (Fig. 3). We observed a similar growth pattern and proliferative capacity of CLEC-muc from different donors and passage numbers (Fig. 2), demonstrating that CLEC-muc maintain comparable culture characteristics. Our current study findings are consistent with our previous studies^14,26 in which CLEC-muc were shown to proliferate rapidly while maintaining their cell integrity during successive passaging. Although CLEC-muc are not immortalized cells, they can be propagated for at least 40 passages in vitro without a significant loss of proliferative capacity or multipotency.²⁶ 

Although bona fide limbal stem cell markers are yet to be defined, to assess the quality of the cultivated limbal cells in terms of gene expression pattern, we examined the expression of several putative stem cell markers and other genes. It is expected that limbal stem cells cultivated ex vivo should retain their stem cell properties to contribute long-term as stem cells grafted in a living system. In our study, we checked several stem cell markers, such as HES1, ABCG2, ΔNP63, and BMI1. HES1 and ΔNP63 are well-recognized putative stem cell markers for both keratinocytes and corneal epithelial cells. ABCG2 and BMI1 are expressed in hematopoietic stem cells and in many other tissue-specific stem cells,^16,27–29 as well as limbal epithelial side population (SP) cells.³⁰ Fluorescence immunostaining and qRT-PCR data revealed positive expression of all these genes in limbal cells cultivated with both 3T3 and CLEC-muc as feeder cells in a similar fashion. This suggests that CLEC-muc possesses sufficient potential to help proliferation and maintain stemness of cultured limbal cells. It is to be noted that the current study compared only CLEC-muc and 3T3 cells; however, other cells proposed by different studies could be considered. 

Further investigation showed that cytokeratins such as CK14, CK15, and CK19 were also expressed in cells cultivated with CLEC-muc; again this comparable to observations in 3T3 cells. CK14, CK15, and CK19 have been demonstrated to be expressed in different stem/progenitor cells and also in limbal basal cells; thus they are considered proposed limbal stem cell markers.³¹ Interestingly, the conjunctival epithelial marker CK4 was less expressed in limbal cells cultivated with CLEC-muc as compared to limbal epithelial cells grown with 3T3 cells (Figs. 6Y–Z2, 6Z–Z5, respectively); however, corneal epithelial marker CK3 expression was similar in both cases. Based on this finding, we propose that CLEC-muc may provide a more favorable niche in the culture system for generating the corneal cell type rather than the conjunctival type. It is to be noted that CK3 and CK4 are considered differentiation markers; hence, expression of these genes indicates that some level of differentiation occurs during ex vivo culture of limbal cells with both types of feeder cells. 

In summary, we propose that CLEC-muc can substitute for 3T3 fibroblasts as feeder layer for the cultivation of limbal stem cells, which would drive down the potential risk of xenogenic contamination to a large extent. However, in order to use a completely animal-free culture system, culture media should be improved by avoiding the use of animal serum. 

Acknowledgments

Supported by grants from Biomedical Research Council, Singapore (BMRC 07/1/35/19/536 to LPKA). 

Disclosure: L.P.-K. Ang, None; P. Jain, None; T.T. Phan, None; H.M. Reza, None