Targeted Cornea Limbal Stem/Progenitor Cell Transfection in an Organ Culture Model

Abstract

purpose. To optimize a nonviral gene transfection system targeting the corneal limbal stem/progenitor cells.

methods. A plasmid containing LacZ gene coding for β-galactosidase (β-gal) was transfected into human corneal epithelial cells (HCECs) and multilineage progenitor cells (MLPCs) with different transfection reagents, to determine the optimal transfection reagent. In an ex vivo study, the bovine corneal epithelium and limbal stem/progenitor cells were transfected with a microinjection system with a 36-gauge needle that delivered plasmid/transfection reagent (Lipofectamine 2000; Invitrogen, Carlsbad, CA) complexes. The transfected corneoscleral discs were cultured in an air-interface culture system. The expression of β-gal was determined with an X-gal staining assay, and images were acquired with light microscopy and transmission electron microscopy. The expression of cytokeratin K5/14 and K3/K12 in corneal and limbal epithelium was determined by immunohistochemistry.

results. The highest percentages of β-gal expression in HCECs and MLPCs were achieved when the transfection reagent Lipofectamine 2000 was used. Corneal epithelial and limbal basal cells were successfully transfected with the reporter gene by targeted microinjection of plasmid/liposomal complexes. The location of the bovine limbal stem/progenitor cells was confirmed by positive K5/K14 labeling and negative K3/12 labeling.

conclusions. Targeted microinjection of plasmid/liposomal complexes resulted in limbal stem/progenitor cell transfection. This technique has potential for the short-term treatment of corneal diseases.

Despite very promising beginnings, the successful transfection of genes into many tissues in the body has been more difficult than initially anticipated. However, potentially the corneal epithelium should be a good model for gene transfection studies. Its advantages include the accessibility of the epithelium and the relative immune privileged status of the tissue; and, because of its transparency, the transfection product can sometimes be monitored. ¹  

Broadly speaking, there are two classes of gene transfection: transient and stable. Transient therapeutic gene expression would be appropriate for short-term treatment strategies, for example, in the treatment of corneal wound healing and corneal inflammation. Long-term or permanent therapeutic gene expression would be necessary for the permanent replacement of a defective gene in inherited corneal dystrophies. 

The corneal epithelium is a rapidly regenerating stratified squamous epithelium. Cellular turnover rate of the corneal epithelium is approximately 14 days, ^{2 3} and therefore transfection into the corneal epithelial cells would result after approximately 2 weeks in all the transfected cells being replaced by new untransfected cells. For the long term, a more sustainable method must be used. Specifically, the limbal stem/progenitor cells must be transfected. 

The cornea limbus is the junctional zone between the cornea and the sclera, starting at the termination of Bowman’s layer. The maintenance of the corneal epithelial cell mass is achieved by a distinct population of stem cells located in the basal epithelium of the limbus. ^{4 5 6} These cells contain a large amount of melanin, which offers protection against ultraviolet light and reactive oxygen species. ^{4 7} Limbal stem/progenitor cells have a high capacity for self-renewal, which is retained throughout life, and the ability to generate transient amplifying cells. ^{8 9}  

Several studies have been made involving corneal epithelial transfection. ^{10 11 12 13 14 15 16} However, we are aware of only two reporting gene transfection into corneal epithelial stem cells, ^{11 17} both of these involved using viral vectors. One involved transfecting keratolimbal tissue ex vivo and regrafting the explants onto the limbal region. The second involved shaving the superficial cells of the limbal epithelium and inoculating viral vector into the basal cells. Although both techniques were successful, viral vectors have some disadvantages; these include limits to the size of DNA that can be transfected, high immunogenicity, and the possible mutagenesis of cells transfected. ¹⁸ Thus, we feel that the development of a transfection system using nonviral vectors targeted specifically at limbal stem/progenitor cells may offer some advantages. ¹⁹ Since, to date, liposomes have only been used successfully for transient transfection, this technique would be of use only for short-term treatment strategies. 

In this study, we evaluated commercial liposome transfection vectors and used the technique of microinjection into the basal limbal epithelium. Although microinjection has been used for gene transfection in the eye, particularly intracameral and intravitreous, it has never been used for gene transfection in the limbal epithelium. The use of a 36-gauge needle (Nanofil; World Precision Instruments Ltd., Stevenage, UK) allowed precise targeting of the basal cell in the limbus while minimizing damage to the overlying epithelium. In this study, we used a sophisticated organ culture model in which to evaluate our transfection system. 

Methods

Cell Culture

The immortalized human corneal epithelial cell (HCEC) line was obtained from LGC Promochem (Teddington, UK) and cultured in medium with human cornea growth supplement (Epilife; Cascade, Biologics, UK) onto 24-well plates precoated with 0.01 mg/mL bovine serum albumin, 0.01 mg/mL fibronectin, and 0.03 mg/mL collagen I. The nonimmortalized multilineage progenitor cells (MLPCs; BioE, Minneapolis, MN) derived from postpartum human umbilical cord blood were cultured in mesenchymal stem cell growth medium (MSCGM; Cambrex, Nottingham, UK) and used within 3 passages. The cells were cultured in 24-well plates in a standard incubator until 60% to 90% confluence before performing gene transfection. 

Gene Transfection

Plasmid.

pGeneGrip plasmids (Gene Therapy System, San Diego, CA) with hCMV IE promoter/enhancer driving β-galactosidase gene were grown in Escherichia coli and purified (HiSpeed Plasmid Mid Kit; Qiagen, Crawley, UK), as described by the manufacturer. The concentration of plasmid was adjusted to 100 ng/μL in TE buffer. 

Transfection of the Reporter Gene to HCECs and MLPCs.

Fugene 6 (Roche Products, Ltd., Welwyn Garden City, UK), Lipofectamine Reagent (Invitrogen, Paisley, UK), and Lipofectamine 2000 (Invitrogen) were all evaluated for transfection and used according to the manufacturers’ instructions. Lipofectamine 2000 (2 μL) with different concentrations of DNA was evaluated to optimize the transfection reagent-to-DNA ratio. DNA (0.17–1.3 μg/600 μL) was used at a ratio of 1:3 with Lipofectamine 2000 to optimize the DNA concentration. Each experiment was repeated three times with triplicate samples. 

Gene Transfection to Bovine Cornea Epithelium and Limbal Stem/Progenitor Cells with Microinjection.

Normal bovine eyes were obtained from a local abattoir within 2 hours of death, transported to the laboratory at 4°C, and used immediately. Plasmid DNA (1.3 μg) and 4 μL of Lipofectamine 2000 were diluted in 50 μL of minimum essential medium (MEM), respectively. The diluted DNA and Lipofectamine 2000 were then mixed gently and incubated for 20 minutes at room temperature. Microinjection of complexes to the bovine corneal epithelium and limbal stem/progenitor cells was performed with a 36-gauge beveled needle (Nanofil; World Precision Instruments, Ltd.) under dissection microscope. For corneal epithelial cell transfection, the complexes of DNA/Lipofectamine 2000 were injected at a constant rate while the needle was inserted progressively into epithelium until reaching Bowman’s layer. For limbal stem/progenitor cell transfection, after the needle reached a depth of 200 to 300 μm, the needle was pushed forward along the limbus slowly, while the transfection complexes were injected (Fig. 1A) . The injection was performed circumferentially, and approximately 3 μL of transfection complex was injected at each time. The injected corneoscleral buttons were dissected by using standard eye bank techniques and cultured with a 3-D organ culture system ²⁰ for 24 hours (Fig. 1B) . 

Corneal Culture with a 3-D Organ Culture System

The culture was performed as previously described by Zhao et al. ²⁰ Briefly, corneoscleral preparations were mounted on a perfusion chamber and secured with the clamping sleeve, which covered the sclera. The culture chamber was perfused with MEM containing 4% fetal calf serum (FCS) at a flow rate of 4 μL/min. The perfusate from the outflow tube was collected in a reservoir and recirculated. The reservoir was elevated 25 cm above the level of the clamped cornea, to create a physiological pressure of 18 mm Hg inside the artificial anterior chamber. The apical surface was irrigated with MEM containing 4% FCS (Fig. 1C 1D) . The system was maintained at 35°C in a class II laminar flow cabinet. 

Detection of β-Gal Expression

β-Gal expression for either cell lines or the corneoscleral preparations were detected (X-gal Staining Assay Kit; Genlantis, San Diego, CA) according to the manufacturer’s instructions. The samples were washed once with PBS and then fixed with fixing buffer for 15 minutes at room temperature. After they were washed twice with PBS, the samples were incubated in X-gal staining solution (0.2% Triton X-100 was added to aid penetration) overnight at 37°C. After the samples were washed with PBS, digital images were taken. Cells/tissues with blue staining were considered to be positive for β-gal expression. Cell counting was performed in five randomly selected fields in each well at 100× magnification, and the percentage of β-gal-expressing cells obtained. Controls consisted of the samples being treated with the liposomes without the plasmid and then subjected to the X-gal staining protocol. 

Light and Transmission Electron Microscopy

The light microscope images were taken digitally either from wholemounts or from sections. For electron microscopy, the specimens were washed three times in PBS before being passed through a graded ethanol series (50%, 70%, 80%, 90%, 95%, and 100%) and then embedded in epoxy resin. Ultrathin (70 nm) sections were collected on copper grids and examined with a transmission electron microscope (JEM 1010; JEOL, Tokyo, Japan). The sections were not counterstained, so as to make the X-gal product easier to visualize under the microscope. The X-gal product appeared as needlelike crystals, as previously described. ²¹  

Immunolabeling for Cytokeratins in Bovine Corneal and Limbal Epithelium

Fresh specimens were dehydrated in a graded series of ethanol and then embedded in resin (LR White; London Resin Company Ltd., Reading, UK). Polymerization was performed at 50°C overnight. Semithin sections (500 nm) were collected onto coated slides and permeabilized for 30 minutes in 1% Triton X-100. Nonspecific binding sites were blocked by bovine serum albumin in PBS for 30 minutes, after which the sections were incubated with primary antibodies directed against cytokeratin 5 (K5, 1:50), cytokeratin 14 (K14, 1:100), and cytokeratin 3/12 (K3/12, 1:100; Progen Biotechnik, Heidelberg, Germany) overnight at 4°C. After three washes in PBS, the sections were incubated with goat anti-mouse or anti-guinea-pig secondary antibody conjugated with FITC at a concentration 1:200 in PBS for 1 hour. After several washings with PBS and ultrapure water, the samples were mounted (Vectashield with propidium iodide; Vector Laboratories, Peterborough, UK) and imaged under a confocal microscope (SB2-AOBS; Leica). Cells exhibiting fluorescence under a FITC filter were considered to have stained positive for the antibodies. Negative controls consisted of replacement of the primary antibody by a nonspecific IgG of the appropriate species, as previously described. ²²  

Statistical Analysis

All numerical results are expressed as the mean ± SEM. Statistical analysis was performed with an unpaired Student’s t-test. Statistical significance was defined as P < 0.05. 

Results

Optimization of Transfection Reagents and Conditions in Cell Lines

This section describes the results of HCEC and MPLC transfection. After transfection for 24 hours, the β-gal expression was identified with the X-gal staining assay kit. Our results demonstrated that the β-gal expression rates in HCECs and MLPCs varied with different DNA/vector ratios and total DNA concentration. When 0.67 μg of DNA was used at a DNA-to-transfection reagent (microliters) ratio of 1:3, β-gal-expressing HCECs were 9.18% ± 1.11% with Fugene 6, 5.76% ± 0.49% with Lipofectamine, and 32.08% ± 1.46% with Lipofectamine 2000. In contrast, the β-gal-expressing MLPCs were 18.77% ± 0.89%, 4.37% ± 0.34%, and 77.03% ± 2.56% with Fugene 6, Lipofectamine, and Lipofectamine 2000, respectively. With Lipofectamine 2000 use in HCECs and MLPCs, the highest efficiency was achieved at a DNA-to-Lipofectamine 2000 ratio of 1:3 (Fig. 2A) . The combination of a DNA concentration of 1.3 μg/600 μL and a DNA-to-Lipofectamine 2000 ratio of 1:3 produced the highest expression in both HCECs (38.10% ± 2.29%) and MLPCs (86.30% ± 2.81%) (Figs. 2B 2C 2D) . Extension of the culture time to 48 hours after transfection did not significantly increase the amount of expression. 

Gene Transfection with Lipofectamine 2000 in the Organ Culture Model

For the cell lines used in this investigation, the most effective transfection was achieved by using Lipofectamine 2000, and so Lipofectamine 2000 was chosen for corneal epithelium and limbal stem/progenitor cell transfection in the organ culture model. After microinjection, the results demonstrated that the reporter gene was expressed in both corneal and limbal epithelium. In the corneal epithelial transfection (n = 6), β-gal was expressed sporadically in all cell layers, including corneal basal, wing and superficial cells (Figs. 3A 3B) . After limbal region injections (n = 6), the results showed that β-gal staining was present in the limbal area (Fig. 3C) . Wholemounts showed that β-gal product localized in pigmented cells within the limbal region (Figs. 3D 3E) . Light micrographs of sections showed that β-gal staining was sporadically present in the basal cell layer in the limbus (Fig. 3F) , which is the location of the limbal stem/progenitor cells. Transmission electron microscopy confirmed that the crystalline β-gal product was present in the cytoplasm of some of the basal limbal epithelial cells (Figs. 4A 4B) . These cells were easily identifiable as limbal stem/progenitor cells by their location and ultrastructure. These were small, poorly differentiated cells containing numerous melanin granules. The cells had a high nucleus-to-cytoplasm ratio. There was no underlying Bowman’s membrane; instead, there was a thin basement membrane that was deeply invaginated in places. 

Immunolabeling for Cytokeratins in Bovine Corneal and Limbal Epithelium

The immunohistochemical results demonstrated that K5 and K14 were expressed at high levels in the limbal basal layer and were not detected in the limbal superficial, or suprabasal, layer (Figs. 5B 5D) . They were not present in the central corneal epithelium (Figs. 5A 5C)and only at very low levels in the peripheral basal epithelium In contrast, K3/12 was expressed in all the corneal epithelium, the limbal superficial cells, and suprabasal cells, but was absent in the limbal basal cells (Figs. 5E 5F) . 

Discussion

The results of this study show that the limbal stem/progenitor cells can be successfully transfected with a microinjection system to deliver plasmid/liposomal complexes into the basal layer of the limbal epithelial cells. 

Previously a variety of gene transfection methods have been used on the corneal epithelium. There are two examples that report successful gene transfection into corneal epithelial progenitor cells. ^{11 17} These involve transducing keratolimbal tissue ex vivo and regrafting the explants onto the limbal region or shaving the superficial cells of the limbal epithelium and inoculating the viral vector into the basal cells. However, while of great potential, ex vivo methods are necessarily complex, whereas mechanical shaving of the epithelium to access the basal cells increases the risk of infection. In addition, viral vectors have some disadvantages, including an acute immune response, potential mutagenesis due to random insertion of viral genome, and limits in the size of DNA transfected. ^{23 24 25} Nonviral methods require a plasmid vector for the delivery of the genes of interest into the cells. Several methods have been developed, including mechanical, electrical, and chemical approaches. Some investigators have delivered DNA into rabbit and mouse corneal epithelial cells in vivo with gene gun technology. ^{12 13 15 16} In this method, plasmid DNA is mixed with gold microparticles and transferred into the target tissue with a gene gun. This approach successfully delivers the marker gene into the target cells without ocular irritation. However, poor efficiency and low target specificity were the main limitations of this method, and limbal stem/progenitor cells were not specifically examined. We experimented with a gene gun, but even with maximum pressure were unable to achieve penetration of the vector into the limbal stem/progenitor cells (Zhao B, Fullwood NJ, unpublished results, 2007). Other workers have reported successful gene transfection by intrastromal corneal injection using naked DNA. ²⁶ We also experimented with injecting naked DNA but were unable to observe any gene expression in the basal limbal epithelium. In another report, the uptake of naked DNA into corneal epithelium was achieved by using adjunctive electroporation. ¹⁰  

Liposomes are efficient and relatively simple to use, with low toxicity. Positively charged cationic lipid binds to the negatively charged DNA and forms a lipid-DNA complex. Cellular uptake of the cationic-DNA complex involves nonspecific interaction with the cell surface, followed by endocytosis into endocytic vesicles, trafficking and release of DNA from the endosomal compartment, nuclei uptake the DNA and gene expression. ^{27 28} However, their main disadvantage is that they have not been proven for long-term gene expression. ¹⁸  

In our experiments, Lipofectamine 2000 was the most efficient in the cell lines used in the study. Microinjection allowed the complexes of plasmid/liposome to infiltrate around the basal cells, as the tight junctions of the corneal epithelium are located only between the most apical cells, and such a barrier does not exist in the deeper layers of the corneal epithelium. ^{29 30}  

To confirm the identify and location of the bovine epithelial stem cells we used keratin markers. These included the high levels of labeling for K5/14 and the absence of labeling for K3/12 in the limbal stem/progenitor cells. Our immunohistochemical results demonstrate that the bovine limbal basal layer of cells was positive for K5/14 and negative for K3/12, this is in agreement with previous work. ³¹ Other studies have independently confined the location of bovine ^{20 32} and human limbal stem/progenitor cells ³³ in the cornea using recently developed microspectroscopic technology. ³⁴  

Our study clearly demonstrates that precise stem cell transfection can be accomplished with the method of microinjection, although liposomes would not be suitable for the treatment of conditions requiring long-term gene expression. As far as we are aware, this is the first study to confirm by electron microscopy the expression of the transfected gene product within the limbal stem/progenitor cells. 

In the past few years, changes in adult stem cell activity have become linked with an increasing number of diseases, ^{35 36} so that interest in modifying stem cell behavior through transfection has intensified. We believe that microinjection of liposomes has potential for clinical use. We realize that many obstacles must be overcome before this method can be used for clinical applications, but our results show that targeted transfection into limbal stem/progenitor cells using a nonviral vector is possible. 

 Supported by The National Centre for the Replacement, Refinement and Reduction of Animals in Research, United Kingdom.

 Submitted for publication September 27, 2007; revised February 21 and April 7, 2008; accepted June 19, 2008.

 Disclosure: B. Zhao, None; S.L. Allinson, None; A. Ma, None; A.J. Bentley, None; F.L. Martin, None; N.J. Fullwood, 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.

 

(A) Gene transfection of limbal stem/progenitor cells by microinjection. (B) Corneoscleral preparation for limbal stem/progenitor cell transfection at 24 hours of culture. The cornea and limbus are in good condition, and the cornea is transparent. (C) Corneal perfusion chamber. The corneoscleral preparation was secured with the clamping sleeve. The posterior culture chamber is perfused, and the apical surface of the cornea irrigated. (D) Schematic diagram of the corneal perfusion chamber.

Figure 2.

 

The β-gal expression after 24 hours of transfection with Lipofectamine 2000 (Invitrogen, Paisley, UK) under different transfection conditions on HCECs and MLPCs. (A) The β-gal expression rates attained with 2 μL of transfection reagent at different ratios of DNA. Data are expressed as the mean ± SEM. *The β-gal expression rate with a DNA to transfection reagent of 1: 3 was significantly better compared with the other ratios (P < 0.05). (B) The β-gal expression rates at different concentrations of DNA with a DNA to transfection reagent ratio of 1:3. The data are expressed as the mean ± SEM. *P < 0.05, indicating that the β-gal expression rate was significantly better when 1.3 μg/600 μL DNA was used. (C) The expression of β-gal products in HCECs with 1.3 μg/600 μL DNA with a DNA-to-Lipofectamine 2000 ratio of 1:3. (D) The expression of β-gal products in MLPCs using 1.3 μg/600 μL DNA with a DNA-to-Lipofectamine 2000 ratio of 1:3. Scale bars: 100 μm.

Figure 3.

 

(A) Light micrograph of a cornea wholemount showing that the blue β-gal product was present in the injection sites along the corneal epithelium. (B) Light micrograph of a cross-section of the corneal epithelium showing that the blue β-gal product was present in the basal, wing, and superficial epithelium. (C) Light micrograph of wholemount showing that the blue β-gal product was present in the injection sites within the limbal region. (D, E) High-magnification light micrographs of wholemounts showing the colocation of the blue β-gal product with melanin deposits within the limbus. (F) Cross-section light micrograph showing the blue β-gal product located within pigmented basal cells in the limbus. Scale bars: (A) 300 μm; (B, D) 100 μm; (C) 500 μm; (E) 50 μm; (F) 200 μm.

Figure 4.

 

Transmission electron micrographs showing β-gal product (crystals) inside bovine limbal basal cells. (A) High magnification of corneal limbal stem/progenitor cells showing a high density of β-gal product (black arrow); melanin granules (white arrow) are also present with the cell. (B) High magnification showing β-gal product (black arrows) and melanin granules (white arrow) clearly visible within the same cell. SC, stem cell; BM, basement membrane. Scale bars, 1 μm.

Figure 5.

 

Immunohistochemical labeling (green) of the bovine epithelium of the central cornea (A, C, E) and the limbus (B, D, F) for K5, K14, and K3/12. Immunostaining for (A, B) K5; (C, D) K14; and (E, F) K3/12. Propidium iodide staining (red) shows the location of the nuclei. Scale bar, 100 μm.

The authors thank John Dent for his assistance with the confocal microscopy.