IOVS News in Physiological Sciences
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

(Investigative Ophthalmology and Visual Science. 2007;48:144-156.)
© 2007 by The Association for Research in Vision and Ophthalmology, Inc.
DOI:  10.1167/iovs.06-0346
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Figueira, E. C.
Right arrow Articles by Wakefield, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Figueira, E. C.
Right arrow Articles by Wakefield, D.

The Phenotype of Limbal Epithelial Stem Cells

Edwin C. Figueira,1 Nick Di Girolamo,1 Minas T. Coroneo,1,2 and Denis Wakefield1

1From the Inflammatory Diseases research unit, University of New South Wales, Australia; and the 2Department of Ophthalmology, Prince of Wales Hospital, New South Wales, Australia.


   Abstract
Top
 Abstract
Materials and Methods
 Results
 Discussion
 Conclusion
 References
 
PURPOSE. The purpose of this study was to identify phenotypic markers of human limbal stem cells in fetal and adult corneas.

METHODS. RNA from microscopically dissected superficial limbal and central fetal (18 weeks) corneas was amplified and used to generate P32-labeled, reverse-transcribed antisense RNA that was linearly amplified and hybridized to a focused stem cell cDNA microarray. Differential gene expression of fetal limbus was compared with the expression of central cornea. Microarray differential expression experiments were performed on P63-expressing primary cultured limbal epithelial cells (passage 1; Pa1) and primary cells passaged 5 times (Pa5). Semiquantitative RT-PCR assay and immunohistochemistry were performed on fetal and adult corneas and cultured primary limbal epithelial cells, to confirm the results of the microarray experiments. Slow-cycling (pulsed bromodeoxyuridine label-retaining) limbal epithelium in corneal organ culture was studied for the expression of four selected upregulated limbal genes.

RESULTS. Of the 266 genes tested, 33 were differentially overexpressed (more than twofold) in the fetal limbus (compared with central cornea) and primary cultured limbal epithelium compared with primary cells after 5 passages. Cytokeratin 15 (CK15) and cytokeratin 14 (CK14) are expressed in limbal basal epithelium and P-cadherin (CDH3) and Wnt-4 expression was restricted to basal and immediate parabasal limbal epithelium of both the adult and fetal corneas). Bromodeoxyuridine label retaining epithelium in corneal organ culture (slow-cycling cells) expressed the four selected limbal upregulated genes.

CONCLUSIONS. For the first time, a focused stem cell pathway microarray analysis has been performed on fetal cornea and cultured limbal explant epithelium. CK15, CK14, CDH3, and Wnt-4 are expressed in the basal limbal epithelial cells.

The corneal epithelium provides an ideal model system for the study of adult stem cells due to the readily visualized specific spatial arrangement of the limbal basal stem cells (located in the limbal palisades), the transient amplifying cells, and terminally differentiated cells.1 2 3 4 5 Identification of undifferentiated stem cells has relied primarily on the presence or absence of specific phenotypic markers. At present, there is no specific phenotypic marker for corneal limbal stem cells. There is an ever-increasing demand for allograft corneal and limbal cells for transplantation and the identification of a phenotype profile of limbal stem cells may facilitate their genetic manipulation and therapeutic use. The use of such cells for transplantation would avoid the risks associated with allograft rejection and transmissible disease.

Investigators have proposed potential markers for limbal epithelial stem cells; however, their studies have been restricted to the demonstration of markers in the basal limbal epithelial layer. Currently, the limbal stem cell profile is defined as p63, ABCG2, integrin-{alpha}9–positive and nestin, E-cadherin, connexin 43, involucrin, CK3, and CK12-negative.6 Several crucial questions remain to be answered: (1) Does the limbal epithelial stem cell have a profile similar to other stem cell populations (adult/embryonic)? (2) Is it possible to use innovative differential expression analysis techniques, like microarray hybridization, to identify gene expression specific to the limbal region of the cornea? and (3) Is the limbal epithelium in adult cornea similar to more primitive undifferentiated corneal epithelium, such as the fetal corneal.

In this study, we compared gene expression in the immature fetal corneal epithelium with that in primary in vitro cultured limbal epithelia. We reasoned that the fetal cornea in the early stage of embryology would express genes similar to those in the adult limbal epithelium. The human fetal corneal epithelium develops through three different stages: the first 17 weeks of gestation, in which cell proliferation and apoptosis are minimal; the second stage (17–28 weeks), when there is high proliferative activity in all parts of the cornea (stage of ongoing differentiation of the corneal epithelium) with minimal apoptotic cell death; and the final stage (beyond 28 weeks’ gestation), when apoptosis first appears in the epithelium.7 These observations form the basis for studying the limbal and the central corneal regions at the second stage of development, due to the high proliferative rate of these cells and combined relatively primitive nature of the stem cells (relative to the third stage of development when the proliferative activity is restricted to the basal layer). We used the slow-cycling concept of stem cells to colocalize expression of the genes of interest in limbal stem cells.


   Materials and Methods
Top
 Abstract
 Materials and Methods
Results
 Discussion
 Conclusion
 References
 
Tissue Preparation
Normal human eyes were obtained from the Lions NSW Eye Bank, Sydney, Australia. Premortem consent for tissue donation for research was obtained from all donors. Fetal eyes (18 weeks gestational age) were obtained postmortem from the Diabetes Transplant Unit, Prince of Wales Hospital, Sydney, under ethics approval by the Human Research Ethics Committee of the University of New South Wales. Tissue acquisition and experimentation were performed according to the guidelines of the Declaration of Helsinki. Whole eyes were washed in sterile phosphate-buffered saline (PBS; Invitrogen-Gibco, Grand Island, NY) in a laminar flow sterile hood. Fetal corneas were dissected into limbal (limbal fetal cornea; LFC) and central (central fetal cornea; CFC) corneal regions and used for RNA extraction. Adult corneas, acquired more than 24 hours postmortem, were dissected into limbal and central corneal regions. The limbal regions were used for limbal explant epithelial cultures. Three whole corneas were subjected to organ culture in serum-free modified Eagle’s medium TMEM (Thermotrace, Noble Park, Australia) and bromodeoxyuridine (BrdU) label-retaining studies. Whole corneas were formalin fixed and paraffin embedded (n = 5 adult corneas, n = 4 fetal corneas) or frozen in OCT (n = 4 adult corneas, n = 3 fetal corneas) for immunohistochemistry.

RNA Extraction and T7 RNA Amplification
Fetal corneas were dissected with a number 21 scalpel blade (Swann-Morton, Sheffield, UK) under sterile conditions into peripheral limbal rims (LFC; 5 mg) and central corneal buttons (CFC; 5 mg). Total RNA was extracted from disrupted and homogenized LFC and CFC tissues, using a commercial kit (RNeasy microkit; Qiagen, Doncaster, VC, Australia). Total RNA was quantified by measuring the 260/280 nm absorbance ratio in a spectrophotometer (GeneQuant II Pro; Pharmacia Biotech, Uppsala, Sweden).

Because of the scarcity of fetal corneal tissue and the small quantities of RNA that could be extracted (0.05 µg RNA / 5 mg wet tissue weight), Linear T7 RNA amplification (RiboAmp kit; Arcturus, Mountain View, CA) was performed on 0.05 µg RNA (extracted from LFC and CFC dissected regions, n = 2).8 9 Briefly, total RNA and Poly-deoxyinosinic-deoxycytidylic acid (200 ng; Sigma-Aldrich, St. Louis, MO) was incubated with hybrid primers containing an oligo-dT/T7 RNA polymerase binding site. Reverse-transcribed RNA was converted into double-stranded cDNA during a second-strand synthesis reaction and purified (MiraCol columns; Arcturus). In vitro transcription into antisense (aRNA) was performed using T7 RNA polymerase, and amplified (aRNA) was eluted from the columns. From 0.05 µg initial total RNA, 50 µg of aRNA was generated after two rounds of amplification. The quality and quantity of aRNA was analyzed using a spectrophotometer (260/280nm-absorbance ratio: 1.8–2.0). aRNA samples were then analyzed by agarose gel electrophoresis. One microgram of eluted aRNA was added to 6 µL of 2x gel loading buffer, incubated at 65°C for 5 minutes and cooled. RNA MOPS (3-[N-morpholino] propanesulfonic acid) running buffer was used for running a 1% agarose gel and filling the electrophoresis unit. The gel was electrophoresed at 5 V/cm, then stained with ethidium bromide for 5 minutes and visualized with a UV transilluminator (Quantity One software for imaging; Bio-Rad, Hercules, CA, software). The final aRNA measured 200 to 800 nucleotides in size, when matched with the RNA ladder size electrophoresed on the same agarose gel.

Limbal Epithelial Cells: Isolation and Culture
Sterile PBS washed 2-mm2 adult corneal limbal rims were dissected from the paracentral cornea and placed epithelial side up in six-well culture plates (Nalge Nunc Int., Herford, UK).10 11 12 The limbal rims were immersed in serum-free modified TMEM (Thermotrace) supplemented with 292 µg/mL L-glutamine, 10,000 U/mL penicillin G, 10,000 U/mL streptomycin sulfate, 25 µg/mL amphotericin B, and ITS mixture (insulin, 10 µg/mL; transferrin, 5.5 µg/mL; and selenium, 5 ng/mL; Sigma-Aldrich). Cultured cells were observed under an inverted phase-contrast microscope for epithelial outgrowth until the confluence reached just short of the well edge (Figs. 1a 1b 1c) . Cells were confluent after 2 weeks in culture. RNA was extracted from primary cultured cells that were detached with 0.05% trypsin-EDTA (Invitrogen-Gibco) at 37°C. Trypsinized cells were split three ways and used for (1) RNA extraction, (2) acquiring culture media–revived (for 6 hours) and centrifuged cell pellets that were formalin fixed (for 24 hours) and paraffin embedded (consecutive trypsinized and cultured cell populations labeled passage Pa1 to Pa5), and (3) resuspension into consecutive cultures at 50 x 106/mL. Primary limbal explant epithelial cells were cultured to near confluence in eight-chamber slides and were studied for nuclear p63 immunostaining (mouse monoclonal IgG2a anti-{Delta}N p63, 1:50 dilution; Santa Cruz Biotechnology, Santa Cruz, CA) and for cytoplasmic CK3/CK12 (AE5 antibody; Dako, Botany, NSW, Australia). Positive staining for {Delta}N p63 and negative for CK3 confirmed their origin from the basal limbal epithelium (Fig. 1d ; P63-positive Pa1 cells).10 Trypsin detached cells (Pa5) were characterized for the differentiation corneal epithelial marker CK3 (antibody AE5 identifies CK3 expression; Fig. 1e ).


Figure 1
View larger version (118K):
[in this window]
[in a new window]

  		FIGURE 1. Characterization of limbal explant cells: morphologically (a–c)—appearance at days 3 (a), 7 (b), and 14 (c)—and phenotypically (d, e). (d) Strong {Delta}N P63 expression, indicating cells of the limbal epithelium nature (left inset: higher power magnification; right inset: IgG isotype control). (e) AE5 antibody detecting the presence of CK3 in the limbal-derived cell line at passage 5 (left inset: higher power magnification; right inset: IgG isotype control). Bars, 100 µm. Inset magnification, x40.

 
Stem Cell–Related Gene Expression in the Fetal Limbus
Five micrograms of aRNA from the LFC and CFC regions were reverse transcribed and amplified using MMLV reverse transcriptase and the ampolabeling LPR (linear polymerase reaction) radioactive detection method protocol of the GE superarray (see: www. superarray.net for protocol; Superarray Bioscience Corp., Frederick, MD). aRNA molecules annealed with random primers at 70°C for 3 minutes were reverse transcribed to cDNA at 37°C for 25 minutes. cDNA was linearly amplified in a DNA thermal cycler (Perkin Elmer Cetus; , Branchburg, NJ) with gene-specific primers and P32 dCTP. The LPR conditions were 85°C for 5 minutes, 30 cycles (85°C for 1 minute, 50°C for 1 minute, 72°C for 1minute), and 72°C for 5 minutes. The linearly amplified DNA was hybridized onto a focused human stem cell cDNA microarray (GE Superarray; catalog HS 601.2; Superarray, Inc.) overnight at 60°C in a hybridization oven. The microarray used was a commercial microarray designed to look at stem cell pathway gene expression (Catalog HS601.2 human stem cell cDNA microarray; Superarray Bioscience, Inc.). Microarray membranes were placed in hybridization cylinders and washed twice for 15 minutes each with 2 x SSC (standard saline citrate in 1% sodium dodecyl sulfate; SDS), and twice for 15 minutes each with 0.1 x SSC in 1% SDS in a hybridization oven at 60°C. A phosphorescence imaging screen (Phosphorimager; Bio-Rad) was then exposed to the hybridized P32 emission signal of the microarray membranes for 12 hours. A reader (Bio-Rad) and software (Quantity one; ver. 4.5.2; Bio-Rad) was used to acquire the image, which was saved as a 16-bit inverted gray-scale, high-resolution electronic TIFF file (Photoshop software, ver. 5; Adobe Systems, Mountain View CA). The intensity of the signal was estimated and converted into numerical values using the Eisen software program Scanlyze (ver. 2.0; online http://rana.lbl.gov/EisenSoftware.html). The numerical values were then exported into a datasheet (Excel; Microsoft, Redmond, WA) datasheet and later imported into the gene array software program (Superarray Bioscience, Inc. which was used for data analysis of the signals. In brief, the brightness intensity of the individual signals was normalized to the housekeeping gene (GAPDH) after individual signal subtraction from the background signal. The microarray experiments were performed in duplicate.

Stem Cell Pathway Microarray Expression Profiling of Limbal Explant Epithelial Cells
Total RNA was extracted (RNeasy microkit protocol) from trypsin-detached cadaveric human limbal explant epithelial cells in primary culture (Pa0) and from cells trypsin detached after 5 passages (Pa5). RNA (5 µg) was reverse transcribed using P32 labeled dCTP (GE Healthcare, Arlington Heights, IL), linearly amplified, and hybridized onto the stem cell pathway cDNA microarray, as described earlier. The hybridization signals were viewed and analyzed for differential gene expression between Pa0 and Pa5 cells, as described previously. The microarray experiments were performed in duplicate.

Approach to Analysis of Microarray Data
Stem cell cDNA microarray hybridization with 32P labeled cDNA probes synthesized from independent aRNA and RNA preparations obtained from the fetal cadaveric corneas (LFC and CFC) and cultured adult limbal explant epithelial cells (Pa0 and Pa 5), respectively, were compared for differences. A simple differential analysis of gene expression (x-fold change) between samples to be compared allowed for a metric for comparing mRNA expression level between two distinct experimental conditions or two distinct anatomic regions.13

Gene expression data obtained from spotted cDNA arrays are often reported as an expression ratio (x-fold change), which is simply the normalized ratio of the hybridized signal of the test sample to that of a reference sample. The intensities of expression signals in both the test and reference samples are first background corrected (subtracted from background intensity) before the expression, normalized to GAPDH, is calculated. Genes that are upregulated twofold have an expression ratio of 2, and genes that are downregulated twofold have an expression of 0.5. All the genes on an array can have such a calculated expression ratio. A gene may be unexpressed in the reference sample but expressed in the test sample. Calculation of an expression ratio in such a case would need a calculation division by 0. These genes are usually of great interest, as in diseased versus normal gene expression or regional variation of gene expression. This lowest allowed signal can be set at an arbitrary value or to the level of noise—for instance, set at a specified number of standard deviations from the background signal.13

aRNA preparations from microscopically dissected superficial limbal versus superficial central fetal cornea were compared for in vivo differential gene expression between two individual data sets (LFC and CFC). The microarray experiments were performed in duplicate. Spearman’s nonparametric rank correlation test was used to confirm the correlation of ranking of the differential gene expression between the replicates.

Independent total RNA extracted from cells cultured primarily in serum-free medium from a limbal explant (Pa0) and epithelial cells after 5 passages (Pa5) yielded two individual data sets. As the microarray experiment was performed in duplicate, a simple differential analysis was performed with the four individual data reported that compared the gene expression between Pa0 and Pa5 cells. The differences in gene expression between the LFC and CFC profile and the Pa0 versus the Pa5 profile were first analyzed. Spearman’s rank correlation test was used to correlate the ranking of the differential gene expression between the replicate microarray experiments, a method used similar to the experiments on the fetal cornea.

Semiquantitative RT-PCR Validation of the Differential Expression of Selected Limbal Genes
The microarray results on the fetal cornea and limbal epithelial explants were confirmed using RT-PCR, for four genes that were shown to be upregulated: cytokeratin 15 (CK15), cytokeratin 14 (CK14), P-cadherin (CDH3), and Wnt-4. Total RNA was used for first-strand synthesis using 2.5% (vol/vol) of purified cDNA (reverse transcribed from 200 ng RNA) as a template for RT-PCR. The primers used are listed in Table 1 .


View this table:
[in this window]
[in a new window]

  		TABLE 1. Primers Used for RT-PCR

 
Semiquantitative RT-PCR Conditions
Amplification was performed using the preamplification system for first-strand cDNA (Superscript III; Invitrogen) and a DNA thermal cycler (PerkinElmer). Total RNA (equal amounts of RNA from all samples for all five genes) was reverse transcribed into cDNA using oligo-dT primers and reverse transcriptase enzyme (Superscript III; Invitrogen). PCR reactions were performed in a 25-µL reaction solution (for CK15, CK14, CDH3, Wnt-4, and GAPDH genes) containing 2.5 µL 10' PCR buffer, 3.5mM MgCl, 10 nanomoles deoxyribonucleoside-triphosphate, 0.6 micromoles of primers, and 2 µL of 10 µg/mL DNA polymerase (Invitrogen). Reactions were cycled 28 times for each primer pair analyzed for all five genes. For semiquantitative analysis, 28 cycles were chosen based on prior optimization, as with the number of cycles the polymerization of cDNA was still in the exponential range for the four genes of interest and GAPDH. The optimized PCR parameters were as follows: 94°C for the first denaturation process (94°C for 5 minutes), 28 cycles of denaturation (94°C for 30 seconds), annealing (65°C for 60 seconds), and extension (72°C for 30 seconds) and a final extension at 72°C for 5 minutes at termination of the PCR cycles. The PCR products were separated by electrophoresis in a 1% agarose gel made up in 1x TAE (triacetate EDTA) running buffer. PCR amplicon size was estimated using a 100- or 50-bp DNA ladder (Promega, Madison, WI).

For semiquantitative analysis, the PCR product of each gene (CK15, CK14, CDH3, and Wnt-4) signal in each sample was evaluated as an expression ratio to the GAPDH expression band (housekeeping gene), and the ratios were compared between samples (Pa0 to Pa5 cells) using the nonparametric Kruskal-Wallis test (statistical software, Prism ver. 4; GraphPad, San Diego, CA) to determine the significance of change in gene expression.

Immunohistochemistry for Single Protein Localization
Corneas dissected from postmortem human fetal and adult donors were either fixed in 10% neutral formalin and paraffin embedded or frozen in liquid nitrogen-cooled 2-methylbutane (OCT tissue-embedding medium; Tissue-Tek; Sakura Finetek, Tokyo, Japan) and stored at –80°C. Frozen sections (8 µm) cut with a cryostat (Microm HM 500; Carl Zeiss Meditec, Inc., Dublin, CA) were collected on electrostatically charged slides (Superfrost; Fisher Scientific, Pittsburgh, PA). Cultured limbal explant epithelial cells (serial passages: Pa1, Pa2, Pa3, Pa4, and Pa5) were centrifuged and revived overnight in serum-free media for 6 hours and the cell pellets were fixed in 10% neutral formalin overnight and paraffin embedded. Paraffin-embedded tissue (4 µm) sections on silane-coated slides were dewaxed in xylene (BDH, Poole, UK) and rehydrated in changes of 100% ethanol ' 2 and 70% ethanol '1, and double distilled water (ddH2O; MilliQ, Millipore Ltd., Tokyo, Japan). Cryostat sections were fixed in ice-cold 100% methanol for 10 minutes, air dried, and stored at –80°C. Before immunostaining, the sections were thawed and preincubated in 0.1% Triton X-100, 1x PBS for 1 hour at room temperature.

Paraffin-embedded fetal and adult human corneal tissue and limbal epithelial explant cell pellets of Pa1 to Pa5 samples on silane-coated slides were dewaxed and rehydrated as just described. A heat-mediated (750-W microwave) antigen retrieval method using a sodium citrate (0.1 M, pH 6.0, three times for 5 minutes each) bath was used. Slides cooled at room temperature for 5 minutes were then washed twice with TBS (Tris-buffered saline). Endogenous peroxidase was blocked with 3% H2O2 (Ajax Finechem, Australia) in methanol (APS, Australia) for 10 minutes. After the slides were washed twice for 5 minutes with 1x TBS, 20% normal goat serum (Vector Laboratories, Burlingame, CA) in 2% BSA in 1x TBS (bovine serum albumin) was used to prevent nonspecific staining by incubating the slides for 30 minutes at room temperature. Slides were then incubated with the primary antibody: LHK 15 clone primary mouse IgG2a anti-human CK15 (Biocare Medical, Concord, CA) in the final concentration of 2 µg/mL, ON359 clone primary mouse IgG3 anti-human CK14 (U.S. Biological, Swampscott, MA) in a final concentration of 1 µg/mL, and 3G51 clone primary mouse IgG1 anti-human P-Cadherin (U.S. Biological) in a final concentration of 2 µg/mL were used as the primary antibodies (used on formalin-fixed paraffin-embedded sections). Sections were incubated with the primary antibody for preoptimized times (60 minutes at 37°C) and then washed three times for 5 minutes each with 1x TBS and then incubated with an appropriate secondary antibody (goat anti-mouse; Dako) for 30 minutes at room temperature. After three slide washes with 1x TBS (5 minutes), the slides were incubated with 1:100 diluted streptavidin conjugated to horseradish peroxidase (Vector Laboratories, Inc.) for 60 minutes at room temperature. After three washes with 1x TBS for 5 minutes, AEC (3-amino-9-ethylcarbazole; Sigma-Aldrich) was used as a substrate chromogen. One microliter of 30% H2O2 used to initiate the reaction was added to the AEC and then placed on the slides for 5 minutes; a positive reaction indicated by a dark red color. The slides were counterstained with Meyer hematoxylin (Dako) for 20 seconds.

Cryostat corneal sections on silane coated slides were thawed and preincubated in 0.1% Triton X-100 in 1x TBS for 60 minutes at room temperature and then fixed for 10 minutes in acetone. Endogenous peroxidases were blocked with 3% H2O2 (Ajax Finechem, Seven Hills, NSW, Australia) in methanol (APS) for 10 minutes. The slides were washed for 5 minutes twice with 1x TBS. Incubation with 20% normal rabbit serum (Vector Laboratories, Inc.) for 30 minutes at room temperature was used to prevent nonspecific staining before incubation with Wnt-4 primary antibody. Tissue slides were incubated with 1:50 diluted primary goat polyclonal anti-Wnt-4 antibody (Santa Cruz Biotechnology) at 37°C for 60 minutes and washed three times for 5 minutes each with 1x TBS and then incubated with secondary rabbit anti-goat antibody for 30 minutes at room temperature. After three TBS washes, slides were incubated with 1:100 diluted streptavidin conjugated to horseradish peroxidase (Vector Laboratories, Inc.) for 60 minutes at room temperature. With AEC used as the chromogen, the rest of the immunohistochemistry steps were as just described.

Colocalization of Slow-Cycling BrdU Label and Protein Expression
Whole human corneas were placed in six-well plates (Nalge Nunc) filled with serum-free TMEM (additives: 120 µg/mL penicillin G, 200 µg/mL streptomycin sulfate, 5 µg/mL amphotericin B, 5 µg/mL insulin, 5 µg/mL transferrin, and 5 ng/mL selenous acid). After 3 days’ incubation, the corneal cultures were treated with 10 micromoles BrdU for 72 hours (media with diluted BrdU was changed every 24 hours for 3 days). After the BrdU pulse treatment, the organ culture specimen was washed with PBS, and the serum-free TMEM was replaced and then changed every third day for 4 weeks. After 4 weeks, organ-culture corneas were formalin fixed and paraffin embedded for immunostaining studies.

Paraffin-embedded sections were dewaxed, rehydrated, and blocked for endogenous peroxidase activity as described above. Nonspecific staining was prevented by 20% goat serum block (rabbit serum used for Wnt-4) for 30 minutes. After three washes with 1x TBS (5 minutes), sections were treated with 2 N HCl for 1 hour at 37°C after 0.1 M sodium citrate antigen retrieval (pH 6.0). Tissue sections were then incubated with 1:50 diluted primary rat IgG2a anti-BrdU antibody in 1% BSA/TBS (10 µg/mL; MCA 2060T, clone BU1/75; ICR1; Serotec, Oxford, UK) at 37°C for 60 minutes. After three washes with TBS, the slides were incubated with alkaline phosphatase–conjugated goat anti-rat antibody diluted 1:100 for 30 minutes and then washed three times with TBS. The sections were treated with BCIP/NBT substrate chromogen (5-bromo-4-chloro-3-indoyl phosphate–nitroblue tetrazolium, K0598; Dako) and observed for blue color. The sections were incubated serially with the primary antibody for either of the second proteins (CK15, CK14, P-cadherin, and Wnt-4, as previously described for single immunohistochemical studies). Consecutive incubation conditions with secondary antibody and streptavidin conjugated to HRP and AEC chromogen were as described previously AEC was used to indicate the presence of the second protein, dark red color indicative of positive staining. The sections were not counterstained. The slides were dried in an oven for 10 minutes at 50°C after an aqueous tissue mount (Crystal Mount; Biomeda, Foster City, CA) was placed over the tissue sections. A microscope (BX51; Olympus, Tokyo, Japan) with a CCD camera (MagnaFire Optronics, Goleta, CA) was used for photographs.


   Results
Top
 Abstract
 Materials and Methods
 Results
Discussion
 Conclusion
 References
 
T7RNA Amplification
With the 0.05 µg (RiboAmp kit; Arcturus) total RNA was amplified to average of 50.04 µg (±2.925 µg) aRNA. The bulk of the aRNA product was 200 to 600 nucleotides in size as viewed on a 1.25% agarose gel electrophoresis after staining the sample (1 µg aRNA/10µL RNase free water) with 0.5 µg/mL ethidium bromide, and the electrophoresis was performed in a 1' RNA MOPS running buffer at 5 V/cm for ~90 minutes.

Limbal Epithelial Cell Cultures
Primary limbal explant cells did not stain for K3 (indicating their relative undifferentiated state), but their nuclei stained strongly for p63 suggesting that these cells were derived from the limbal basal epithelium. Cultured cells (after five serially trypsinized culture passages) were positive for K3, indicating that these cells retained their corneal epithelial nature.

Simple Change Analysis of Differential Gene Expression from Microarray Experiments and Correlation of Results between Duplicate Experiments
Stem cell pathway–related cDNA microarray hybridization with radiolabeled cDNA probes after RT-PCR on independent aRNA and RNA preparations from the cadaveric fetal corneas (LFC versus CFC) and cultured adult limbal explant epithelial cells (Pa0 versus Pa5) were compared. Two independent aRNA preparations (amplified by T7 RNA polymerase) from microscopically dissected superficial limbal versus superficial CFC (n = 2 replicates) were compared for in vivo gene expression yielding 2 individual data sets (LFC and CFC). Two independent total RNA samples extracted from cells cultured primarily in serum-free media from a limbal explant (Pa0) and epithelial cells after 5 passages (Pa5) yielded two individual data sets (n = 2 replicates). As the microarray experiment was performed in duplicate on each of the four samples, a simple analysis as performed comparing the LFC with CFC gene expression (x-fold difference) and the Pa0 cells gene expression with that of the Pa5 cells. This analysis was a metric for comparing a gene’s mRNA expression level between the two distinct experimental conditions.12 13 This simple differential expression helps identify a gene that is differentially expressed (up or down) and is described as an expression ratio or a log-2 ratio.

Phenotypic Profile of Limbal and Central Corneal Epithelium
As the microarray experiment was performed in duplicate for the two comparative experiments (in vitro LFC versus CFC and in vivo Pa0 cells versus Pa5 cells), a simple comparison was performed on data comparing the gene expression of the LFC with the CFC and of the Pa0 cells with the Pa5 cells. The differences in gene expression (more than twofold difference considered significant) between the LFC and CFC regional profile and Pa0 cells versus Pa5 cellular profile were studied. The raw data were background subtracted from the average signal of the three blank spots on the microarray membrane, and the corrected signals were then normalized to the housekeeping gene GAPDH. A simple change analysis was then performed to demonstrate the differential expression of 266 genes comparing in vivo LFC versus CFC and in vitro Pa0 versus Pa5 cells. Thirteen genes that were expressed only in the fetal limbus (when compared to the central fetal cornea) were also expressed in primary limbal explant epithelial cells. Ten genes that were highly expressed in the fetal limbus (when compared to the central fetal cornea) were expressed only in primary limbal explant epithelial cells. Eleven genes that were overexpressed in the fetal limbus (compared to the fetal central cornea) were also overexpressed in Pa0 cells compared with Pa5 cells (Table 2) .


View this table:
[in this window]
[in a new window]

  		TABLE 2. Expression Profile of the Fetal Limbus and Primary Cultured Adult Limbal Explant Epithelial Cells

 
Spearman’s Rank Correlation of Ranked Differentially Expressed Genes between Duplicate Experiments
Data from duplicate experimental sets were reproducible as indicated by a positive Spearman correlation (r = 0.829, 95% CI, 0.788–0.863; P < 0.0001) between rank orders of differentially expressed genes in the microarray experiments comparing LFC with the CFC. Thus, there is a consistency in the ranking order of the differentially expressed genes between the duplicate experiments comparing the fetal limbus with the fetal cornea. From the duplicate microarray experiments performed on Pa0 and Pa5 limbal epithelial populations, the reproducibility of the ranking order of the differentially expressed genes was also suggested by a significant Spearman correlation (r = 0.859, 95% CI, 0.824–0.887; P < 0.0001) between the rank orders of differentially expressed genes. Thus, although these experiments were performed in duplicate, there was consistency between data sets.

The upregulated genes in the fetal limbus (when compared to the central cornea) are displayed with their x-fold expression in Tables 2a and 2b . Table 2a lists the genes that are uniquely expressed in the human fetal limbus. Table 2b lists the rest of the genes that are upregulated more than twofold in human fetal limbus, when compared with the central fetal cornea. Genes expressed uniquely by the primary cultured limbal explant epithelial cells (Pa0) are displayed in Table 2c (compared with the limbal epithelial cells after five serial trypsinized culture passages). Genes expressed more than twofold higher in the Pa0 cells, when compared with the Pa5 cells, are listed in Table 2d .

From Table 2 , it can be observed that 13 genes that are uniquely expressed in the fetal limbus (compared to the CFC) are also uniquely expressed in the primary cultured cadaveric human limbal epithelium (compared with the Pa5 limbal epithelial cells). Thus there is no expression of these genes at the fifth trypsinized passage. Nine genes that were highly expressed (more than twofold) in the human fetal limbus (compared with the central fetal cornea) were also highly expressed in the primary cultured human limbal epithelial cells (compared with the Pa5 limbal epithelial cells) with no expression in the Pa5 cells. Eleven genes that were highly expressed (more than twofold) in the human fetal limbus when compared to the central cornea were also highly expressed (more than twofold) in primary cultured human limbal epithelial cells when compared with the expression of cells after the fifth passage in culture.

RT-PCR Validation and Semiquantitative Analysis
RT-PCR study of the differential expression of the four selected genes revealed positive expression at the mRNA level in the fetal limbal cornea compared to no gene expression in the central cornea (Table 3 ; Figs. 2a 3a 4a 5a , for CK15, CK14, CDH3, and Wnt-4, respectively). A GAPDH normalized semiquantitative analysis revealed a highest expression of these four genes in primary cultured limbal epithelial explants (Pa0) and these gene expressions decreased significantly when measured using a GAPDH-normalized semiquantitative analysis, as these cells are passaged from the first to the fifth time (Pa5) (see Figs. 2h 3h 4h 5h 6 and Table 3 ). A non-parametric Kruskal-Wallis test comparing the GAPDH normalized gene expression across samples (Pa0 to Pa5 populations) revealed a statistically significant (P < 0.05) decrease in gene expression for all four genes (Table 3 , Fig. 6 ).


View this table:
[in this window]
[in a new window]

  		TABLE 3. Gene Expression in Cultured Human Limbal Epithelial Cells and Fetal Limbus

 

Figure 2
View larger version (60K):
[in this window]
[in a new window]

  		FIGURE 2. CK15 expression. (a) RT-PCR confirming the expression of CK15 in the limbal region of the fetal cornea at 18 weeks’ gestational age (versus fetal central cornea). (b) Eighteen-week-old fetal cadaveric cornea: immunohistochemistry revealing the presence of CK15 in basal limbal epithelial cells in the limbal palisades. (c) Isotype-matched control fetal limbal section; (d) absence in the central corneal epithelium. (e) Adult human cadaveric cornea: immunohistochemistry revealing presence of CK15 in basal limbal epithelium cells in the limbal palisades. (f) Isotype-matched control adult limbal section. (g) Absence in the central corneal epithelium. (h) Semiquantitative RT-PCR demonstrating the decrease to absent expression of CK15 in the primary limbal explant epithelial cells from primary explant through to 5 passages (Pa0–Pa5). (hi) CK15 expression; (hii) GAPDH expression. Kruskal-Wallis test comparing the cellular expression between groups (P < 0.05). (i–m) Immunohistochemical comparison of CK15 expression in primary limbal epithelial explants through to passage 5 (Pa1–Pa5). Bars, 100 µm. Inset magnification, x40.

 

Figure 3
View larger version (70K):
[in this window]
[in a new window]

  		FIGURE 3. CK14 expression. (a) RT-PCR confirming the expression of CK14 in limbal region of the fetal cornea at 18 weeks’ gestational age (versus fetal central cornea). (b) Eighteen-week-old fetal cadaveric cornea: immunohistochemistry revealing the presence of CK14 in basal limbal epithelium cells in the limbal palisades. (c) Isotype-matched control fetal limbal section. (d) Absence in the central corneal epithelium (right). (e) Adult human cadaveric cornea: immunohistochemistry revealing the presence of CK14 in basal limbal epithelial cells in the limbal palisades. (f) Isotype-matched control adult limbal section. (g) Absence in the central corneal epithelium. (h) Semiquantitative RT-PCR demonstrating the decreasing expression of CK14 in the primary limbal explant epithelial cells from primary explant through 5 passages (Pa0–Pa5). (hi) CK14 expression; (hii) GAPDH expression. Kruskal-Wallis test comparing the cellular expression between groups (P < 0.05). (i–m) immunohistochemical comparison of CK14 expression in primary limbal epithelial explants: primary explant to passage 5 (Pa1–Pa5). Bars, 100 µm. Inset magnification, x40.

 

Figure 4
View larger version (64K):
[in this window]
[in a new window]

  		FIGURE 4. P-cadherin expression. (a) RT-PCR confirming the expression of P-cadherin in the limbal region of the fetal cornea at 18 weeks’ gestational age (versus fetal central cornea). (b) Eighteen-week-old fetal cadaveric cornea: immunohistochemistry revealing the presence of P-cadherin in basal and immediate parabasal limbal epithelium cells in the limbal palisades. (c) Isotype-matched control fetal limbal section and (d) absence in the central corneal epithelium. (e) Adult human cadaveric cornea: immunohistochemistry revealing the presence of P-cadherin in basal and immediate parabasal limbal epithelial cells in the limbal palisades. (f) Isotype-matched control adult limbal section and (g) absence in the central corneal epithelium. (h) Semiquantitative RT-PCR demonstrating the decreasing expression of P-cadherin in the primary limbal explant epithelial cells from primary explant through to 5 passages (Pa0–Pa5). (hi) P-cadherin expression; (hii) GAPDH expression. Kruskal-Wallis test comparing the cellular expression between groups P < 0.05). (i–m) Immunohistochemical comparison of P-cadherin expression in primary limbal epithelial explants: primary explant through to passage 5 (Pa1–Pa5). Bars, 100 µm. Inset magnification, x40.

 

Figure 5
View larger version (66K):
[in this window]
[in a new window]

  		FIGURE 5. Wnt4 expression. (a) RT-PCR confirming the expression of Wnt4 in limbal region of the fetal cornea at 18 weeks’ gestational age (versus fetal central cornea). (b) Eighteen-week-old fetal cadaveric cornea: immunohistochemistry revealing the presence of Wnt4 in basal and immediate parabasal limbal epithelial cells in the limbal palisades. (c) Isotype-matched control fetal limbal section and (d) absence in the central corneal epithelium. (e) Adult human cadaveric cornea: immunohistochemistry revealing the presence of Wnt-4 in basal and immediate parabasal limbal epithelial cells in the limbal palisades, (f) Isotype-matched control adult limbal section and (g) absence in the central corneal epithelium. (h) Semiquantitative RT-PCR demonstrating the decrease to absence of the expression of Wnt-4 in the primary limbal explant epithelial cells from primary explant through to 5 passages (Pa0–Pa5). (hi) Wnt-4 expression; (hii) GAPDH expression. Kruskal-Wallis test comparing the cellular expression between groups (P < 0.05). (i–m) Immunohistochemical comparison of Wnt-4 expression in primary limbal epithelial explants: primary explant through to passage 5 (Pa1–Pa5). Bars, 100 µm. Inset magnification, x40.

 

Figure 6
View larger version (32K):
[in this window]
[in a new window]

  		FIGURE 6. Graphic representation of the semiquantitative analysis of gene expression (GAPDH normalized) in human limbal epithelial populations (Pa0–Pa5): (a) CK15, (b) CK14, (c) CDH3, and (d) Wnt-4 expression.

 
Single-Protein Immunohistochemistry Studies
Immunostaining for CK15 was positive in the basal layer of both the human fetal limbal and adult limbal epithelium. The cells within the crypts were stained more strongly than the others. The central corneal epithelium in both the human fetal (n = 4, 18 weeks’ gestational age) and human adult cornea (n = 5) did not express CK15 (Figs. 2b 2c 2d 2e 2f 2g) .

CK14 was expressed at the protein level in the limbal basal epithelium of both fetal corneas (n = 4) and adult human corneas (n = 5) and was not expressed in the central corneal epithelium (Figs. 3b 3c 3d 3e 3f 3g) . P-cadherin and Wnt-4 are expressed in the limbal basal and immediate parabasal epithelial layer of fetal (n = 3) and adult human cornea (n = 4 for frozen corneal sections; Figs. 4b 4c 4d 4e 4f 4g 5b 5c 5d 5e 5f 5g , respectively). P-cadherin protein was absent in the central cornea of both the adult and fetal human corneas. Immunostaining was negative for Wnt-4 in the central human fetal and adult corneal epithelia.

Double Immunostaining in Organ-Cultured Corneas
Hematoxylin and eosin staining of the corneal organ cultures revealed a five- to six-layered nonkeratinizing epithelium in the limbus, with the palisade-like structure still obvious at the end of 4 weeks (Figs. 7a 7b) . After a chase for 4 weeks, the corneal organ cultures initially pulse treated with BrdU for 72 hours. Four weeks later slow-cycling BrdU retaining cells (blue reaction in the nucleus demonstrated presence of BrdU caused by reaction of alkaline phosphatase with the chromogen BCIP/NBT) were found to express CK15, CK14, CDH3 (P-cadherin) and Wnt-4 in double immunolocalization studies (Fig. 7c 7d 7e 7f 7g 7h 7i 7j) , red color indicating positive staining for the protein of interest. These slow-cycling cells in the limbus, an indication of stemness, can be thus described to express these four proteins.


Figure 7
View larger version (82K):
[in this window]
[in a new window]

  		FIGURE 7. Colocalization of the four selected genes. Double immunostaining of BrdU pulse-labeled organ culture and chase of the serum-free organ culture after 4 weeks. Hematoxylin and eosin staining of adult corneal (organ culture) (a) limbus and (b) central cornea 4 weeks after initial BrdU pulse labeling. (c, e, g) Isotype controls for (d), (f), and (h), respectively. (d) Alkaline phosphatase reaction (blue) demonstrated the presence of BrdU (blue) and peroxidase reaction (red) indicating (d) CK15, (f) CK14, (h) P-cadherin and (j) Wnt-4 expression. (i) Negative control for (j). All insets: magnified views. Bars, 100 µm.

 

   Discussion
Top
 Abstract
 Materials and Methods
 Results
 Discussion
Conclusion
 References
 
In this study, we used a stem cell gene pathway cDNA microarray to analyze the transcriptome of the stem cell–rich limbal cornea and primary limbal explant epithelial cells and compared it with the central fetal cornea and fifth-passage limbal epithelial cells, respectively. Genes on the microarray were chosen based on previous reports of their expression in other stem cell populations. We found 33 genes to be differentially highly expressed in the limbal region of fetal corneas compared with the central cornea. These genes were also highly expressed in cultured primary limbal explant epithelial cells when compared to the cells after five passages in culture. The differential expression of four selected genes revealed that CK15, CK14, CDH3, and Wnt-4 were upregulated in the limbal cornea and limbal explant cultured primary cells. This finding was confirmed in the basal limbal epithelium by RT-PCR and immunohistochemistry. These genes were also found to be expressed in the slow-cycling (BrdU retaining) limbal epithelial of corneal organ culture tissues. As with other stem cells, a molecular signature of these cells may ultimately include a number of phenotypic markers that will help define a unique single-cell population. Genes were selected for further study by semiquantitative RT-PCR, based on three different functions: (1) cytoskeletal support (CK15 and CK14), (2) adhesion to the basement membrane (P- cadherin; CDH3), and (3) morphogenetic signaling (Wnt-4 of the Wnt signaling pathway).

Cytokeratin 15 (CK15) has been described as a hair follicle stem cell marker.14 It is expressed in the basal ectoderm layer of the developing fetal epidermis and is seen in certain isolated regions of the developed epidermis. CK15 expression occurs at an early stage of keratinocyte differentiation, a stage that precedes the commitment of epidermal cells to an epidermal or hairlike lineage.15 The fact that these relatively primitive epidermal cells express CK15, can be speculated at this stage that a similar CK15 expression by the relatively primitive basal limbal epithelium may hold true due to the common embryonic origin of these two tissues. CK15 is mainly expressed in the basal keratinocytes of stratified epithelial tissues and developing tissues like the fetal epidermis, nail, and the outer root sheath.14 We have demonstrated for corneal limbal stem cells, that with serial passage the expression of CK15 decreases, a finding that has also been described in actively proliferating keratinocytes.16 17 18 Graft versus host disease (GVHD) is known to affect the stem cells by triggering an inflammatory response and initiating the differentiation and proliferation of these cells. In the cornea, GVHD affects the limbus and its local stem cell population, which we speculate at this stage, is similar to the GVHD involvement of the retelike prominences of the lingual epithelium in mice. The GVHD in mice lingual epithelium initiates a reduction in CK15 expression in these cells.17 Though it has not yet been studied, we hypothesize that a similar process may be involved in the basal limbal epithelium in GVHD affecting the eye. The high prevalence of ocular dermoids at the limbus, containing hair and sebaceous glands, might correlate with the fact that these limbal basal cells express CK15, which is considered a marker for hair follicle stem cells.

CK14 has been described as a proliferative marker in cells of the basal layer of the stratified epidermis, and its expression has been related to the ectoderm development of the fetal epidermis. As described, we postulate that due to the common ectodermal origin of the epidermis and the corneal epithelium and the expression of CK14 in the cells of both tissues, it is possible that CK14 is a phenotypic marker of these stem cells. The fact that CK14 continues to be expressed in limbal derived explant epithelial cultures through five trypsinized passages, suggests that CK14 is expressed by both limbal and transient amplifying cells. Mutations in the CK14 gene have been described in blistering skin disease epidermolysis bullosa, which also affects mucosal surfaces and corneal epithelium. Transgenic mice null for CK14 and CK16 blister extensively and die within 2 days after birth. Mice purely null for CK14 display alopecia and chronic epidermal ulcers. CK14 has been described as a marker of murine epidermal stem cells, along with p63 and integrin ß1.19 CK14 in the human cornea is associated with hemidesmosome formation and has been used to study the attachment of limbal cell cultures to sheets of amniotic membrane.20 The common embryonic origin of the skin and corneal epithelium from surface ectoderm, suggests they may share stem cell markers.

P-cadherin (CDH3) is a member of the classic cadherin family and is known to be coexpressed with E-cadherin in the retinal pigment epithelium,21 as well as in the skin and follicular epithelium.22 CDH3 plays a role in cellular adhesion, motility, invasion, and signaling in malignant cells. It is mainly expressed in the placenta.23 Stem cells and malignant cells share some properties such as a high proliferative potential and a relatively primitive state (lack of differentiated state). As in malignancies, we speculate that P-cadherin plays a similar role in limbal basal epithelium as in the malignant cells. Changes in expression of the E- and/or P-cadherin expression probably participate in the delay of terminal differentiation of keratinocytes in wound healing.24 CDH3 is strongly upregulated in some bullous skin diseases and Darier’s disease.25 It functions in homotypic–homophilic cellular adhesiveness.26 27 28 29 30 31 The expression of CDH3 is restricted to basal and stems cells of stratified epithelia, such as the epidermis and urothelium. P-cadherin is present in the membranes of myoepithelial and basal stem cells of the ducts and terminal ductolobular units. An association between CDH3 expression in human breast carcinoma cells and an embryonic myoepithelial stem cell–like phenotype has been reported.27 28 29

The Wnt genes encode short-range secreted signaling molecules that regulate cell fate, adhesion, shape, proliferation, differentiation and movement. They are required for the development of multiple organ systems.32 33 34 35 Wnt-4 is diffusely expressed in the matrix and on immature hair shaft precursor cells. Wnts 10b, 10a, 3a, and 4 lie adjacent to and include precursor cells of the hair shaft cortex and cuticle and therefore may signal to hair shaft precursors. Studies identifying expression of Wnt genes in mouse embryonic skin at the onset of hair follicle induction, revealed the expression of Wnts 3, 4, 5a, 6, 7b, 10a, 10b, and 11.36 37 38 39 40 41 Wnt-4 is important for the mesenchymal to epithelial transition of the developing murine kidney.42 43 It is expressed in the nephrogenic mesenchyme in response to signals originating from the ureteric bud.44 45 Wnt-4 works in an autocrine manner to activate endogenous frizzled receptor(s) and downstream Wnt pathways.45 Wnt-4-dependent mesenchyme to epithelial transition results in the formation of tubules that further branch to form each nephron’s collecting elements.42 46 47 48 Wnt-4 is also expressed in tissues such as the surface ectoderm and neural tube.47 48 49 Wnt-4 advances the tubulogenesis and branching of mammary gland.49 50 51 In summary, Wnt-4 functions in distinct developmental aspects ultimately sharing common morphogenic objectives. It is thus not surprising that this important molecule is expressed by the basal limbal epithelium which may play a crucial role in differentiation.


   Conclusion
Top
 Abstract
 Materials and Methods
 Results
 Discussion
 Conclusion
References
 
A focused stem cell pathway microarray analysis was performed on fetal cornea and cultured limbal explant epithelium. Our results indicate that the limbal basal epithelium expresses CK15, CK14, CDH3, and Wnt-4. The expression profile was concordant with stratified epithelial tissues like the skin, both originating from the same embryonic layer of ectoderm. A definition of the expression profiles of the limbal epithelium should help their isolation and further characterization. Identification of limbal stem cells by their phenotype will facilitate studies directed toward therapeutic expansion and replacement procedures in states of limbal stem cell deficiency and transplantation.


   Acknowledgements
 
The authors thank Peter McCluskey (Inflammatory Diseases Research Unit, University of New South Wales [UNSW]) for the acquisition of corneal tissue used for limbal explant cultures in this study; Kenneth Hsu and Anusha Hettiaratchi for their assistance on the P32 work; and John Hunt (Inflammatory Diseases Research Unit, UNSW), and Deborah Black (School of Medicine, School of Public Health and Medicine, UNSW), for their advice and assistance with statistical analysis of the data.


   Footnotes
 
Submitted for publication March 30, 2006; revised August 2, 2006; accepted November 20, 2006.

Disclosure: E.C. Figueira, None; N. Di Girolamo, None; M.T. Coroneo, None; D. Wakefield, 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.

Corresponding author: Edwin C. Figueira, PO Box 280, Macquarie St, Parramatta, NSW 2124, Australia; e_c_figueira{at}hotmail.com.


   References
Top
 Abstract
 Materials and Methods
 Results
 Discussion
 Conclusion
 References
 

  1. Davanger M, Evensen A. Role of the pericorneal papillary structure in renewal of corneal epithelium. Nature. 1971;229:560–561.[CrossRef][Medline][Order article via Infotrieve]
  2. Tseng SC. Concept and application of limbal stem cells. Eye. 1989;3:141–157.
  3. Cotsarelis G, Cheng SZ, Dong G, Sun TT, Lavker RM. Existence of slow-cycling limbal epithelial basal cells that can be preferentially stimulated to proliferate: implications on epithelial stem cells. Cell. 1989;57:201–209.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  4. Watt FM, Hogan BL. Out of Eden: stem cells and their niches. Science. 2000;287:1427–1430.[Abstract/Free Full Text]
  5. Dua HS, Azuara-Blanco A. Limbal stem cells of the corneal epithelium. Surv Ophthalmol. 2000;44:415–425.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  6. Chen Z, de Paiva CS, Luo L, Kretzer FL, Pflugfelder SC, Li DQ. Characterization of putative stem cell phenotype in human limbal epithelia. Stem Cells. 2004;22:355–366.[Abstract/Free Full Text]
  7. Yew DT, Sha O, Li WW, Lam TT, Lorke DE. Proliferation and apoptosis in the epithelium of the developing human cornea and conjunctiva. Life Sci. 2001;68:2987–3003.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  8. Eberwine J, Yeh H, Miyashiro K, et al. Analysis of gene expression in single live neurons. Proc Natl Acad Sci USA. 1992;89:3010–3014.[Abstract/Free Full Text]
  9. Phillips J, Eberwine JH. Antisense RNA amplification: a linear amplification method for analyzing the mRNA population from single living cells. Methods. 1996;10:283–288.[CrossRef][Medline][Order article via Infotrieve]
  10. Joseph A, Powell-Richards AO, Shanmuganathan VA, Dua HS. Epithelial cell characteristics of cultured human limbal explants. Br J Ophthalmol. 2004;88:393–398.[Abstract/Free Full Text]
  11. Kim HS, Jun Song X, de Paiva CS, Chen Z, Pflugfelder SC, Li DQ. Phenotypic characterization of human corneal epithelial cells expanded ex vivo from limbal explant and single cell cultures. Exp Eye Res. 2004;79:41–49.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  12. Allison DB, Cui X, Page GP, Sabripour M. Microarray data analysis: from disarray to consolidation and consensus. Nat Rev Genet. 2006;7:55–65.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  13. Bjorkbacka H. Microarray Gene Expression Data. 2004;19–43. Kluwer Academic Publishers Boston.
  14. Lyle S, Christofidou-Solomidou M, Liu Y, Elder DE, Albelda S, Cotsarelis G. The C8/144B monoclonal antibody recognizes cytokeratin 15 and defines the location of human hair follicle stem cells. J Cell Sci. 1998;111:3179–3188.[Abstract]
  15. Whitbread LA, Powell BC. Expression of the intermediate filament keratin gene, K15, in the basal cell layers of epithelia and the hair follicle. Exp Cell Res. 1998;244:448–459.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  16. Waseem A, Dogan B, Tidman N, et al. Keratin 15 expression in stratified epithelia: downregulation in activated keratinocytes. J Invest Dermatol. 1999;112:362–369.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  17. Jakic-Razumovic J, Browne MD, Sale GE. Proliferation rates in epidermis of patients with graft-versus-host disease, non-specific inflammation and normal skin. Bone Marrow Transplant. 1992;10:27–31.[Medline][Order article via Infotrieve]
  18. Werner S, Munz B. Suppression of keratin 15 expression by transforming growth factor beta in vitro and by cutaneous injury in vivo. Exp Cell Res. 2000;254:80–90.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  19. Zhou JX, Jia LW, Yang YJ, Peng S, Cao YJ, Duan EK. Enrichment and characterization of mouse putative epidermal stem cells. Cell Biol Int. 2004;28:523–529.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  20. Meller D, Pires RT, Tseng SC. Ex vivo preservation and expansion of human limbal epithelial stem cells on amniotic membrane cultures. Br J Ophthalmol. 2002;86:463–471.[Abstract/Free Full Text]
  21. Burke JM, Cao F, Irving PE, Skumatz CM. Expression of E-cadherin by human retinal pigment epithelium: delayed expression in vitro. Invest Ophthalmol Vis Sci. 1999;40:2963–2970.[Abstract/Free Full Text]
  22. Muller-Rover S, Tokura Y, Welker P, et al. E- and P-cadherin expression during murine hair follicle morphogenesis and cycling. Exp Dermatol. 1999;8:237–246.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  23. Nose A, Takeichi M. A novel cadherin cell adhesion molecule: its expression patterns associated with implantation and organogenesis of mouse embryos. J Cell Biol. 1986;103:2649–2658.[Abstract/Free Full Text]
  24. Koizumi M, Matsuzaki T, Ihara S. Expression of P-cadherin distinct from that of E-cadherin in re-epithelialization in neonatal rat skin. Dev Growth Differ. 2005;47:75–85.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  25. Kovacs A, Schmidt E, Begany A, Hunyadi J, Szegedi A. Immunohistochemical examination of P-cadherin in bullous and acantholytic skin diseases. Acta Derm Venereol. 2004;84:116–119.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  26. Paredes J, Milanezi F, Viegas L, Amendoeira I, Schmitt F. P-cadherin expression is associated with high-grade ductal carcinoma in situ of the breast. Virchows Arch. 2002;440:16–21.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  27. Gamallo C, Moreno-Bueno G, Sarrio D, Calero F, Hardisson D, Palacios J. The prognostic significance of P-cadherin in infiltrating ductal breast carcinoma. Mod Pathol. 2001;14:650–654.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  28. Madhavan M, Srinivas P, Abraham E, et al. Cadherins as predictive markers of nodal metastasis in breast cancer. Mod Pathol. 2001;14:423–427.[CrossRef][Web of Science]
  29. Palacios J, Benito N, Pizarro A, et al. Anomalous expression of P-cadherin in breast carcinoma: correlation with E-cadherin expression and pathological features. Am J Pathol. 1995;146:605–612.[Abstract]
  30. Han AC, Soler AP, Knudsen KA, Salazar H. Distinct cadherin profiles in special variant carcinomas and other tumors of the breast. Hum Pathol. 1999;30:1035–1039.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  31. Peralta Soler A, Knudsen KA, Salazar H, Han AC, Keshgegian AA. P-cadherin expression in breast carcinoma indicates poor survival. Cancer. 1999;86:1263–1272.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  32. Du SJ, Purcell SM, Christian JL, McGrew LL, Moon RT. Identification of distinct classes and functional domains of Wnts through expression of wild-type and chimeric proteins in Xenopus embryos. Mol Cell Biol. 1995;15:2625–2634.[Abstract]
  33. Cadigan KM, Nusse R. Wnt signaling: a common theme in animal development. Genes Dev. 1997;11:3286–3305.[Free Full Text]
  34. Wodarz A, Nusse R. Mechanisms of Wnt signaling in development. Annu Rev Cell Dev Biol. 1998;14:59–88.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  35. Ungar AR, Kelly GM, Moon RT. Wnt4 affects morphogenesis when misexpressed in the zebrafish embryo. Mech Dev. 1995;52:153–164.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  36. Adamson MC, Dennis C, Delaney S, et al. Isolation and genetic mapping of two novel members of the murine Wnt gene family, Wnt11 and Wnt12, and the mapping of Wnt5a and Wnt7a. Genomics. 1994;24:9–13.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  37. Christiansen JH, Dennis CL, Wicking CA, Monkley SJ, Wilkinson DG, Wainwright BJ. Murine Wnt-11 and Wnt-12 have temporally and spatially restricted expression patterns during embryonic development. Mech Dev. 1995;51:341–350.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  38. Gavin BJ, McMahon JA, McMahon AP. Expression of multiple novel Wnt-1/int-1-related genes during fetal and adult mouse development. Genes Dev. 1990;4:2319–2332.[Abstract/Free Full Text]
  39. Roelink H, Wagenaar E, Lopes da Silva S, Nusse R. Wnt-3, a gene activated by proviral insertion in mouse mammary tumors, is homologous to int-1/Wnt-1 and is normally expressed in mouse embryos and adult brain. Proc Natl Acad Sci USA. 1990;87:4519–4523.[Abstract/Free Full Text]
  40. Roelink H, Nusse R. Expression of two members of the Wnt family during mouse development: restricted temporal and spatial patterns in the developing neural tube. Genes Dev. 1991;5:381–388.[Abstract/Free Full Text]
  41. Wang LC, Liu ZY, Gambardella L, et al. Regular articles: conditional disruption of hedgehog signaling pathway defines its critical role in hair development and regeneration. J Invest Dermatol. 2000;114:901–908.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  42. Kispert A, Vainio S, McMahon AP. Wnt-4 is a mesenchymal signal for epithelial transformation of metanephric mesenchyme in the developing kidney. Development. 1998;125:4225–4234.[Abstract]
  43. Stark K, Vainio S, Vassileva G, McMahon AP. Epithelial transformation of metanephric mesenchyme in the developing kidney regulated by Wnt-4. Nature. 1994;372:679–683.[CrossRef][Medline][Order article via Infotrieve]
  44. Barasch J, Yang J, Ware CB, et al. Mesenchymal to epithelial conversion in rat metanephros is induced by LIF. Cell. 1999;99:377–386.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  45. Carroll TJ, McMahon AP. Secreted molecules in metanephric induction. J Am Soc Nephrol. 2000;11(suppl 16)S116–S119.[Abstract/Free Full Text]
  46. Vainio SJ. Nephrogenesis regulated by Wnt signaling. J Nephrol. 2003;16:279–285.[Web of Science][Medline][Order article via Infotrieve]
  47. Karavanova ID, Dove LF, Resau JH, Perantoni AO. Conditioned medium from a rat ureteric bud cell line in combination with bFGF induces complete differentiation of isolated metanephric mesenchyme. Development. 1996;122:4159–4167.[Abstract]
  48. Kispert A, Vainio S, Shen L, Rowitch DH, McMahon AP. Proteoglycans are required for maintenance of Wnt-11 expression in the ureter tips. Development. 1996;122:3627–3637.[Abstract]
  49. Fan CM, Lee CS, Tessier-Lavigne M. A role for WNT proteins in induction of dermomyotome. Dev Biol. 1997;191:160–165.[CrossRef][Web of Science][Medline][Order article via Infotrieve]
  50. Brisken C, Heineman A, Chavarria T, et al. Essential function of Wnt-4 in mammary gland development downstream of progesterone signaling. Genes Dev. 2000;14:650–654.[Abstract/Free Full Text]
  51. Edwards PA. Control of the three-dimensional growth pattern of mammary epithelium: role of genes of the Wnt and erbB families studied using reconstituted epithelium. Biochem Soc Symp. 1998;63:21–34.[Medline][Order article via Infotrieve]




This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via ISI Web of Science (4)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Figueira, E. C.
Right arrow Articles by Wakefield, D.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Figueira, E. C.
Right arrow Articles by Wakefield, D.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS