Regional variation in distribution of EGF receptor in developing and ...

Journal of Cell Science 106, 145-152 (1993) Printed in Great Britain © The Company of Biologists Limited 1993

145

Regional variation in distribution of EGF receptor in developing and adult corneal epithelium James D. Zieske* and Michael Wasson Cornea Unit, Schepens Eye Research Institute, 20 Staniford Street, Boston, Massachusetts 02114, and Department of Ophthalmology, Harvard Medical School, USA *Author for correspondence

SUMMARY Epidermal growth factor receptor has been localized to the proliferative cell layers in a variety of stratified squamous epithelia. In the current study, the rat cornea was used as an experimental model to determine if epidermal growth factor receptor is concentrated in epithelial stem cells. Epidermal growth factor receptor was localized using immunofluorescence microscopy in adult and neonatal (1-day to 4-week) rat corneas. Antibody binding to epidermal growth factor receptor was present in basal cells across the adult cornea but was more intense in the limbal zone. In rats 1 day to 1 week of age, the corneal epithelium consisted of one or two layer of cells that were intensely labeled by anti-epidermal growth factor receptor. Following epithelial stratification, which occurred just prior to eyelid opening (~12 days), expression of epidermal growth factor receptor was greatly reduced in central corneal epithelium and gained an adult pattern by 3 weeks of age. Expression of epidermal growth factor receptor was also

examined by incubating 1 mm slices of adult corneas with 125I-epidermal growth factor (4 nM) for 90 minutes, followed by washing and autoradiography. Basal cells in the limbal zone contained 4.5-fold more silver grains per cell than did basal cells in the central cornea. These data suggest that cells with high potential for proliferation, i.e. limbal basal cells and all basal cells in developing rats, express high epidermal growth factor receptor levels. High levels of receptor may allow these cells to be rapidly stimulated by growth factors to undergo cell division during development and following wounding in adult corneas. High epidermal growth factor receptor levels may also provide a mechanism whereby limbal basal cells are maintained in an undifferentiated stem cell state rather than entering the pathway of terminal differentiation.

INTRODUCTION

and insulin-like growth factor receptor (Yarden and Ullrich, 1988). After binding of their specific ligands, the growth factor receptors dimerize and undergo autophosphorylation of specific tyrosine residues. These phosphorylated tyrosine residues in turn become binding sites for a group of cytoplasmic signaling proteins that share the Src homology 2 domain (Pawson and Gish, 1992; Fantl et al., 1992; Kashishian et al., 1992). These signaling proteins, which include phospholipase C, phosphatidylinositol 3-kinase, and the GTPase activating protein Ras, are then capable of activating pathways that instruct the target cell to undergo proliferation and/or differentiation (Pawson and Gish, 1992). In stratified squamous epithelium, EGFR has been localized to the plasma membranes of cells in the basal layer (Green et al., 1983; Gusterson et al., 1984; Nanney et al., 1984). In these studies, which used specific antibodies and/or 125I-EGF binding assays, EGFR levels diminished rapidly in each subsequent layer superficial to the basal layer, suggesting that EGFR is lost as epithelial cells undergo terminal differentiation. Nonepithelial cells such as

Epidermal growth factor (EGF) is a small polypeptide (6,045 Da molecular mass) originally isolated from the submandibular glands of male mice (Cohen, 1962). EGF stimulates both proliferation (Cohen, 1965) and differentiation (Cohen and Elliott, 1963) in epidermis as well as in a variety of epithelial and nonepithelial tissues (Carpenter and Cohen, 1979). Since the discovery of EGF, a family of structurally related growth factors has been identified that includes transforming growth factor α, heparin-binding EGF, cripto, heregulin and amphiregulin (Carpenter and Cohen, 1990; Prigent and Lemoine, 1992). This family of growth factors requires a cell surface receptor in order to express their effects. The EGF receptor (EGFR) is a 170 kDa protein that includes an external growth-factor-binding domain, a transmembrane domain, and an internal tyrosine kinase domain. EGFR is a member of a family of receptor tyrosine kinases that includes platelet-derived growth factor receptor, fibroblast growth factor receptor,

Key words: EGF receptor, stem cell, corneal epithelium

146

J. D. Zieske and M. Wasson

muscle cells (Olwin and Hauschka, 1988) and chondrocytes (Kinoshita et al., 1992) also appear to lose EGFR as they undergo terminal differentiation. Based on three lines of evidence, the number of EGFR present on a cell appears to be correlated with the rate of epithelial cell proliferation: (1) in rat epidermis the number of EGFR decreases with age of the animal, which correlates with the rate of epidermal cell proliferation (Green et al., 1983); (2) the number of EGFR is related to the rate of proliferation of normal and transformed keratinocytes in cell culture (Boonstra et al., 1985); and (3) many kinds of carcinoma express high levels of EGFR (Mukaida et al., 1991; Yonemura et al., 1992). Stratified squamous epithelium exhibits rapid and continuous turnover of the terminally differentiated cells throughout the life span of the animal. These cells are replaced through proliferation of a group of cells known as stem cells (Potten, 1976; Cotsarelis et al., 1989; Hall and Watt, 1989). A stem cell can be most simply defined as any cell with a high capacity for self-renewal extending throughout adult life (Hall and Watt, 1989). Along with their high proliferative potential, stem cells apparently have a long cell cycle and divide infrequently. Stem cells are also thought to be capable of asymmetric cell division, giving rise to a daughter cell that is somewhat more differentiated. These daughter cells, termed “transit” (Potten, 1976) or “transient” (Cotsarelis et al., 1989) amplifying (TA) cells, have a limited proliferative potential and a cell cycle shorter than that of stem cells. Following an undefined signal, the TA cells lose their ability to undergo mitosis and enter the pathway toward terminal differentiation. Misumi and Akiyoshi (1990, 1991) have reported that EGFR is concentrated in the nonserrated cells of the human epidermis. These cells are thought to be the epidermal stem cells (Lavker and Sun, 1982, 1983). In the current study, we localized EGFR in the rat cornea to examine whether EGFR levels showed regional variation among stem, TA and terminally differentiated cells. The cornea is an ideal model in which to study this question, since the corneal epithelial stem cells are present only in the limbus, the transition zone between corneal and conjunctival epithelium (Schermer et al., 1986; Cotsarelis et al., 1989; Zieske et al., 1992a). In addition, we examined EGFR localization during corneal epithelial development between birth and the time of eyelid opening. We have previously shown that stemlike cells are present across the entire cornea prior to eyelid opening (Chung et al., 1992). The results of the current study suggest that high levels of EGFR are present in corneal epithelial stem cells and that this level consistently diminishes as the cells progress through the pathway of terminal differentiation from stem to early TA to terminally differentiated cells.

MATERIALS AND METHODS Antibodies Antibody against EGFR was obtained from Upstate Biotechnology (Lake Placid, NY). The antibody was prepared against recombinant human EGFR fusion protein in sheep. According to the

manufacturer, the polyclonal antibody shows wide species crossreactivity including human, rat and chick. We have previously reported (Zieske et al., 1992a,b) on the production of a monoclonal antibody specific to α-enolase, a glycolytic enzyme that catalyzes the conversion of 2-phosphoglycerate to phosphoenolpyruvate. This antibody (4G10.3) reacts specifically with basal cells in the limbal portion of the corneal epithelium (Zieske et al., 1992b) and was used in the current study as a marker of these cells.

Immunofluorescence microscopy Adult Sprague-Dawley rats and neonates aged 1, 4, 7, 10, 12 and 14 days, and 3 and 4 weeks, were used. Eyelid opening occurred between 12 and 15 days of age. The rats were killed by an intraperitoneal injection of a lethal dose of sodium pentobarbital and their eyeballs enucleated. Corneas were excised and frozen in Tissue Tek II OCT compound (Lab Tek Products, Naperville, IL). Cryostat sections (6 µm) were placed on gelatin-coated slides, airdried overnight at 37°C, rehydrated in PBS, and blocked in 1% bovine serum albumin (BSA) for 10 minutes. Polyclonal antibody against EGFR or monoclonal antibody against α-enolase was placed on the slides, which were then incubated for 1 hour in a moist chamber. The slides were then rinsed for 10 minutes in PBS with 1% BSA. FITC-conjugated donkey anti-sheep IgG or goat anti-mouse IgG (Jackson Immuno Research, West Grove, PA) was applied for 1 hour in a moist chamber. Coverslips were mounted with a medium consisting of PBS, glycerol and paraphenylene diamine. Negative control tissue sections (primary antibody omitted) were routinely run with every antibody binding experiment. Controls were also done with unrelated monoclonal and polyclonal antibodies to ensure specificity. The sections were viewed and photographed using a Zeiss Axiophot (Thornwood, NY) equipped for epi-illumination.

Autoradiography Slices (1 mm thick) of adult corneas were incubated with 125IEGF (4 nM) (ICN Biomedicals, Costa Mesa, CA) for 90 minutes at 37°C in minimum essential medium (MEM) containing 1% BSA and processed for autoradiography using Kodak NTB-3 emulsion (IBI, New Haven, CT). To control for nonspecific binding, corneas were incubated in the presence of a 100-fold excess of unlabelled human recombinant EGF (Upstate Biotechnology). Sections were exposed at 4°C for 2-4 weeks, developed with Kodak D-19 (IBI), and counterstained with hematoxylin and eosin (Green et al., 1983; Gusterson et al., 1984; Nanney et al., 1984). To quantify the relative levels of 125I-EGF binding to various regions of the cornea, silver grains were counted following autoradiography in sections exposed to emulsion for 2 weeks. Grains were counted in four regions: (1) limbus, the area immediately above limbal blood vessels; (2) periphery, the area immediately adjacent to limbus; (3) mid-cornea, the area halfway between the limbus and the center of the cornea; and (4) central, the center of the cornea. At least four sections in each of four corneas were examined, with the number of grains present in 40 cells per region being determined. The numbers of cells per region were analyzed for statistically significant differences using Student’s t-test.

Electrophoresis and immunoblotting To harvest central versus limbal and peripheral epithelium, a 4 mm trephine was used to demarcate central cornea in adult rats. A small scalpel was then used to remove the epithelium inside the central 4 mm zone (central region). Epithelium outside the 4 mm central zone was harvested similarly (peripheral region). Epithelium was harvested from rat corneas prior to eyelid opening in an identical way except that the eyelids were surgically opened and a 3 mm trephine was used to demarcate the central

Distribution of EGF receptor area. Epithelial samples were solubilized in 1% SDS and protein amounts determined using the Pierce (Rockford, IL) BCA protein assay. Epithelial protein samples were analyzed by SDS-PAGE and transferred to polyvinylidene difluoride paper (Millipore, Bedford, MA) as described by Towbin et al. (1979). The paper was then placed in a blocking solution containing 5% BSA in 0.1% Tween20 in Tris-buffered saline (100 mM Tris-HCl, 0.9% NaCl, pH 7.5; TTBS) and incubated for 30 minutes at 37°C. The paper was washed and then incubated for 30 minutes in anti-EGFR in TTBS at a 1:30 dilution at room temperature. Antibody binding was detected with the Vectastain ABC Kit (Vector Laboratories, Burlingame, CA). Color was detected using 0.05% diaminobenzidine, 0.04% NiCl2, and 0.015% H2O2. The relative amount of EGFR present was semiquantitated using laser densitometry (Molecular Dynamics Computing Densitometer, Model 300A, Sunnyvale, CA). Since purified EGFR was not available for use as a standard, only relative amounts of EGFR could be determined.

RESULTS Corneal epithelium is a 5- to 7-layered, stratified squamous epithelium with several unique characteristics related to

147

maintaining transparency. The corneal epithelium is flat with no papillary structures. In addition, the epithelium is remote from a capillary network and grows over an avascular stroma. The nearest vascular source is in the limbus, which is a transition zone between the cornea and the conjunctiva. Limbal epithelium, which is thought to contain the corneal epithelial stem cells, is adjacent to the limbal blood vessels. The characteristic immunofluorescence staining pattern of anti-EGFR in adult corneal epithelium is seen in Fig. 1. EGFR is concentrated in the basal cell layer of the limbal epithelium with the labeling intensity in corneal basal cells decreasing progressively with the distance from the limbal region (Fig. 1C,D). In all areas, EGFR was found primarily in cells of the basal layer; weak labeling extended into the suprabasal cell layers. The immunoreactive distribution of EGFR appeared to be primarily membranous, although cytoplasmic localization adjacent to the membrane cannot be distinguished at these magnifications. When expression of EGFR was compared with α-enolase, a marker of limbal basal cells, we observed that cells that expressed high levels of α-enolase also expressed high levels of EGFR (Fig. 2). Both EGFR and α-enolase

Fig. 1. Immunolocalization of EGFR in adult rat cornea. (A) Low magnification shows intense labeling of anti-EGFR in limbal (l) region, with labeling diminishing in the immediately adjacent peripheral (p) region and diminishing further in central (c) cornea. (B) Higher magnification of limbal region shows intense membranous labeling of anti-EGFR. Note the absence or low levels of immunoreactive EGFR on the basal membrane of the basal cells. (C) Higher magnification of peripheral region shows reduced labeling of anti-EGFR. (D) Higher magnification of central cornea shows greatly reduced level of immunoreaction of anti-EGFR. Bars, 50 µm.

148


Fig. 2. Colocalization of EGFR and α-enolase. Adjacent sections show localization of (A) EGFR and (B) α-enolase in limbal region. Bar, 50 µm.

Fig. 3. Autoradiography of 125I-EGF. (A) Dark-field autoradiograph shows 125I-EGF binding to adult rat cornea. High levels of silver grains in limbal epithelial region diminish toward central cornea. (B) Bright-field micrograph of same area as in (A). (C) Dark-field autoradiograph shows 125I-EGF binding to adult rat cornea following incubation with 100-fold excess of unlabeled EGF. Note low levels of silver grains over epithelium. (D) Bright-field micrograph of same area as in (C). Bar, 50 µm; arrows indicate direction of central cornea.

appeared to be concentrated in the basal cell layer of the limbal epithelium. Since EGFR could potentially be masked in a manner to prevent immunoreaction with anti-EGFR, autoradiography

using 125I-EGF was done to localize active EGFR. We found that 125I-EGF binding was concentrated in the limbal and peripheral zones (Fig. 3). High levels of silver grains were present in the limbal region and diminished toward

Distribution of EGF receptor

SILVER GRAINS/ 10 BASAL CELLS

300

149

DISCUSSION

200

100

0 LIMBUS PERIPHERY

MID

CENTRAL

Fig. 4. Number of silver grains localized over basal cells in four areas of the cornea following 125I-EGF autoradiography. Bars indicate mean ± s.d. of counts of four sections of each of four corneas.

the central cornea (Fig. 3A,B). Incubation of corneas with 125I-EGF with a 100-fold excess of unlabeled EGF removed specific binding to the epithelium (Fig. 3C,D). Semiquantitation of EGFR levels yielded 203.2 ± 48.6, 166.9 ± 24.9, 88.8 ± 11.1 and 45.2 ± 10.0 (mean ± s.d.) silver grains per 10 cells, respectively (Fig. 4). The numbers in the mid and central areas were significantly less (P ≤ 0.001) than the numbers in the peripheral and limbal areas. The number of grains found over basal cells in the control sections, which had been incubated with a 100-fold excess of unlabeled EGF, was less than one per cell. EGFR was also immunolocalized in developing epithelium (Fig. 5). In the rat, corneal epithelium undergoes major morphological changes in the period between birth and eyelid opening. At birth, the corneal epithelium consists of one or two cell layers. Between 7 and 10 days of age, the epithelium begins to stratify; approximately 4-5 cell layers are present at the time of eyelid opening and 5-7 layers are present in the adult. Anti-EGFR labeled all cells across the cornea at 1, 4 and 7 days of age (Fig. 5A-D) with the intensity of labeling appearing to be maximal at 7 days. As the epithelium stratifies, immunoreaction of anti-EGFR rapidly diminished in the central cornea, taking on the adult-like pattern of EGFR expression (Fig. 5E,F). By 4 weeks of age, EGFR expression could not be distinguished from the adult pattern (data not shown). High levels of labeling of antiEGFR were seen in the limbal and peripheral regions of all developing corneas examined regardless of the age of the rat (Fig. 5G,H). Finally, EGFR levels were semiquantitated using immunoblotting followed by laser densitometry. In all samples tested, anti-EGFR reacted with a single band of approximately 170 kDa (Fig. 6). Epithelium harvested from the peripheral and limbal regions of developing corneas contained the highest levels of EGFR. In a representative experiment, relative levels for developing periphery:developing central:adult periphery:adult central were 100:69:62:13.

EGFR was originally discovered through its ability to stimulate epidermal keratinization and has since been shown to stimulate proliferation in a variety of epithelial tissues (Cohen, 1962; Carpenter and Cohen, 1979). EGF’s activity is regulated through a cell surface receptor that is localized to the proliferative cell layers in stratified squamous epithelium (Green et al., 1983; Gusterson et al., 1984; Nanney et al., 1984). Using fetal palmar skin, Misumi and Akiyoshi (1991) demonstrated that EGFR is expressed heterogeneously in the basal cell layer of developing epidermis. Their data suggest that levels of EGFR may be related to the degree of differentiation of the basal cells. In the current study, we used corneal epithelium as a model to examine whether EGFR levels are altered as stratified squamous epithelial cells differentiate along the pathway of stem cell to TA cell to terminally differentiated cell. Corneal epithelium is an ideal model in which to study this question in that the epithelial stem cells are thought to reside in the limbus at the edge of the cornea (Schermer et al., 1986; Cotsarelis et al., 1989; Zieske et al., 1992a). In addition, epithelial cells are thought to migrate centripetally from the limbus toward the center of the cornea (Kinoshita et al., 1981; Buck, 1982; Thoft and Friend, 1983). The combination of these two phenomena provides a spatial arrangement that has stem cells present at the edge of the cornea and an increasing degree of maturation of basal cells towards the center of the cornea. Our present results indicate that high levels of EGFR are present in the stem and early TA cells and that this level rapidly drops during maturation. Autoradiographic data suggest that 4.5-fold more EGFR are present in limbal than in central corneal epithelial basal cells, while immunoblot dot data indicate that limbal and peripheral corneal epithelium contains 4.8-fold more EGFR than does central corneal epithelium in the adult rat. These data suggest that stem cells in adult corneal epithelium express high levels of EGFR, which may inhibit differentiation by signaling the cells to maintain their proliferative potential rather than enter the pathway of terminal differentiation. EGF has been known for many years to stimulate the proliferation of corneal epithelial cells in culture (Savage and Cohen, 1973) and to stimulate the rate of epithelial migration by means of an increase in cell number during wound healing (Savage and Cohen, 1973; Frati et al., 1972; Soong et al., 1989; Kitazawa et al., 1990). While EGFR has been shown to be present in isolated (Frati et al., 1972) and cultured (Hongo et al., 1992) corneal epithelial cells, it has not, to our knowledge, been localized to specific cell layers or regions of the cornea. The localization of EGFR in the basal cell layer across the cornea, with a concentration in the periphery and limbal cells, is consistent with several observations concerning the effects of EGF on corneal epithelial proliferation and wound repair. Savage and Cohen (1973) found that topical application of EGF had no morphological effect on epithelium in unwounded corneas. However, following wounding, EGF decreased the time required for wound closure and stimulated hyperplasia. This hyperplasia was lost following epithelial wound closure. These data, in conjunction with our localization of

150


Fig. 5. Immunolocalization of EGFR in developing-rat corneas at (A) 1, (B) 4, and (C) 7 days of age. (D) Phase-contrast micrograph of area shown in (B). Note one or two layers of epithelial cells. (E) Central cornea, 10 days of age. (F) Phase-contrast micrograph of area shown in (E). Note multilayered epithelium. (G) Limbal region, 10 days of age. Basal cells maintain intense anti-EGFR immunoreaction. (H) Phase-contrast micrograph of area shown in (G). Bar, 50 µm.

Distribution of EGF receptor

Fig. 6. Immunoreaction of antiEGFR with: (A) 80 µg of adult peripheral and limbal corneal epithelium; (B) 80 µg of adult central corneal epithelium; (C) 160 µg of developing peripheral and limbal corneal epithelium; and (D) 160 µg of developing central corneal epithelium. Developing epithelium is pooled tissue harvested from corneas of animals prior to eyelid opening (4-12 days of age). Molecular mass markers are indicated at right in kDa.

EGFR, suggest that topically applied EGF does not penetrate to the basal cell layer unless the epithelial integrity has been broached. Our data showing EGFR concentration in the limbal zone is also consistent with the findings that, following epithelial wounding, the peripheral and limbal cells preferentially undergo mitosis (Cotsarelis et al., 1989; Kitazawa et al., 1990) and that treatment with EGF results in hyperplasia in the region adjacent to the limbus (Savage and Cohen, 1973). It is also noteworthy that ocular surface neoplasms are almost entirely associated with the limbal region (Waring et al., 1984). This provides provocative correlation with the finding that members of the EGFR family are often amplified or overexpressed in human cancer (Mukaida et al., 1991; Yonemura et al., 1992). We have previously demonstrated that the glycolytic enzyme α-enolase is restricted to the limbal basal cells of the cornea and that, following wounding, the number of cells expressing α-enolase approximately doubles, suggesting that α-enolase is a marker of stem and early TA cells (Zieske et al., 1992a,b). In the present study, we demonstrated that cells expressing high levels of α-enolase also express high levels of EGFR. It is tempting to speculate that the expression of these proteins might be linked, since EGF has been shown to stimulate the synthesis of α-enolase in fibroblasts (Matrisian et al., 1985; Nikaido et al., 1991). This linkage might involve c-myc in that EGF, acting presumably through EGFR, stimulates c-myc protein production, which in turn appears to stimulate α-enolase synthesis (Warwar et al., 1992). In the second aspect of our study, we investigated the localization of EGFR in the neonatal rat corneal epithelium. In the rat, a major portion of corneal epithelial development occurs between birth and the time of eyelid opening. We have previously shown that α-enolase is expressed at high levels in all cells of the developing corneal epithelium until approximately 1 week of age (Chung et al., 1992), suggesting that all cells are in a stem or stem-like state. As the epithelium begins terminally differentiating and undergoes stratification, α-enolase expression is lost in epithelial cells of the central cornea, becoming sequestered in the basal cells of the limbal epithelium. This loss of expression is concurrent with a change in cell shape from flattened

151

ovoid to columnar. In the current study, we observed that expression of EGFR underwent similar changes during development, with expression of EGFR in epithelial cells extending across the cornea in the undeveloped, 1- to 2cell layer thick epithelium. As the epithelium begins to stratify, EGFR levels drop rapidly. This decrease in EGFR level is concurrent with a surge in proliferation as the number of cell layers increases from 1.6 at day 7 to 4.7 after eyelid opening (day 12-14) (Watanabe et al., 1993). These data suggest that the developing corneal epithelial cells are in a stem or early TA state, allowing them to proliferate rapidly just before eyelid opening, providing the eye with a protective layer of epithelium as it becomes exposed to the external environment. In addition, the data suggest that following this surge of proliferation, the basal cells in the central cornea are no longer in the stem or early TA mode but have matured into late TA and terminally differentiated cells. In summary, we have shown that EGFR is localized in all cells of the developing corneal epithelium and in the stem cells of the adult rat corneal epithelium. These findings are consistent with the hypothesis that stem cells of corneal epithelium express high levels of growth factor receptors to help them maintain their proliferative potential, and ‘protecting’ them from entering the pathway of terminal differentiation. EGFR is one of many growth factor receptors that may play pivotal roles in stem cell proliferation, and we are currently investigating the distribution of other growth factor receptors. We thank Ilene Gipson for critically reading the manuscript, Patricia Pearson for preparing cryostat sections, Gale Unger for typing and proofreading the manuscript, and Peter Mallen for photographic assistance in printing the figures. This work was supported by NIH grant R01-EY05665 to James D. Zieske.

REFERENCES Boonstra, J., De Laat, S. W. and Ponec, M. (1985). Epidermal growth factor receptor expression related to differentiation capacity in normal and transformed keratinocytes. Exp. Cell Res. 161, 421-433. Buck, R. C. (1982). Hemidesmosomes of normal and regenerating mouse corneal epithelium. Virchow. Arch. [B] Cell Pathol. 41, 1-16. Carpenter, G. and Cohen, S. (1979). Epidermal growth factor. Annu. Rev. Biochem. 48, 193-216. Carpenter, G. and Cohen, S. (1990). Epidermal growth factor. J. Biol. Chem. 265, 7709-7712. Chung, E.-H., Bukusoglu, G. and Zieske, J. D. (1992). Localization of corneal epithelial stem cells in the developing rat. Invest. Ophthalmol. Vis. Sci. 33, 2199-2206. Cohen, S. (1962). Isolation of a mouse submaxillary gland protein accelerating incisor eruption and eyelid opening in the new-born animal. J. Biol. Chem. 237, 1555-1562. Cohen, S. (1965). The stimulation of epidermal proliferation by a specific protein (EGF). Dev. Biol. 12, 394-407. Cohen, S. and Elliott, G. A. (1963). The stimulation of epidermal keratinization by a protein isolated from the submaxillary gland of the mouse. J. Invest. Dermatol. 40, 1-5. Cotsarelis, G., Cheng, S.-Z., Dong, G., Sun, T.-T. and Lavker, R. M. (1989). Existence of slow-cycling limbal epithelial basal cells that can be preferentially stimulated to proliferate: implications on epithelial stem cells. Cell 57, 201-209. Fantl, W. J., Escobedo, J. A., Martin, G. A., Turck, C. W., del Rosario, M., McCormick, F. and Williams, L. T. (1992). Distinct phosphotyrosines on a growth factor receptor bind to specific molecules that mediate different signaling pathways. Cell 69, 413-423.

152


Frati, L., Daniele, S., Delogu, A. and Covelli, I. (1972). Selective binding of the epidermal growth factor and its specific effects on the epithelial cells of the cornea. Exp. Eye Res. 14, 135-141. Green, M. R., Basketter, D. A., Couchman, J. R. and Rees, D. A. (1983). Distribution and number of epidermal growth factor receptors in skin is related to epithelial cell growth. Dev. Biol. 100, 506-512. Gusterson, B., Cowley, G., Smith, J. A. and Ozanne, B. (1984). Cellular localization of human epidermal growth factor receptor. Cell Biol. Int. Rep. 8, 649-658. Hall, P. A. and Watt, F. M. (1989). Stem cells: the generation and maintenance of cellular diversity. Development 106, 619-633. Hongo, M., Itoi, M., Yamaguchi, N. and Imanishi, J. (1992). Distribution of epidermal growth factor (EGF) receptors in rabbit corneal epithelial cells, keratocytes and endothelial cells, and the changes induced by transforming growth factor-β1. Exp. Eye Res. 54, 9-16. Kashishian, A., Kazlauskas, A. and Cooper, J. A. (1992). Phosphorylation sites in the PDGF receptor with different specificities for binding GAP and PI3 kinase in vivo. EMBO J. 11, 1373-1382. Kinoshita, S., Friend, J. and Thoft, R. A. (1981). Sex chromatin of donor corneal epithelium in rabbits. Invest. Ophthalmol. Vis. Sci. 21, 434441. Kinoshita, A., Takigawa, M. and Suzuki, F. (1992). Demonstration of receptors for epidermal growth factor on cultured rabbit chondrocytes and regulation of their expression by various growth and differentiation factors. Biochem. Biophys. Res. Commun. 183, 14-20. Kitazawa, T., Kinoshita, S., Fujita, K., Araki, K., Watanabe, H., Ohashi, Y. and Manabe, R. (1990). The mechanism of accelerated corneal epithelial healing by human epidermal growth factor. Invest. Ophthalmol. Vis. Sci. 31, 1773-1778. Lavker, R. M. and Sun, T.-T. (1982). Heterogeneity in epidermal basal keratinocytes: morphological and functional correlations. Science. 215, 1239-1241. Lavker, R. M. and Sun, T.-T. (1983). Epidermal stem cells. J. Invest. Dermatol. 81 (1 Suppl.), 121s-127s. Matrisian, L. M., Rautmann, G., Magun, B. E. and Breathnach, R. (1985). Epidermal growth factor or serum stimulation of rat fibroblasts induces an elevation in mRNA levels for lactate dehydrogenase and other glycolytic enzymes. Nucl. Acids Res. 13, 711-726. Misumi, Y. and Akiyoshi, T. (1990). Consistently nonoverlapping distribution of epidermal growth factor receptors in adult human skin detected by various monoclonal antibodies. Acta Anat. 137, 202-207. Misumi, Y. and Akiyoshi, T. (1991). Epidermal ridge formation in the human fetus: a correlation to the appearance of basal cell heterogeneity and the expression of epidermal growth factor receptor and cytokeratin polypeptides in the epidermis. Amer. J. Anat. 191, 419-428. Mukaida, H., Masakazu, T., Hirai, T., Yamashita, Y. and Toge, T. (1991). Clinical significance of the expression of epidermal growth factor and its receptor in esophageal cancer. Cancer 68, 142-148. Nanney, L. B., McKanna, J. A., Stoscheck, C. M., Carpenter, G. and King, L. E. (1984). Visualization of epidermal growth factor receptors in human epidermis. J. Invest. Dermatol. 82, 165-169. Nikaido, T., Bradley, D. W. and Pardee, A. B. (1991). Molecular cloning

of transcripts that accumulate during the late G1 phase in cultured mouse cells. Exp. Cell Res. 192, 102-109. Olwin, B. B. and Hauschka, S. D. (1988). Cell surface fibroblast growth factor and epidermal growth factor receptors are permanently lost during skeletal muscle terminal differentiation in culture. J. Cell Biol. 107, 761769. Pawson, T. and Gish, G. D. (1992). SH2 and SH3 domains: from structure to function. Cell 71, 359-362. Potten, C. S. (1976). Identification of clonogenic cells in the epidermis and the structural arrangement of the epidermal proliferative unit (EPU). In Stem Cells of Renewing Cell Populations (ed. P. K. Lala and D. G. Osmond), pp. 91-102. Academic Press, New York. Prigent, S. A. and Lemoine, N. R. (1992). The type 1 (EGFR-related) family of growth factor receptors and their ligands. Prog. Growth Factor Res. 4, 1-24. Savage, C. R. Jr and Cohen, S. (1973). Proliferation of corneal epithelium induced by epidermal growth factor. Exp. Eye Res. 15, 361-366. Schermer, A., Galvin, S. and Sun, T.-T. (1986). 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. 103, 49-62. Soong, H. K., McClenic, B., Varani, J., Hassan, T., Huang, S. C. and Brenz, R. (1989). EGF does not enhance corneal epithelial cell motility. Invest. Ophthalmol. Vis. Sci.30, 1808-1812. Thoft, R. A. and Friend, J. (1983). The X, Y, Z hypothesis of corneal epithelial maintenance. Invest. Ophthalmol. Vis. Sci. 24, 1442-1443. Towbin, H., Staehelin, T. and Gordon, J. (1979). Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc. Nat. Acad. Sci. USA 76, 4350-4354. Waring, G. O. III, Roth, A. M. and Ekins, M. B. (1984). Clinical and pathologic description of 17 cases of corneal intraepithelial neoplasia. Amer. J. Ophthalmol. 97, 547-559. Warwar, R. E., Kim, R. Y., Wistow, G. J. and Zelenka, P. S. (1992). The tau-crystallin/alpha-enolase gene: a candidate for regulation by C-MYC. Invest. Ophthalmol. Vis. Sci. 33(Suppl):794. Watanabe, H., Tisdale, A. S. and Gipson, I. K. (1993). Eyelid opening induces expression of a glycocalyx glycoprotein of rat ocular surface epithelium. Invest. Ophthalmol. Vis. Sci. (in press). Yarden, Y. and Ullrich, A. (1988). Growth factor receptor tyrosine kinases. Annu. Rev. Biochem. 57, 443-478. Yonemura, Y., Takamura, H., Ninomiya, I., Fushida, S., Tsugawa, K., Kaji, M., Nakai, Y., Ohoyama, S., Yamaguchi, A. and Miyazaki, I. (1992). Interrelationship between transforming growth factor-α and epidermal growth factor receptor in advanced gastric cancer. Oncology 49, 157-161. Zieske, J. D., Bukusoglu, G. and Yankauckas, M. A. (1992b). Characterization of a potential marker of corneal epithelial stem cells. Invest. Ophthalmol. Vis. Sci.33, 143-152. Zieske, J. D., Bukusoglu, G., Yankauckas, M. A., Wasson, M. E. and Keutmann, H. T. (1992a). α-Enolase is restricted to basal cells of stratified squamous epithelium. Dev. Biol. 151, 18-26. (Received 15 March 1993 - Accepted 20 May 1993)