Molecular Vision 2007; 13:1711-1721 <http://www.molvis.org/molvis/v13/a191/> Received 7 August 2007 | Accepted 13 September 2007 | Published 18 September 2007

Download Reprint

Proteomic analysis of epithelium-denuded human amniotic membrane as a limbal stem cell niche

Hossein Baharvand,¹ Manzar Heidari,² Marzieh Ebrahimi,¹ Tahereh Valadbeigi,¹ Ghasem Hosseini Salekdeh²
 
 

¹Department of Stem Cells, P.O. Box: 19395-4644, Royan Institute, Tehran, Iran; ²Department of Physiology and Proteomics, Agricultural Biotechnology Research Institute of Iran, Karaj, Iran.

Correspondence to: Ghasem Hosseini Salekdeh Department of Physiology and Proteomics, Agricultural Biotechnology Research Institute of Iran (ABRII), P.O.Box 31535-1897, Karaj, Iran; Phone: +98-261-2702893; FAX: +98-261-2704539; email: h_salekdeh@abrii.ac.ir

Abstract

Purpose: A new strategy of treating ocular surface reconstruction is to transplant a bioengineered graft by expanding limbal stem cells (SCs) ex vivo on the amniotic membrane (AM). The reasons for the exceptional success on the AM are not fully understood but are believed to be related to its unique composition. We investigated the proteome of the epithelium-denuded AM to increase our understanding of the mechanisms by which AM may confer its beneficial effects.

Methods: We compared the epithellialy denuded-human AM with matrigel and collagen on the expansion of limbal SCs by evaluating the expression of specific markers. The protein pattern of the epithelium-denuded AM was analyzed using two dimensional electrophoresis (2-DE) coupled with mass spectrometry (MS) identification of proteins.

Results: Epithelial outgrowth of limbal explants on AM expressed more p63 and K19 (SC markers) and less K3 and connexin 43 (corneal differentiation markers) in comparison with other extracellular matrices (ECMs). Moreover, in all groups, the cells expressed ABCG2, K19, K12, p63, and Pax6 as shown by reverse transcription polymerase chain reaction (RT-PCR). Out of about 600 protein spots analyzed on six 2-DE gels, 515 spots could be detected in all replicates. A high average correlation coefficiency (CC) of 0.926 implied good intra-sample reproducibility. Forty major proteins of AM were identified using MALDI TOF/TOF MS of which different isoforms of lumican and osteoglycin were responsible for around 23% of the total proteome on gels.

Conclusions: Our results showed that epithelium-denuded AM provides a superior niche for limbal SC proliferation and phenotype maintenance in vitro and the denuded human AM is a protein enriched ECM. This will prove critical to the future understanding of the biological and therapeutic mechanisms involved in AM transplantation and regeneration. The identification of highly abundant proteins in denuded-AM, such as lumican, osteoglycin/memican, collagen α type IV, and fibrinogen, further explains its unique properties and will assist in the efforts to generate bioengineered and artificial AM constructs.

Introduction

It has been hypothesized that limbal stem cells (SCs), a population of SCs located in the basal epithelium at the corneoscleral limbus, may be maintained and controlled by intrinsic and extrinsic factors in their local microenvironment, the so called SC niche. The SC niche hypothesis, first proposed by Schofield [1], suggests that SCs are maintained in an environment that prevents their differentiation in several tissues and that is the case for the limbal SC niche [2,3]. Environmental factors that may synergistically act to regulate gene expression and to maintain "stemness" include the limbal extracellular matrix (ECM), particularly the basement membrane, cell-matrix interactions, and cell-cell contacts. Although the specific features of such a limbal niche remain elusive, characterization of the microenvironment, including ECM components and growth factors, has already begun to be explored [4].

Using ex vivo cultured limbal SCs transplanted onto the human amniotic membrane (AM) the deficiencies or destructions of the limbal SC population as well as their local microenvironment dysfunction due to a disease or injury can be treated successfully as first reported by Pellegrini et al. [5] and subsequently by others. To achieve this objective, limbal biopsy specimens were cultivated on an intact cellular AM [6-9] or epithelium-denuded AM [10-15]. At the present time, no study has been conducted to compare these cultivation variables thoroughly to determine which ones are vital in achieving effective expansion of limbal epithelial progenitor cells.

In this study, we have endeavored to compare the epithelium-denuded human AM as a limbal SC niche in vitro with commercial ECMs (matrigel and collagen). The expansion of limbal SCs on the ECMs was evaluated by the expression analysis of limbal SC-specific markers. Matrigel is an ECM, which is extracted from the Engelbreth-Holm-Swarm (EHS) mouse sarcoma, and contains a variety of growth factors, laminin (60%), collagen IV (30%), heparan sulfate proteoglycans (8%), entactin or nidogen (1%), and fibronectin [16]. When used in cell culture, matrigel polymerizes to produce biologically active matrix material. Collagen and laminin are major basement membrane components of the limbal and corneal epithelia [17-20]. Therefore, matrigel and collagen were investigated for their suitability at ex vivo expansion of limbal SCs. Indeed, previous studies have already shown that collagen, laminin, and fibronectin can be used to differentiate mouse and human embryonic SC into corneal epithelial cells [21,22].

The ECMs surrounding the cells are as important for growth control as the interactions between soluble, growth-regulating molecules and their cellular receptors. By interacting with matrix molecules, the growth factors may become sequestered from their signaling receptors. This interaction may also cause the activity of growth factors or change their bioactivity. Several reports showed that ECM molecules may exhibit a direct signaling function either via interactions with matrix receptors such as integrins or via signaling through growth factor receptors themselves [23,24]. The identification of key proteins in denuded-AM provides unique insight into properties of AM. Using a 2-DE based proteomics approach, we identified several candidate proteins, which may contribute in the biological function of AM on limbal cell proliferation.

The identification of highly abundant proteins such as lumican, osteoglycin/memican, collagen α type IV, and fibrinogen in denuded-AM goes some way to explain the unique properties of AM and will assist in efforts to generate bioengineered and artificial AM constructs.

Methods

Isolation and culturing of limbal explants

Human tissue was handled according to the Declaration of Helsinki and the approval of the Institutional Review Board. Normal human eye globes were derived from human cadavers aged 55-75 years (n=20). These were obtained from the Central Eye Bank of Iran (Tehran, Iran) and were preserved for less than 24 h post mortem. The tissue was rinsed in Dulbecco's modified Eagle's medium/Ham's F12 nutrient mixture (DMEM/F12; 16141-079; Gibco, Grand Island, NY) containing gentamicin (G1387; Sigma, St Louis, MO) and amphotericin B (A9528; Sigma). After careful removal of excessive sclera, conjunctiva, iris, and corneal endothelium, the remaining tissue was placed in a culture dish and exposed to 10 mg/mL Dispase II (17105-041; Gibco) in DMEM/F12 containing 10% fetal bovine serum (FBS; 21331-20; Gibco) at 37 °C under humidified 5% CO₂ for 10 min. After three 10 min rinses with the medium, each corneoscleral rim was trimmed to obtain limbal tissue cubes of approximately 1x1.5x2.5 mm in size. The whole experiment was conducted using three corneoscleral rims.

The limbal explants were cultured on one of the three different ECMs (AM, matrigel, and/or collagen) in DMEM/Ham's F-12 (1:1) supplemented with 5% FBS, 0.5% dimethyl sulphoxide (DMSO; D2650; Sigma), 2 ng/ml epidermal growth factor (EGF; E9644; Sigma), insulin-transferrin-selenite supplement (ITS; 55261; Sigma), 0.5 μg/ml hydrocortisone (H0888-56; Sigma), 30 ng/ml cholera toxin (subunit A; G115; Biomol, Tokyo, Japan), 50 μg/ml gentamicin, and 1.25 μg/ml amphotericin B. Cultures were incubated in a humidified incubator in 95% air and 5% CO₂ and maintained for two weeks during which the medium was replaced every two days. The extent of each outgrowth was monitored by using an inverted phase contrast microscope (CKX 41, Olympus, Tokyo, Japan).

Preparation of amniotic membrane

Human AMs were obtained in accordance with the tenets of the Declaration of Helsinki and following the approval of the Institutional Review Board and after obtaining informed consent from women undergoing elective caesarean section at term. The tissues were processed as previously reported [25,26]. Briefly, the tissues were washed with PBS containing oflaxocine (0.3%) and gentamicine (50 μg/ml), then fastened on nitrocellulose, and cut into pieces of approximately 3x3 cm². The AM pieces were stored in PBS containing 1.5% DMSO at -70 °C for up to five months. For experiments, AM tissue was thawed, washed with PBS, and incubated with 0.2% EDTA at 37 °C for 15 min to eliminate cellular adhesion, followed by gentle scraping in 5% ammonium chloride to remove the epithelium without breaking the basement membrane. Acellularity of AM was confirmed by phase contrast inverted microscopy. Membrane segments were thawed and thoroughly cleaned by washing three times in 10 ml PBS for 10 min each.

Preparation of matrigel and collagen

Matrigel (E1270; Sigma) was slowly thawed at 4 °C to avoid the formation of gel and then diluted in cold DMEM to the final concentration of 1:30. The collagen solution consisted of collagen type I (2 ml), (3 mg/ml; purecol; Inamed Biomaterials Fremont, CA), which was mixed with 250 μl 10X PBS/cell culture medium (DMEM/Ham's F-12) in the proportion of 8:1 (v/v).

Finally, each well of a six well plate was coated with matrigel and/or collagen solution and incubated for 1 h at 37 °C. Matrigel and/or collagen solutions were removed immediately before use.

Fluorescent immunostaining

The cultured cells were rinsed twice with PBS, fixed with methanol/acetone (3:1) at -20 °C, or 4% paraformaldehyde at room temperature prior to incubation with the respective primary antibody (anti-p63 [MAB4135; 1:50; Chemicon, Temecula, CA], anti-connexin43 [CX43; 1:100; C8093; Sigma], anti-K3/K12 [CBL-218; 1:100; Chemicon], anti-K19 [M0772; Dako, Glostrup, Denmark]) for 60 min at 37 °C in a humid chamber. Cells were then rinsed with PBS and incubated with the secondary antibody, fluorescence isothiocyanate (FITC)-conjugated anti-mouse IgG (F9006; 1:200; Sigma) for 60 min at 37 °C. Cells were rinsed again with PBS and analyzed under a fluorescent microscope (Nikon, Kanagawa, Japan). Negative controls were incubated in mouse IgG rather than the primary antibody. The nuclei were counterstained with 5 μg/ml Hoechst 33342 (B2261; Sigma) and/or propidium iodide (P4170; 10 μg/ml; Sigma). After immunocytochemical staining, positive cells were blindly counted in 10-15 random fields of view and the percentage of positive cells was calculated. All assays were performed in triplicate. The results were given as means±standard deviation (SD). Statistical analysis was performed by using a Student's t-test with significance reported when p<0.05.

Reverse transcription-polymerase chain reaction analysis

Total RNA was isolated from undifferentiated cells and at varying stages of induction as described above using the Nucleospin RNA II kit (Macherey-Nagel, Düren, Germany). Prior to reverse transcription (RT), to eliminate residual DNA, a sample of the isolated RNA was treated with 1 U/μl of RNase free DNaseI (EN0521; Fermentas, Opelstrasse, Germany) per 1 μg of RNA in the presence of 40 U/μl of ribonuclease inhibitor (E00311; Fermentas) in 1X reaction buffer with MgCl₂ for 30 min at 37 °C. To inactivate the DNaseI, 1 μl of 25 mM EDTA was added and incubated at 65 °C for 10 min. Standard reverse-transcription reaction was performed with 2 μg total RNA using Oligo (dT)18 as a primer and RevertAid^TM First Strand cDNA Synthesis Kit (K1622; Fermentas) according to the manufacturer's instructions. For every reaction set, one RNA sample is performed without RevertAid^TM M-MuLV Reverse Transcriptase (RT^- reaction) to provide a negative control in the subsequent PCR. To minimize variation in the reverse transcription reaction, all RNA samples from a single experimental setup were reverse transcribed simultaneously.

Reaction mixtures for PCR included 2 μl cDNA, 1X PCR buffer (AMS^TM, CinnaGen Co., Tehran, Iran), 200 μM dNTPs, 0.5 μM of each antisense and sense primer and 1 U Taq DNA polymerase. The PCR primers and the annealing temperature of the amplified products are shown in Table 1. The PCR reactions were performed on a Mastercycler gradient machine (Eppendorf, Humborg, Germany). Amplification conditions consisted of: initial denaturation, 94 °C for 5 min followed by 35 cycles (25 cycles for GAPDH) of denaturation at 94 °C for 45 s, 55 °C annealing for 45 s, extension at 72 °C for 30 s, and a final polymerization at 72 °C for 10 min. Each PCR was performed under linear conditions with GAPDH used as an internal standard. Products were electrophoresed on a 1.7 % agarose gel. The gels were stained with ethidium bromide (0.5 μg/ml) and photographed on a UV transilluminator (Uvidoc, Cambridge, UK). Gel images were analyzed using the UVI bandmap program (Uvitec, Cambridge, UK).

Protein extraction and two-dimensional gel electrophoresis

Samples collected from six individuals (0.5 g) were pulverized to a fine powder with liquid nitrogen and a mortar and pestle. The powder was resuspended directly in 2.5 ml of Tris pH 8.8 buffered phenol and an equal volume of extraction buffer (0.1 M Tris-HCl pH 8.8, 10 mM EDTA, 0.4% 2-mercaptoethanol, and 0.9 M sucrose). The homogenate was mixed for 30 min at 4 °C. After the 15 min centrifugation at 5,000 g at 4 °C, the phenol phase was removed and proteins were precipitated with 5 volumes of ice-cold 0.1 M ammonium acetate in 100% methanol at -20 °C for 16 h. Then, after 10 min of centrifugation at 5,000 g, the protein pellet was thoroughly washed twice in 20 ml of 0.1 M ammonium acetate in 100% methanol, followed by 2 washes in ice-cold 80% acetone and 10 mM DTT. The pellet was lyophilized and the sample powder was then solubilized in lysis buffer (9.5 M urea, 2% [w/v] CHAPS, 0.8% [w/v] Pharmalyte pH 3-10, 1% [w/v] DTT) and the protein concentration was determined by the Bradford assay (Bio-Rad, Hercules, CA) with BSA as the standard.

Isoelectric focusing (IEF) of approximately 120 μg (for preparative gels 1-1.5 mg) total protein was carried out on immobilized pH gradient 24 cm pH 4-7L and 18 cm pH 6-11L strips on Pharmacia Multiphor II (Amersham Pharmacia Biotech, Uppsala, Sweden). The running condition was as follows: 500 V for 1 h, followed by 1000 V for 1 h, and finally 3500 V for 16 h. The focused strips were equilibrated twice for 15 min in 10 ml equilibration solution. The first equilibration was performed in a solution containing 6M urea, 30% (w/v) glycerol, 2% (w/v) SDS, 1% (w/v) DTT, and 50 mM Tris-HCl buffer, pH 8.8. The second equilibration was performed in a solution modified by the replacement of DTT by 2.5% (w/v) iodoacetamide. Separation in the second dimension was performed by SDS-PAGE in a vertical slab of acrylamide (12% total monomer, with 2.6% crosslinker) using a Dodeca Cell (BioRad). The analytical 2-DE gels were stained with silver nitrate as described by Blum et al. [27] with some modifications. After termination of the second dimension run, the gels were immersed in fixative solution (methanol/distilled water/acetic acid, 40/50/10). The gels were sensitized by exposure to thiosulfate reagent (0.02% sodium thiosulfate), followed by impregnation with silver nitrate reagent (0.2% silver nitrate, 0.02% formaldehyde [37%]) for 30 min and developing in the developing solution (3% sodium carbonate, 0.05% formaldehyde [37%], 0.0005% sodium thiosulfate). The development was stopped by using 5% Glycin for 5 min and gels were rinsed with water several times prior to densitometry. Preparative gels were stained with colloidal coomassie brilliant blue (CBB) G 250 [28].

Image analysis

The silver stained gels were scanned at a resolution of 600 dots per inch on a GS-800 densitometer (Bio-Rad). The scanned gels saved as TIF images for subsequent analysis. Spot quantitation was carried out using the Melanie 3 software (GeneBio, Geneva, Switzerland). The parameters for protein spot detection as follows: number of smooths, 2; Laplacian threshold, 3; partial threshold, 3; saturation, 90; peakness increase, 100; minimum perimeter, 35. After the image treatment, spot detection, protein quantification, and spot pairing were carried out based on Melanie-3 default settings. Then, spot pairs were investigated visually and the scatter plots between gels of each data point were displayed to estimate gel similarity and experimental errors. The molecular masses of proteins on gels were determined by coelectrophoresis of standard protein markers (Amersham Pharmacia Biotech) and pI of the proteins was determined by migration of the protein spots on 18 cm IPG (pH 4-7 and 6-11 linear) strips. Six gels were run and the percentage volume of each spot was estimated and analyzed. Proteins were sorted based on their abundance (percentage volume).

Protein identification and database search

Protein spots were excised from CBB and silver stained gels and analyzed using the Applied Biosystems 4700 Proteomics Analyzer at Protein and Proteomics Centre in National University of (Mass Spectrometry Services, Protein, and Proteomics Centre, Department of Biological Sciences, Singapore) as described in Supplemental Materials and methods. GPS Explorer^TM software Version 3.5 (Applied Biosystems, Foster City, CA) was used to create and search files in the MASCOT search engine (version 2.0; Matrix Science, London, UK) for peptide and protein identification. The S/N ratio in MS/MS mode for peak identification was greater than 40. Combined MS-MS/MS searches were conducted with the selection of following criteria: NCBInr database 12 Feb 2007 (4383774 sequences; 1510495646 residues), all entries, parent ion mass tolerance at 50 ppm, MS/MS mass tolerance of 0.2 Da, carbamidomethylation of cysteine (fixed modification), and methionine oxidation (variable modification). The threshold for positive identification was a molecular weight search (MOWSE) score of >79 (p<0.05).

Results

Effects of the extracellular matrix on ex vivo expanded limbal stem cells

The epithelial sheet grown from the limbal explant on ECMs was usually more or less circular and in two weeks, it assumed a diameter of 1.5 cm. The cells on AM were polygonal and more cuboidal (Figure 1A) and had a nuclear to cytoplasm area of almost 1:1 ratio when observed under the phase contrast microscopy. On matrigel and collagen however, some cells exhibited changes in their size, shape, and nuclear to cytoplasmic ratio with the embossed nucleus (Figure 1B,C).

P63, a transcription factor, and K19, an intermediate filament, have been reported to be expressed in the nuclei and the cytoplasm of limbal basal cells, respectively [29,30]. Using the monoclonal antibody, p63 was detected in the nuclei of outgrowing cells on day 14 (Figure 1D,G, 76.16%±8.47%, out of 1,555 cells) and was reduced by 34.68%±19.27% (out of 2,025 cells) and 47.14%±10.85% (out of 2,569 cells) in matrigel and collagen groups, respectively (p<0.05). K19 was also detected in 28.37%±16.26% of cells (out of 8,597 cells, Figure 1E,H), 29.2%±10.25% of cells (out of 6,540 cells), and 7.17±2.40% of cells (out of 2,073 cells) on AM, matrigel, and collagen, respectively (p<0.01).

Connexin 43 (Cx43), a gap junction transmembrane protein, is involved in direct cell-cell communication [31], and has been implicated in corneal functions and repair [32]. In the outgrowth, a few cells were found to express Cx43 (2.08%±0.89% on AM [out of 2,771 cells, Figure 1F,I], 4.31%±3.01% on matrigel [out of 8463 cells], and 3.89%±2.90% on collagen [out of 1,523 cells]). Positive Cx43 staining was punctuated in appearance and confined to the cell membrane of adjacent cells, compatible with its known function in the formation of gap-junction channels. It was also observed to be in close proximity to the nucleus. The percentage of K3 positive cells was determined as 18.66%±6.02% (out of 3,629 cells), 22.62%±9.21% (out of 2,542 cells), and 53.50%±11.13% (out of 2,621 cells) in the AM group (Figure 1J), matrigel group, and collagen group, respectively (p<0.05). Cytokeratin K3 is generally believed to be a specific marker for mature corneal epithelial cells [33,34].

The phenotypes of cultivated limbal cells at day 14 were evaluated by RT-PCR for ABCG2, an ATP-binding cassette transporter, which has been identified as a specific marker for SCs in limbal epithelium [35,36], for K12 (an intermediate filament), and for a corneal epithelial marker [33,34] such as K19, p63. Moreover, the expression of Pax6, which is necessary for normal clonal growth during corneal development, normal limbal SC activity, correct corneal epithelial cell migration, and proper generation and maintenance of the adult cornea [37,38], was assessed by RT-PCR. The results showed the expression of all of these genes in all groups (Figure 1K) except ABCG2, which was not found to be expressed in the corneal tissue.

Proteome pattern of amniotic membrane

Protein extracts from six individuals were separated by 2-DE. Multiple gels were run to ensure reproducibility of the protein homogenates on the 2-DE gels. The analytical gels were visualized by silver staining (Figure 2A). 2-DE images were analyzed using software and the percentage volume of spots were estimated and compared across the gels. Out of about 600 protein spots analyzed on six 2-DE gels, 515 spots could be detected in all replicates. A high average correlation coefficiency (CC) of 0.926 implied good intra-sample reproducibility (Figure 3).

Protein identification

Using MS, we analyzed 43 highly abundant proteins of which 24 spots showed similar expression across all replications whereas 16 spots represented variable expression patterns across different gels (Figure 2B). MALDI-TOF-TOF MS/MS, on the basis of a combined peptide mass fingerprinting and MS/MS analysis, led to identification of 40 protein spots (Table 2 and Table 3). The most abundant proteins on 2-DE gels were identified as lumican (spots 35, 39, 43, 93, and 11), osteoglycin (spots 24, 26, and 27), collagen VI α-1 (spots 1, 5, 6, 8, 9, 306, 355, 358, 359, and 370), and fibrinogen beta chain (spots 10, 12, 28, 29, 95, 105, and 361; Figure 4). Several of the highly abundant proteins also seemed to be the modified forms of lumican (Figure 2C) and osteoglycin (Figure 2D). Other identified proteins include transglutaminase 2 isoform A (spot 15), heat shock 70 kDa protein 5 (spots 56 and 57), periplakin (spot 106), β-actin variant (spots 107 and 109), nidogen 2 (spot 217), collagen VI α-2 (spot 31), integrin α-6 (spot 51 and 293), TGF-β-induced protein ig-h3 (spot 149), and tubulointerstitial nephritis (spot 160).

Discussion

The effects of extracellular matrix on the ex vivo expanded limbal stem cells

Since the initial successful reconstruction of the corneal surface by transplantation of limbal epithelial cells expanded on human AM in limbus-deficient patients [5], several slightly modified methods for reconstruction of the corneal surface have been reported. The clinical success of these protocols in reconstructing the corneal surface suggests that limbal epithelial SCs may be preserved in the AM-based culture system. However, there has been little evidence of the function of the denuded-human AM as a limbal SC niche in vitro. In this study, we compared the growth of human limbal explants for 14 days on AM, matrigel, and collagen. Our results indicate that the epithelial outgrowth of the limbal explant on AM assumes a phenotype similar to the in vivo limbal epithelium [39] while the cells tend to differentiate more on matrigel and collagen.

AM is the innermost layer of the fetal membranes. It consists of a single layer of epithelial cells that are attached to a thick basement membrane and an avascular stromal matrix. The structural integrity, transparency, and elasticity of the AM make it currently the most widely accepted tissue replacement for ocular surface reconstruction. Moreover, AM has been used for human embryonic SC differentiation into neural cells [40] as well as for supporting chondrocyte proliferation and phenotype maintenance in vitro and the regeneration of osteochondral defect in rabbits [41].

Although not fully elucidated, several different mechanisms of action have been attributed to explain the beneficial effects of the AM, some of which are inferred from the membrane compositions [42-44]. It was reported that the AM's basement membrane (BM) contains collagen types I-VII, Ln-1, Ln-5, fibronectin, and collagen IV and VII [45]. Fibronectin and collagen IV and VII promote epithelial adhesion and migration [46,47]. Laminin (Ln) plays an important role in corneal epithelial cell adhesion [48].

Identification of major amniotic membrane proteins

We analyzed the proteome pattern of AM to identify possible candidate proteins which may contribute to the biological function of AM on proliferation and maintenance of limbal SCs. A comparison between proteins identified in our study with those reported in epithelially intact transplant-ready AM (TRAM) [49] revealed both qualitative and quantitative differences in proteome patterns. The most abundant proteins on 2-DE gels of epithelially intact TRAMs were identified as annexin II, keratin, actin, and BIG-H3 [49] whereas proteins like lumican, osteoglycing/memican, collagen α type IV, and fibrinogen were among the most abundant proteins in our study as well as periplakin, pidogen 2, transglutaminase 2, and tubulointerstitial nephritis. Although only 0.25% of proteins (17 spots) were detected as lumican and osteoglycin, they composed about 23% of total percentage volume of spots on 2-DE gels. Several isoforms of BIG-H3 had been reported with relatively high abundance on 2-DE gels while in our study this protein was detected as a single spot. Further studies will be required to elucidate how the differences and similarities between proteome patterns of epithelium-denuded and intact-AMs are related to their biological functions. This discrepancy can be partly explained by differences in samples preparation (epithelium-denuded AMs versus intact AMs) and the proteomics techniques applied.

Our results showed that two members of the proteoglycan (PG) family, limican and mimican/osetoglycin, are among the most abundant proteins of the AM and contribute to about 11.5% and 11.7% of the total spot volumes on 2-DE gels, respectively. However, PGs had a very small contribution in total proteome matter of intact AM [49], which can be due to different protein extraction methods applied or masking effects of AM epithelial proteins. PGs are macromolecules consisting of a core protein and glycosaminoglycan side chains and are widely distributed in stromal tissues in the human body. PGs of the ECM have been long recognized as the organizers of collagenous networks and also as the molecules that exhibit cell signaling properties, thereby influencing cellular growth, differentiation, and migration [50,51]. PGs belonging to the small leucine-rich proteoglycan (SLRP) family have relatively small molecular sizes with core proteins of approximately 40 kDa and possess 6-10 tandem leucine-rich repeating units between the flanking cystein-rich disulfide-bonded domains at the NH₂- and COOH-termini of the core protein [52-54]. Based on the intron/exon organization of cognate genes, the spacing of the four cysteine residues in NH₂-terminal regions, and the number of leucine-rich repeats in the central domains, they can be divided into three subclasses. Class I PGs comprise the chondroitin/dermatan sulfate chain-containing PGs including decorin and biglycan as well as the glycoprotein asporin. Class II PGs comprise the keratan sulfate chain-containing PGs such as fibromodulin, lumican, keratocan, and osteoadherin as well as the glycoprotein (PRELP). Class III PGs consist of osteoglycin (also known as mimecan) and epiphycan both of which carry a glycosaminoglycan chain and the glycoprotein opticin.

Lumican is the major keratan sulfate proteoglycan of the cornea but is also distributed in interstitial collagenous matrices throughout the body. It is also widely present as a non- or low-sulfated glycoprotein in connective tissues of many other organ systems e.g. skeleton, heart, kidney, and lung [55-58]. Lumican may regulate collagen fibril organization and circumferential growth. Recently, lumican expression has been studied in several pathological conditions including tumor tissues. In corneal injury, the corneal epithelium ectopically and transiently expresses lumican during the early phases of wound healing, suggesting a potential function unrelated to collagen fibrillogenesis, such as the modulation of epithelial cell adhesion or migration, and repair of corneal injury [59].

Osteoglycin, also known as mimecan, might play roles in many biological processes including cellular growth, angiogenesis, and inflammation. It has been suggested that osteoglycin may play roles in adaptive responses, growth control, apoptosis, and aging. It has been shown that growth factors and cytokines modulate mimecan mRNA expression in bovine and human cells [60-63]. It is also upregulated after vascular injury and after low-level laser irradiation of osteoblasts, indicating that the corresponding protein may play a role in wound healing in vascular smooth muscle cells and in osteoblasts [61,64]. In most cancer cell lines and tumors, osteoglycin is either absent or only expressed at low levels [65]. Knockout of osteoglycin in mice leads to an increased collagen fibril diameter in the cornea and skin, illustrating a role in collagen fibrillogenesis [66]. Further studies including the use of a recombinant lumican and osteoglycin protein may clarify the exact role of these proteins in limbal SCs.

Several of the proteins on 2-DE gels were identified as collagen VI and most of them were type α1 (IV) chain (spots 1, 5, 6, 8, 9, 306, 355, 358, 359, and 370). Collagen is the major insoluble fibrous protein in the ECM and in connective tissue. In fact, it is the single most abundant protein in the animal kingdom. Collagen type VI is a unique member within the family of collagenous proteins and is characterized by a rather short triple helix, flanked on each side by a globular domain. To date, six different α (IV) chains have been identified, α1 (IV) to α6 (IV), of which the α1 (IV) and α2 (IV) chains are ubiquitous components of all basement membranes [67,68].

Two of the AM proteins were identified as integrin α6 proteins (spots 51 and 293). Integrins are integral cell-surface proteins composed of one α and one β subunit. The two subunits collaborate to bind ligands, which are ECM proteins or counter-receptors of the Ig superfamily. For example, α6 may combine with β4 in the integrin referred to as TSP180 or with β1 in the integrin VLA-6. Integrins are known to participate in cell adhesion as well as cell-surface mediated signaling. It was reported that the SC or progenitor cell populations of other self-renewing tissues also express integrin molecules. The ECM receptors of the integrin family have been identified as important regulators of epidermal homeostasis, influencing the balance between SC renewal and differentiation [69].

Transglutaminase 2 (TG2) is the most ubiquitously expressed member of the transglutaminase family of proteins. It can modulate several biological events and has been implicated in apoptosis [70], the regulation of cell differentiation [71], and cell survival [72,73]. TG2 is normally secreted into the ECM in relatively low amounts. The increase of TG2 in the matrix causes both direct and indirect effects either through direct protein crosslinking leading to matrix stabilization or indirectly via the activation of matrix-bound TGFβ1 leading to matrix deposition. Matrix-bound enzymes can also act as an independent cell-adhesion protein when bound to fibronectin, preventing cell death by anoikis, wound healing, and maintenance of tissue integrity [74].

A comparison between three ECMs in our study revealed that epithelium-denuded AM provides a superior niche for limbal SC proliferation and phenotype maintenance in vitro. Using the proteomics approach, we showed that the denuded-human AM is a protein enriched ECM. This will prove critical to the future understanding of the biological and therapeutic mechanisms involved in AM transplantation and regeneration. The identification of key proteins in denuded-AM goes some way to explain its unique properties and will assist in the efforts to generate bioengineered and artificial AM constructs. Furthermore, a particularly intriguing subject for future study is to determine how several activities common to the epithelium-denuded and intact-AM are related at the proteome level.

Acknowledgements

We gratefully thank Dr. N.F. Dolatshad for her manuscript proofreading and Ehsan Taghiabadi for his technical assistance. This project was funded by grants from the Royan Institute.

References

1. Schofield R. The stem cell system. Biomed Pharmacother 1983; 37:375-80.

2. Schlotzer-Schrehardt U, Kruse FE. Identification and characterization of limbal stem cells. Exp Eye Res 2005; 81:247-64.

3. Li W, He H, Kuo CL, Gao Y, Kawakita T, Tseng SC. Basement membrane dissolution and reassembly by limbal corneal epithelial cells expanded on amniotic membrane. Invest Ophthalmol Vis Sci 2006; 47:2381-9.

4. Grueterich M, Espana EM, Tseng SC. Ex vivo expansion of limbal epithelial stem cells: amniotic membrane serving as a stem cell niche. Surv Ophthalmol 2003; 48:631-46.

5. Pellegrini G, Traverso CE, Franzi AT, Zingirian M, Cancedda R, De Luca M. Long-term restoration of damaged corneal surfaces with autologous cultivated corneal epithelium. Lancet 1997; 349:990-3.

6. Grueterich M, Tseng SC. Human limbal progenitor cells expanded on intact amniotic membrane ex vivo. Arch Ophthalmol 2002; 120:783-90.

7. Grueterich M, Espana EM, Touhami A, Ti SE, Tseng SC. Phenotypic study of a case with successful transplantation of ex vivo expanded human limbal epithelium for unilateral total limbal stem cell deficiency. Ophthalmology 2002; 109:1547-52.

8. Ti SE, Grueterich M, Espana EM, Touhami A, Anderson DF, Tseng SC. Correlation of long term phenotypic and clinical outcomes following limbal epithelial transplantation cultivated on amniotic membrane in rabbits. Br J Ophthalmol 2004; 88:422-7.

9. Tsai RJ, Li LM, Chen JK. Reconstruction of damaged corneas by transplantation of autologous limbal epithelial cells. N Engl J Med 2000; 343:86-93.

10. Koizumi N, Inatomi T, Quantock AJ, Fullwood NJ, Dota A, Kinoshita S. Amniotic membrane as a substrate for cultivating limbal corneal epithelial cells for autologous transplantation in rabbits. Cornea 2000; 19:65-71.

11. Koizumi N, Inatomi T, Suzuki T, Sotozono C, Kinoshita S. Cultivated corneal epithelial stem cell transplantation in ocular surface disorders. Ophthalmology 2001; 108:1569-74.

12. Schwab IR, Reyes M, Isseroff RR. Successful transplantation of bioengineered tissue replacements in patients with ocular surface disease. Cornea 2000; 19:421-6.

13. Shimazaki J, Aiba M, Goto E, Kato N, Shimmura S, Tsubota K. Transplantation of human limbal epithelium cultivated on amniotic membrane for the treatment of severe ocular surface disorders. Ophthalmology 2002; 109:1285-90.

14. Nakamura T, Inatomi T, Sotozono C, Koizumi N, Kinoshita S. Successful primary culture and autologous transplantation of corneal limbal epithelial cells from minimal biopsy for unilateral severe ocular surface disease. Acta Ophthalmol Scand 2004; 82:468-71.

15. Daya SM, Watson A, Sharpe JR, Giledi O, Rowe A, Martin R, James SE. Outcomes and DNA analysis of ex vivo expanded stem cell allograft for ocular surface reconstruction. Ophthalmology 2005; 112:470-7.

16. Kleinman HK, Luckenbill-Edds L, Cannon FW, Sephel GC. Use of extracellular matrix components for cell culture. Anal Biochem 1987; 166:1-13.

17. Ljubimov AV, Atilano SR, Garner MH, Maguen E, Nesburn AB, Kenney MC. Extracellular matrix and Na+,K+-ATPase in human corneas following cataract surgery: comparison with bullous keratopathy and Fuchs' dystrophy corneas. Cornea 2002; 21:74-80.

18. Ljubimov AV, Burgeson RE, Butkowski RJ, Michael AF, Sun TT, Kenney MC. Human corneal basement membrane heterogeneity: topographical differences in the expression of type IV collagen and laminin isoforms. Lab Invest 1995; 72:461-73.

19. Marshall GE, Konstas AG, Lee WR. Collagens in ocular tissues. Br J Ophthalmol 1993; 77:515-24. Erratum in: Br J Ophthalmol 1994; 78:80.

20. Ihanamaki T, Pelliniemi LJ, Vuorio E. Collagens and collagen-related matrix components in the human and mouse eye. Prog Retin Eye Res 2004; 23:403-34.

21. Homma R, Yoshikawa H, Takeno M, Kurokawa MS, Masuda C, Takada E, Tsubota K, Ueno S, Suzuki N. Induction of epithelial progenitors in vitro from mouse embryonic stem cells and application for reconstruction of damaged cornea in mice. Invest Ophthalmol Vis Sci 2004; 45:4320-6.

22. Ahmad S, Stewart R, Yung S, Kolli S, Armstrong L, Stojkovic M, Figueiredo F, Lako M. Differentiation of human embryonic stem cells into corneal epithelial-like cells by in vitro replication of the corneal epithelial stem cell niche. Stem Cells 2007; 25:1145-55.

23. Ornitz DM. FGFs, heparan sulfate and FGFRs: complex interactions essential for development. Bioessays 2000; 22:108-12.

24. Schonherr E, Hausser HJ. Extracellular matrix and cytokines: a functional unit. Dev Immunol 2000; 7:89-101.

25. Rama P, Bonini S, Lambiase A, Golisano O, Paterna P, De Luca M, Pellegrini G. Autologous fibrin-cultured limbal stem cells permanently restore the corneal surface of patients with total limbal stem cell deficiency. Transplantation 2001; 72:1478-85.

26. Rama P, Giannini R, Bruni A, Gatto C, Tiso R, Ponzin D. Further evaluation of amniotic membrane banking for transplantation in ocular surface diseases. Cell Tissue Bank 2001; 2:155-63.

27. Blum H, Beier H, Gross HJ. Improved silver staining of plant proteins, RNA and DNA in polyacrylamide gels. Electrophoresis 1987; 8:93-9.

28. Neuhoff V, Arold N, Taube D, Ehrhardt W. Improved staining of proteins in polyacrylamide gels including isoelectric focusing gels with clear background at nanogram sensitivity using Coomassie Brilliant Blue G-250 and R-250. Electrophoresis 1988; 9:255-62.

29. Kasper M. Patterns of cytokeratins and vimentin in guinea pig and mouse eye tissue: evidence for regional variations in intermediate filament expression in limbal epithelium. Acta Histochem 1992; 93:319-32.

30. Kasper M, Moll R, Stosiek P, Karsten U. Patterns of cytokeratin and vimentin expression in the human eye. Histochemistry 1988; 89:369-77.

31. Kumar NM, Gilula NB. The gap junction communication channel. Cell 1996; 84:381-8.

32. Ratkay-Traub I, Hopp B, Bor Z, Dux L, Becker DL, Krenacs T. Regeneration of rabbit cornea following excimer laser photorefractive keratectomy: a study on gap junctions, epithelial junctions and epidermal growth factor receptor expression in correlation with cell proliferation. Exp Eye Res 2001; 73:291-302.

33. Barnard Z, Apel AJ, Harkin DG. Phenotypic analyses of limbal epithelial cell cultures derived from donor corneoscleral rims. Clin Experiment Ophthalmol 2001; 29:138-42.

34. Harkin DG, Barnard Z, Gillies P, Ainscough SL, Apel AJ. Analysis of p63 and cytokeratin expression in a cultivated limbal autograft used in the treatment of limbal stem cell deficiency. Br J Ophthalmol 2004; 88:1154-8.

35. de Paiva CS, Chen Z, Corrales RM, Pflugfelder SC, Li DQ. ABCG2 transporter identifies a population of clonogenic human limbal epithelial cells. Stem Cells 2005; 23:63-73.

36. Watanabe K, Nishida K, Yamato M, Umemoto T, Sumide T, Yamamoto K, Maeda N, Watanabe H, Okano T, Tano Y. Human limbal epithelium contains side population cells expressing the ATP-binding cassette transporter ABCG2. FEBS Lett 2004; 565:6-10.

37. Davis J, Duncan MK, Robison WG Jr, Piatigorsky J. Requirement for Pax6 in corneal morphogenesis: a role in adhesion. J Cell Sci 2003; 116:2157-67.

38. Collinson JM, Chanas SA, Hill RE, West JD. Corneal development, limbal stem cell function, and corneal epithelial cell migration in the Pax6(+/-) mouse. Invest Ophthalmol Vis Sci 2004; 45:1101-8.

39. Baharvand H, Ebrahimi M, Javadi MA. Comparison of characteristics of cultured limbal cells on denuded amniotic membrane and fresh conjunctival, limbal and corneal tissues. Dev Growth Differ 2007; 49:241-51.

40. Ueno M, Matsumura M, Watanabe K, Nakamura T, Osakada F, Takahashi M, Kawasaki H, Kinoshita S, Sasai Y. Neural conversion of ES cells by an inductive activity on human amniotic membrane matrix. Proc Natl Acad Sci U S A 2006; 103:9554-9.

41. Jin CZ, Park SR, Choi BH, Lee KY, Kang CK, Min BH. Human amniotic membrane as a delivery matrix for articular cartilage repair. Tissue Eng 2007; 13:693-702.

42. Gomes JA, Romano A, Santos MS, Dua HS. Amniotic membrane use in ophthalmology. Curr Opin Ophthalmol 2005; 16:233-40.

43. Tosi GM, Massaro-Giordano M, Caporossi A, Toti P. Amniotic membrane transplantation in ocular surface disorders. J Cell Physiol 2005; 202:849-51.

44. Dua HS, Gomes JA, King AJ, Maharajan VS. The amniotic membrane in ophthalmology. Surv Ophthalmol 2004; 49:51-77.

45. Fukuda K, Chikama T, Nakamura M, Nishida T. Differential distribution of subchains of the basement membrane components type IV collagen and laminin among the amniotic membrane, cornea, and conjunctiva. Cornea 1999; 18:73-9.

46. Fujikawa LS, Foster CS, Gipson IK, Colvin RB. Basement membrane components in healing rabbit corneal epithelial wounds: immunofluorescence and ultrastructural studies. J Cell Biol 1984; 98:128-38.

47. Murakami J, Nishida T, Otori T. Coordinated appearance of beta 1 integrins and fibronectin during corneal wound healing. J Lab Clin Med 1992; 120:86-93.

48. Kurpakus MA, Daneshvar C, Davenport J, Kim A. Human corneal epithelial cell adhesion to laminins. Curr Eye Res 1999; 19:106-14.

49. Hopkinson A, McIntosh RS, Shanmuganathan V, Tighe PJ, Dua HS. Proteomic analysis of amniotic membrane prepared for human transplantation: characterization of proteins and clinical implications. J Proteome Res 2006; 5:2226-35.

50. Iozzo RV. The biology of the small leucine-rich proteoglycans. Functional network of interactive proteins. J Biol Chem 1999; 274:18843-6.

51. Kresse H, Schonherr E. Proteoglycans of the extracellular matrix and growth control. J Cell Physiol 2001; 189:266-74.

52. Blochberger TC, Vergnes JP, Hempel J, Hassell JR. cDNA to chick lumican (corneal keratan sulfate proteoglycan) reveals homology to the small interstitial proteoglycan gene family and expression in muscle and intestine. J Biol Chem 1992; 267:347-52.

53. Fisher LW, Termine JD, Young MF. Deduced protein sequence of bone small proteoglycan I (biglycan) shows homology with proteoglycan II (decorin) and several nonconnective tissue proteins in a variety of species. J Biol Chem 1989; 264:4571-6.

54. Oldberg A, Antonsson P, Lindblom K, Heinegard D. A collagen-binding 59-kd protein (fibromodulin) is structurally related to the small interstitial proteoglycans PG-S1 and PG-S2 (decorin). EMBO J 1989; 8:2601-4.

55. Ying S, Shiraishi A, Kao CW, Converse RL, Funderburgh JL, Swiergiel J, Roth MR, Conrad GW, Kao WW. Characterization and expression of the mouse lumican gene. J Biol Chem 1997; 272:30306-13.

56. Funderburgh JL, Funderburgh ML, Mann MM, Conrad GW. Arterial lumican. Properties of a corneal-type keratan sulfate proteoglycan from bovine aorta. J Biol Chem 1991; 266:24773-7.

57. Funderburgh JL, Caterson B, Conrad GW. Distribution of proteoglycans antigenically related to corneal keratan sulfate proteoglycan. J Biol Chem 1987; 262:11634-40.

58. Grover J, Chen XN, Korenberg JR, Roughley PJ. The human lumican gene. Organization, chromosomal location, and expression in articular cartilage. J Biol Chem 1995; 270:21942-9.

59. Saika S, Shiraishi A, Liu CY, Funderburgh JL, Kao CW, Converse RL, Kao WW. Role of lumican in the corneal epithelium during wound healing. J Biol Chem 2000; 275:2607-12.

60. Tasheva ES, Funderburgh ML, McReynolds J, Funderburgh JL, Conrad GW. The bovine mimecan gene. Molecular cloning and characterization of two major RNA transcripts generated by alternative use of two splice acceptor sites in the third exon. J Biol Chem 1999; 274:18693-701.

61. Shanahan CM, Cary NR, Osbourn JK, Weissberg PL. Identification of osteoglycin as a component of the vascular matrix. Differential expression by vascular smooth muscle cells during neointima formation and in atherosclerotic plaques. Arterioscler Thromb Vasc Biol 1997; 17:2437-47.

62. Long CJ, Roth MR, Tasheva ES, Funderburgh M, Smit R, Conrad GW, Funderburgh JL. Fibroblast growth factor-2 promotes keratan sulfate proteoglycan expression by keratocytes in vitro. J Biol Chem 2000; 275:13918-23.

63. Tasheva ES, Conrad GW. Interferon-gamma regulation of the human mimecan promoter. Mol Vis 2003; 9:277-87 <http://www.molvis.org/molvis/v9/a39/>.

64. Hamajima S, Hiratsuka K, Kiyama-Kishikawa M, Tagawa T, Kawahara M, Ohta M, Sasahara H, Abiko Y. Effect of low-level laser irradiation on osteoglycin gene expression in osteoblasts. Lasers Med Sci 2003; 18:78-82.

65. Tasheva ES, Maki CG, Conrad AH, Conrad GW. Transcriptional activation of bovine mimecan by p53 through an intronic DNA-binding site. Biochim Biophys Acta 2001; 1517:333-8.

66. Tasheva ES, Koester A, Paulsen AQ, Garrett AS, Boyle DL, Davidson HJ, Song M, Fox N, Conrad GW. Mimecan/osteoglycin-deficient mice have collagen fibril abnormalities. Mol Vis 2002; 8:407-15 <http://www.molvis.org/molvis/v8/a48/>.

67. Hudson BG, Reeders ST, Tryggvason K. Type IV collagen: structure, gene organization, and role in human diseases. Molecular basis of Goodpasture and Alport syndromes and diffuse leiomyomatosis. J Biol Chem 1993; 268:26033-6.

68. Yurchenco PD, Ruben GC. Basement membrane structure in situ: evidence for lateral associations in the type IV collagen network. J Cell Biol 1987; 105:2559-68.

69. Watt FM. Role of integrins in regulating epidermal adhesion, growth and differentiation. EMBO J 2002; 21:3919-26.

70. Tucholski J, Johnson GV. Tissue transglutaminase differentially modulates apoptosis in a stimuli-dependent manner. J Neurochem 2002; 81:780-91.

71. Tucholski J, Lesort M, Johnson GV. Tissue transglutaminase is essential for neurite outgrowth in human neuroblastoma SH-SY5Y cells. Neuroscience 2001; 102:481-91.

72. Boehm JE, Singh U, Combs C, Antonyak MA, Cerione RA. Tissue transglutaminase protects against apoptosis by modifying the tumor suppressor protein p110 Rb. J Biol Chem 2002; 277:20127-30.

73. Antonyak MA, Singh US, Lee DA, Boehm JE, Combs C, Zgola MM, Page RL, Cerione RA. Effects of tissue transglutaminase on retinoic acid-induced cellular differentiation and protection against apoptosis. J Biol Chem 2001; 276:33582-7.

74. Griffin M, Casadio R, Bergamini CM. Transglutaminases: nature's biological glues. Biochem J 2002; 368:377-96.