Molecular Vision 2004; 10:867-873 <http://www.molvis.org/molvis/v10/a104/> Received 21 June 2004 | Accepted 15 November 2004 | Published 16 November 2004

Download Reprint

Cytokeratin 12 in human ocular surface epithelia is the antigen reactive with a commercial anti-Gαq antibody

Christopher S. Boehlke,¹ Ching Yuan,¹ Winston W.-Y. Kao,² Andrew J. W. Huang¹
 
 

¹Department of Ophthalmology, University of Minnesota, Minneapolis, MN; ²Department of Ophthalmology, University of Cincinnati, Cincinnati, OH

Correspondence to: Andrew J. W. Huang, MD, MPH, Department of Ophthalmology, University of Minnesota, 420 Delaware Street SE, MMC493, Minneapolis, MN, 55455; Phone: (612) 625-4400; FAX: (612) 626-4455; email: huang088@umn.edu

Abstract

Purpose: In our initial attempt to identify differentiation markers for ocular surface epithelia, we observed a unique staining pattern by a commercial anti-Gαq antibody. We further isolate and characterize the protein reactive with this anti-Gαq antibody in human ocular surface epithelia.

Methods: Human donor corneoscleral buttons were sectioned and stained with a battery of commercial antibodies against Gα proteins. Western blot analysis of cell lysates of corneal epithelial cells and HEK 293 cells transfected with Gαq cDNA was used to determine the identity of the protein reactive with the anti-Gαq antibody (E-17). Comparisons were made with another anti-Gαq antibody (G4415) and an anti-cytokeratin 12C (J7) antibody. The isolated proteins reactive with E17 and J7 were then analyzed with two dimensional isoelectric focusing. Polypeptide sequences were identified using matrix assisted laser desorption/ionization time of flight mass spectrometry (MALDI-TOF MS) after in-gel protein digestion.

Results: The E-17 anti-Gαq antibody preferentially stained the entire corneal epithelia and the suprabasal layers of the limbus with complete absence of staining in the basal limbus and conjunctiva. Western blot analysis of corneal epithelial cells showed that E-17 antibody identified a protein with a molecular weight of 55 kDa. However, the antibody did not react with the purported antigen, Gαq protein (42 kDa) produced by Gαq cDNA. Another anti-Gαq antibody (G4415) did not react with the 55 kDa protein but did react with the 42 kDa Gαq protein. Further comparison of the E-17 antibody with the J7 antibody revealed that both recognized the 55 kDa band in one and two dimensional analysis. MALDI-TOF MS analysis confirmed that the 55 kDa protein of interest was actually cytokeratin 12 (CK12), rather than Gαq protein.

Conclusions: The commercial E-17 anti-Gαq antibody did not react with Gαq protein, its purported antigen. Instead, it recognized a 55 kDa protein, which was characterized to be cytokeratin 12 by isoelectric focusing and peptide fingerprinting with mass spectrometry. Based on its reactivity with CK12, this commercial E-17 can be used as a differentiation marker to study ocular surface epithelia.

Introduction

Antibodies are required for research techniques such as immunohistochemistry, western blot analysis, and immunoprecipitation. Commercially available antibodies help to facilitate laboratory investigations by eliminating the need for antibody production. However, there are a limited number of antibodies currently available that are specific for epithelial proteins in the human ocular surface.

G-proteins are ubiquitous heterotrimeric guanine nucleotide binding proteins that function as signal transducers for several distinct intracellular signaling pathways. G-proteins are composed of α, β, and γ subunits, of which there are several α subunits including Gαs, Gαi, and Gαq. Gαs stimulates adenylyl cyclase to produce cyclic AMP (cAMP) to activate protein kinase A (PKA), whereas Gαi inhibits the same cascade [1]. Gαq stimulates phospholipase C (PLC) to produce the intracellular second messengers inositol triphosphate (IP3) and diacylgylcerol (DAG) [2]. IP3 causes the release of intracellular calcium while DAG activates protein kinase C (PKC) [2].

To the best of our knowledge, the roles of G proteins in the ocular surface epithelia have never been explored. We surmise that G-proteins may be involved in signal transduction in the ocular surface epithelia. For example, in corneal epithelium, mucin-like glycoprotein secretion is mediated by cAMP and PKC signaling pathways [3]. PKC may play a role in corneal epithelial gene expression and wound healing, as increased PKC activity has been observed in corneal epithelial wound healing and inhibition of PKC results in a significant delay in corneal re-epithelialization [4-6].

In our attempt to identify G-protein and its subunits as epithelial differentiation markers specific to ocular surface epithelia, we initially characterized the specific immunohistochemical staining pattern of a battery of commercially available anti-G protein antibodies in the human ocular surface. We noted that a commercial anti-Gαq (α subunit) antibody (E-17), differing from other anti-Gαq antibody, had a unique staining pattern in the ocular surface epithelia. Herein, we performed a detailed biochemical analysis to identify the unique corneal epithelial protein reactive with this E-17 antibody.

Methods

Tissue

Twelve fresh human donor corneoscleral buttons with small conjunctival skirts were obtained from the Minnesota Lions Eye Bank. Tissues were stored in the standard Optisol-GS corneal storage media (Bausch & Lomb, New York, NY). An exemption was obtained from the Human Subjects Committee of the University of Minnesota.

For protein analyses using western blot analysis and two dimensional isoelectric focusing (IEF), the donor corneoscleral buttons were prepared by first trimming away the conjunctival skirts. Using a cell scraper, corneal epithelial cells were then gently scraped from the entire corneal surface into a solution of 1X PBS and protease inhibitor cocktail (Sigma Co., St. Louis, MO). After centrifuging for 15 min, the supernatant was removed and the epithelial cells were preserved at -80 °C until further analysis.

Antibodies

Anti-Gαq (E-17; Santa Cruz Biotechnology, Inc, Santa Cruz, CA), a rabbit polyclonal antibody was purportedly developed against a peptide within the N-terminal domain of Gαq. Two distinct batches of the Gαq (E-17) antibody from the same company were tested.

The following antibodies were obtained from Sigma (St. Louis, MO); Anti-Gαq (G4415), developed against a synthetic peptide corresponding to amino acids 277-294 of Gαq, alkaline phosphatase conjugated goat anti-rabbit IgG, and goat anti-mouse IgG (as secondary antibodies).

The anti-cytokeratin 12C antibody (J7) was previously developed against the C-terminal domain of cytokeratin 12 (CK12) by one of the co-authors (WWYK).

Immunohistochemistry

Corneoscleral buttons were vertically bisected, embedded in optimal cutting temperature (OCT) compound, and snap frozen in liquid nitrogen. Tissues were cryosectioned to 6 μm thickness and mounted on colorfrost plus slides (Fisher Scientific, Pittsburgh, PA). At room temperature, the tissues were blocked with 10% goat serum in phosphate buffered saline (PBS, pH 7.4) for 1.5 h and then incubated with primary antibody diluted in PBS (1:100) for 1 h. The sections were washed twice for 15 min each with PBS/0.05% Tween 20 and then incubated for 1 h with alkaline phosphatase conjugated secondary antibodies. The 5-bromo-4-chloro-3-indolyl-phosphate/nitrotetrazolium blue chloride (BCIP/NBT) liquid substrate system (Sigma) was used to develop the tissues. Control slides were prepared with primary antibody substituted by pre-immune serum, but were otherwise subjected to the same protocol. Photomicrographs were obtained using a Zeiss Axiovert 200 microscope (Zeiss, Thornwood, NY). Tissues from three unrelated donors were used to confirm the specific staining pattern for each antibody and the immunostaining was repeated twice for each antibody to ensure the findings were reproducible. Two different batches of E-17 from the same manufacturer were also tested to reconfirm the staining pattern in ocular surface epithelia.

HEK 293 cell transfection

Full length human Gαq subunit cDNA in pCDNA 3.1 was obtained from Guthrie cDNA resource center (Sayre, PA). Plasmid DNA (5 μg) was mixed with 50 μl of lipofectamine (Invitrogen, Carlsbad, CA) in 0. 5 ml of Opti-Mem and incubated at room temperature for 30 min. The liposome-DNA complex was then added to a 10 cm plate of HEK 293 cells (ATCC, Manassas, VA) at approximately 80% confluency and incubated at 37 °C for 4 h. Cells were harvested 48 h later, gently scraped, centrifuged, and re-suspended in sample buffer for collection of Gαq protein.

Gel electrophoresis and immunoblotting

Lysates from corneal epithelial cells and HEK 293 cells transfected with Gαq cDNA were used for comparison. Cells were homogenized in an SDS containing buffer (1X PBS, 2% SDS, and β-mercaptoethanol) and the protein concentration was measured using the bicinchoninic acid (BCA) protein assay kit (Pierce, Rockford, IL) according to the manufacturer's instructions. One dimensional SDS polyacrylamide gel electrophoresis (SDS-PAGE) was used to characterize the molecular weight of the protein recognized by the E-17 antibody. Equal amounts of protein (about 20 μg/lane) were electrophoresed on 10% SDS-PAGE gels. Proteins were then transferred to nitrocellulose membranes (Immobilon; Millipore, Bedford, MA) and the membranes were blocked with a 1X Tris buffered saline (TBS, 10 mM Tris, pH 8.0 and 150 mM NaCl)/5% dry milk solution. Membranes were incubated with primary antibody for 1 h at room temperature or overnight at 4 °C and then washed twice with a solution of 1X TBS and 0.05% Tween 20. The blots were then incubated with secondary antibodies for 1 h at room temperature, washed, and developed with the BCIP/NBT liquid substrate system.

Preparative 2D electrophoresis

Definitive analysis of the corneal epithelial protein of interest was performed with two dimensional isoelectric focusing, silver staining, and in-gel digestion by trypsin for mass spectrometry. Soluble proteins were removed by combining corneal epithelial cells with a radioimmunoprecipitation assay solution (RIPA, 50 mM Tris-HCl, pH 7.5, 150 mM NaCl, 1% Triton X-100, 0.5% sodium deoxycholate, and 0.1% SDS), protease inhibitor, and phenylmethylsulfonyl fluoride (PMSF) for 1 h at 4 °C. After centrifugation and removal of supernatant, the pellet was then resuspended in an SDS containing sample buffer and the protein concentration was measured using bovine serum albumin (BSA) as a standard.

Two dimensional electrophoresis was performed using precast immobilized pH gradient (IPG) strips (pH 3-10, 7 cm; Bio-Rad, Hercules, CA) in the first dimension. Samples were applied via active rehydration of IPG strips using a PROTEAN IEF cell (Bio-Rad) according to the manufacturer's recommendations. Typically 500-600 μg of proteins were loaded on each IPG strip and focusing was carried out for 20,000 volt-hour. After IEF separation, the strips were immediately equilibrated twice for 15 min in an SDS containing buffer (6 M urea, 0.375 M Tris, pH 8.8, 2% SDS, and 20% glycerol). In the first equilibration solution DTT (2%) was included and 2.5% w/v iodoacetamide was added in the second equilibration step to alkylate thiols. SDS-PAGE was performed using 1.0 mm thick, 10% SDS/polyacrylamide gels and electrophoresis was carried out at constant voltage (100 V). Prior to silver staining, gels were fixed in 50% methanol/10% acetic acid, incubated in 20% ethanol, and then washed with water one time each. Following incubation with sodium thiocyanate (0.1 g/500 ml) for one min and then silver nitrate (0.25 g/125 ml) for 30 min, the gels were developed in buffer containing sodium carbonate, formaldehyde, and sodium thiocyanate. The reaction was stopped with 1% acetic acid and the proteins of interest were excised from the gel.

In-gel digestion

Digestion of protein in excised gel pieces was performed after destaining with a 1:1 solution of 30 mM potassium ferricyanide/100 mM sodium thiosulphate. Gel pieces were washed with 100% acetonitrile and dried by vacuum centrifugation. Modified porcine trypsin (12.5 ng/μl, sequence grade; Promega, Madison, WI) in digestion buffer (50 mM ammonium bicarbonate/5 mM calcium chloride) was added to the dry gel pieces and incubated on ice for 1 h for rehydration. After removing the supernatant, 30 μl of digestion buffer was added and digestion was continued at 37 °C overnight.

The resulting peptide mixture was extracted with 20 mM ammonium bicarbonate once, followed by 50% acetonitrile/5% formic acid three times. The supernatants were pooled and the peptides were dried by vacuum centrifugation.

Mass spectrometry

Polypeptide sequences were identified using matrix assisted laser desorption/ionization time of flight mass spectrometry (MALDI-TOF MS) after in-gel protein digestion. Immediately prior to mass spectrometry, the protein samples were reconstituted with a solution of acetonitrile/water (5:95, vol:vol) and trifluoroacetic acid. A ziptip was hydrated, the sample loaded, and water used to wash the sample. An elution solution of acetonitrile/water (60:40), and 0.1% trifluoroacetic acid was then loaded three times, α-cyano-4-hydroxycinnamic acid (CCA) matrix was added, and 1 μl of the sample mixture was spotted directly on a MALDI target for analysis. Peptide mass mapping of tryptic digests was carried out with a Brooker Biflex III MALDI-TOF mass spectrometer (Billarica, MA) in reflectron mode. Protein identification based on peptide mass fingerprint was performed with Mascot search software (Matrix Science Inc., Boston, MA; version 1.8).

Results

With the E-17 anti-Gαq antibody, a unique staining pattern was observed in ocular surface epithelia (Figure 1). The antibody preferentially stained the suprabasal epithelial layers of the limbus (Figure 1A) and entire corneal epithelium (Figure 1B). Complete absence of staining in the conjunctival epithelium (Figure 1C) was noted.

In order to confirm the identity of the protein stained by the E-17 antibody, we first transfected HEK 293 cells with Gαq cDNA to generate genuine Gαq protein (42 kDa). With western blot analysis, the E-17 antibody recognized a unique protein of 55 kDa from corneal epithelial lysates (Figure 2A, lane 1), but it did not recognize any proteins from the HEK 293 cells transfected with Gαq cDNA (Figure 2A, lane 2). Two different batches of E-17 from the same manufacturer showed similar results in immunohistochemistry and western blot analysis. For comparison, further analysis was also performed with an anti-Gαq antibody (G4415) from another manufacturer (Sigma). The G4415 anti-Gαq antibody did not recognize any protein from the corneal epithelial lysates (Figure 2B, lane 1). Instead, the G4415 antibody did react to the 42 kDa protein (Gαq) in the HEK 293 cells transfected with Gαq cDNA (Figure 2B, lane 2). The data indicated the two commercial antibodies, both presumably reactive with the Gαq, instead recognized two distinct proteins.

Since our immunohistochemical data with E-17 antibody showed an epithelial staining pattern that is in striking resemblance to the well known distribution of cytokeratin pair CK3/CK12 in ocular surface epithelia, we further analyzed the identity of the unknown protein recognized by E-17. Western blots using the E-17 antibody (Figure 2C) and a J7 (anti-cytokeratin 12C) antibody (Figure 2D) revealed that both antibodies recognized an equivalent 55 kDa band in the corneal epithelial cells.

To confirm that the identity of the epithelial protein recognized by E-17 was indeed CK12, 2D IEF and subsequent western blots were performed. Both E-17 and J7 antibodies stained equivalent bands of 55 kDa at a low pH (Figure 3A,B). This is consistent with CK12 being a member of the acidic cytokeratins.

The band corresponding to the 55 kDa protein of the epithelial lysates was excised from the 2D gel (Figure 4A) and subjected to in-gel digestion, followed by MALDI-TOF MS analysis. MALDI-TOF MS revealed a monoisotopic peptide mass fingerprint spectrum (Figure 4B). When the spectrum was analyzed with Mascot search software, it matched 25 peptides (51% sequence coverage) of human cytokeratin-12 (Figure 4C). A probability based Mowse score was 147 (Figure 4D), where the score is derived from (-10)log₁₀(P) in which P is the probability that the observed match is a random event. Mowse scores greater than 63 (or p<0.05) indicate a significant match of the protein sequences (Mowse scoring is described in detail on the Matrix Science web site) [7]. These results confirmed that the isolated peptides matched the CK12 sequence with a high degree of certainty and the protein recognized by E-17 was indeed CK12 with very high probability.

Discussion

Our results demonstrated that E-17 stained the entire corneal epithelium and suprabasal epithelia of the limbus with complete absence of staining in the conjunctiva and basal limbus. It reacted with a unique 55 kDa protein in corneal epithelial cells. Surprisingly, this presumed anti-Gαq antibody did not react with the bona fide 42 kDa Gαq from transfected HEK293 cells. Taken together, the data thus suggest that E-17 antibody preferentially stains the more differentiated cells of corneal epithelial lineage, sparing the limbal stem cell zones and conjunctiva. Such a unique pattern is in striking resemblance to the well known distribution of cytokeratin pair CK3/CK12 in ocular surface epithelia [8,9].

Cytokeratins are a family of water insoluble polypeptides that form 10 nm intermediate filaments, major components of the epithelial cell cytoskeleton. In vivo, an acidic keratin is usually paired with a basic keratin to form a heterodimer [10]. Both cytokeratins, acidic CK12 (55 kDa) and basic CK3 (64 kDa), are known to be expressed in the epithelia of the suprabasal limbus and entire cornea, but not in the basal limbal epithelium where corneal stem cells are thought to reside [8,9]. This CK3/CK12 pair is expressed in the corneal epithelia of humans, cows, guinea pigs, rabbits, and chickens and they have been regarded as specific markers of terminal differentiation in the corneal epithelium [11-19]. Our extensive biochemical analyses indicate that the 55 kDa protein was indeed CK12, which reacted equally with the E-17 and J7 antibodies. Consequently, we believe that E-17 represents the first commercially available rabbit derived antibody to recognize CK12 (55 kDa). Monoclonal antibodies to CK12 such as J7 [17,19] have been developed in the past, but are not commercially available. In addition to J7 and E-17, there are other anti-CK12 antibodies available such as goat derived sc-17098, sc-17099 and sc-17101 from Santa Cruz Biotechnology, Inc and a less selective monoclonal antibody (reacting to CK9, CK10, CK11, and CK12) from Biomeda Corporation (Foster City, CA).

Similar to other self renewing epithelial tissues such as skin or oral mucosa, the ocular surface epithelia are in a state of constant renewal, requiring stem cell participation. In contrast to human skin in which epidermal stem cells are uniformly distributed along the basal layer of the epidermis, the progenitors of the cornea reside in the narrow transition zone of limbal cells between the cornea and the conjunctiva [20-23]. The theory that corneal stem cells reside in the basal layers of the limbus was first proposed based, in part, on labeling patterns of CK3 by the monoclonal antibody AE5 [8,20]. It was observed that CK3 exists in the epithelia of the suprabasal limbus and entire cornea, but not in the epithelia of the basal limbus or conjunctiva. CK3 was also expressed in post-confluent stratified cell cultures but was not present in undifferentiated cells in culture. These data suggest that the basal limbal epithelium is less differentiated than the suprabasal limbal epithelium and entire corneal epithelium, and that the putative corneal stem cells reside in the basal limbus. The commercial AE5 monoclonal antibody has been used widely to study epithelial differentiation. The availability of commercial E-17 antibody which reacts to CK12 (the counterpart of CK3) should further augment the armamentarium for research on ocular surface epithelial differentiation.

The study of CK12 is important because the keratin 12 gene is the second most abundant gene isolated from a full length cDNA library of human corneal epithelium [24]. Furthermore, CK12 knockout mice have fragile corneal epithelia that are easily abraded [19], and mutations of human CK3 and CK12 genes are known to result in autosomal dominant Meesmann's corneal epithelial dystrophy [25,26]. Thus, the identification of a commercial antibody such as E-17 as a specific marker of corneal epithelial differentiation will help to facilitate further study of ocular surface epithelial differentiation and stem cell development.

In summary, we set out to investigate the unique immunohistochemical staining pattern of a commercial anti-Gαq antibody (E-17) in human ocular surface epithelia, and found that CK12, instead of the Gαq protein, is the protein in ocular surface epithelia responsible for the selective staining by E-17. Based on a protein sequence search and further comparison with other antibodies such as J7 and G4415, we did not find any antigenic similarity or crossreactivity between Gαq and CK12. Thus, we conclude that E-17 antibody represents another commercially available differentiation marker of corneal epithelial phenotype.

Acknowledgements

Supported in part by the Minnesota Medical Foundation (Research Grant 3180-9227-02 [AJWH]), Student Research Grant 3171-9295-02 (CSB), and by an unrestricted grant from Research to Prevent Blindness (AJWH, CY).

References

1. Sunahara RK, Dessauer CW, Gilman AG. Complexity and diversity of mammalian adenylyl cyclases. Annu Rev Pharmacol Toxicol 1996; 36:461-80.

2. Wu DQ, Lee CH, Rhee SG, Simon MI. Activation of phospholipase C by the alpha subunits of the Gq and G11 proteins in transfected Cos-7 cells. J Biol Chem 1992; 267:1811-7.

3. Nakamura M, Endo K, Nakata K. Mucin-like glycoprotein secretion is mediated by cyclic-AMP and protein kinase C signal transduction pathways in rat corneal epithelium. Exp Eye Res 1998; 66:513-9.

4. Chandrasekher G, Bazan NG, Bazan HE. Selective changes in protein kinase C (PKC) isoform expression in rabbit corneal epithelium during wound healing. Inhibition of corneal epithelial repair by PKCalpha antisense. Exp Eye Res 1998; 67:603-10.

5. Lin N, Bazan HE. Protein kinase C subspecies in rabbit corneal epithelium: increased activity of alpha subspecies during wound healing. Curr Eye Res 1992; 11:899-907.

6. Hirakata A, Gupta AG, Proia AD. Effect of protein kinase C inhibitors and activators on corneal re-epithelialization in the rat. Invest Ophthalmol Vis Sci 1993; 34:216-21.

7. Pappin DJ, Hojrup P, Bleasby AJ. Rapid identification of proteins by peptide-mass fingerprinting. Curr Biol 1993; 3:327-32.

8. Schermer A, Galvin S, Sun TT. Differentiation-related expression of a major 64K corneal keratin in vivo and in culture suggests limbal location of corneal epithelial stem cells. J Cell Biol 1986; 103:49-62.

9. Kurpakus MA, Stock EL, Jones JC. Expression of the 55-kD/64-kD corneal keratins in ocular surface epithelium. Invest Ophthalmol Vis Sci 1990; 31:448-56.

10. Moll R, Franke WW, Schiller DL, Geiger B, Krepler R. The catalog of human cytokeratins: patterns of expression in normal epithelia, tumors and cultured cells. Cell 1982; 31:11-24.

11. Tseng SC, Jarvinen MJ, Nelson WG, Huang JW, Woodcock-Mitchell J, Sun TT. Correlation of specific keratins with different types of epithelial differentiation: monoclonal antibody studies. Cell 1982; 30:361-72.

12. Tseng SC, Hatchell D, Tierney N, Huang AJ, Sun TT. Expression of specific keratin markers by rabbit corneal, conjunctival, and esophageal epithelia during vitamin A deficiency. J Cell Biol 1984; 99:2279-86.

13. Cooper D, Sun TT. Monoclonal antibody analysis of bovine epithelial keratins. Specific pairs as defined by coexpression. J Biol Chem 1986; 261:4646-54.

14. Rodrigues M, Ben-Zvi A, Krachmer J, Schermer A, Sun TT. Suprabasal expression of a 64-kilodalton keratin (no. 3) in developing human corneal epithelium. Differentiation 1987; 34:60-7.

15. Chaloin-Dufau C, Sun TT, Dhouailly D. Appearance of the keratin pair K3/K12 during embryonic and adult corneal epithelial differentiation in the chick and in the rabbit. Cell Differ Dev 1990; 32:97-108.

16. Cooper D, Schermer A, Sun TT. Classification of human epithelia and their neoplasms using monoclonal antibodies to keratins: strategies, applications, and limitations. Lab Invest 1985; 52:243-56.

17. Liu CY, Zhu G, Westerhausen-Larson A, Converse R, Kao CW, Sun TT, Kao WW. Cornea-specific expression of K12 keratin during mouse development. Curr Eye Res 1993; 12:963-74.

18. Liu CY, Zhu G, Converse R, Kao CW, Nakamura H, Tseng SC, Mui MM, Seyer J, Justice MJ, Stech ME, Hansen GM, Kao WW-Y. Characterization and chromosomal localization of the cornea-specific murine keratin gene Krt1.12. J Biol Chem 1994; 269:24627-36.

19. Kao WW, Liu CY, Converse RL, Shiraishi A, Kao CW, Ishizaki M, Doetschman T, Duffy J. Keratin 12-deficient mice have fragile corneal epithelia. Invest Ophthalmol Vis Sci 1996; 37:2572-84.

20. 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; 57:201-9.

21. Zieske JD. Perpetuation of stem cells in the eye. Eye 1994; 8 (Pt 2):163-9.

22. Kruse FE. Stem cells and corneal epithelial regeneration. Eye 1994; 8 (Pt 2):170-83.

23. Pellegrini G, Golisano O, Paterna P, Lambiase A, Bonini S, Rama P, De Luca M. Location and clonal analysis of stem cells and their differentiated progeny in the human ocular surface. J Cell Biol 1999; 145:769-82.

24. Nishida K, Adachi W, Shimizu-Matsumoto A, Kinoshita S, Mizuno K, Matsubara K, Okubo K. A gene expression profile of human corneal epithelium and the isolation of human keratin 12 cDNA. Invest Ophthalmol Vis Sci 1996; 37:1800-9.

25. Irvine AD, Corden LD, Swensson O, Swensson B, Moore JE, Frazer DG, Smith FJ, Knowlton RG, Christophers E, Rochels R, Uitto J, McLean WH. Mutations in cornea-specific keratin K3 or K12 genes cause Meesmann's corneal dystrophy. Nat Genet 1997; 16:184-7.

26. Nishida K, Honma Y, Dota A, Kawasaki S, Adachi W, Nakamura T, Quantock AJ, Hosotani H, Yamamoto S, Okada M, Shimomura Y, Kinoshita S. Isolation and chromosomal localization of a cornea-specific human keratin 12 gene and detection of four mutations in Meesmann corneal epithelial dystrophy. Am J Hum Genet 1997; 61:1268-75.