Lacrimal Gland HGF, KGF, and EGF mrna Levels Increase after Corneal Epithelial Wounding

Transcription

1 Lacrimal Gland HGF, KGF, and EGF mrna Levels Increase after Corneal Epithelial Wounding Steven E. Wilson, 1,2 Qianwa Liang, 2 and Woo Jung Kim 2,3 PURPOSE. To evaluate the effect of corneal epithelial wounding on lacrimal gland expression of hepatocyte growth factor (HGF), keratinocyte growth factor (KGF), and epidermal growth factor (EGF) in the rabbit model. METHODS. Rabbits had corneal epithelial scrape injuries, and the lacrimal gland was removed at different times after wounding. HGF, KGF, and EGF mrna expression was examined by quantitative RNase protection assay. HGF, KGF, and EGF proteins were detected in rabbit lacrimal tissue using immunoprecipitation and western blot analysis. RESULTS. HGF mrna and EGF mrna were significantly increased in rabbit lacrimal gland tissue within 8 hours after corneal epithelial injury. The increase in KGF mrna expression was small and reached significance 1 day after corneal injury. Lacrimal gland expression peaked at 3 days after wounding for each growth factor mrna, the same day, on average, that the epithelial defect healed. After the peak increase in expression, there was a progressive decline in expression of each growth factor mrna, but production was still increased compared with prewound levels. HGF protein, KGF protein, and EGF proteins were detected in rabbit lacrimal gland tissue. CONCLUSIONS. Levels of HGF, KGF, and EGF mrnas increase in rabbit lacrimal gland tissue in response to corneal epithelial wounding. The results of this study are consistent with the existence of a cornea nervous system lacrimal gland regulatory loop modulating expression of these growth factor mrnas. The lacrimal gland is a likely source of increased HGF and EGF proteins detected in tears in previous studies. (Invest Ophthalmol Vis Sci. 1999;40: ) A number of studies performed over the past few years have suggested that the lacrimal gland produces growth factors that have a role in lacrimal gland physiology and maintenance of the ocular surface Some of these investigations have noted increases in either tear growth factor levels or lacrimal gland growth factor production during the wound-healing response to ocular surface injury. 7,15,17 For example, epidermal growth factor (EGF) expression in lacrimal gland has been shown to increase in response to corneal wounding. 7 Cholinergic stimulation of the lacrimal gland increases EGF secretion, suggesting that upregulation of EGF in the lacrimal gland and tears in response to corneal epithelial wounding is controlled by a nervous system mediated reflex arc. 6,7 Corneal wounding also upregulates tumor necrosis factor (TNF)- production by the lacrimal gland. 15 From the 1 Department of Ophthalmology, University of Washington School of Medicine, Seattle; the 2 Eye Institute and Department of Cell Biology, The Cleveland Clinic Foundation, Ohio; and the 3 Department of Ophthalmology, Sungkyunkwan University School of Medicine, Seoul, Korea. Supported in part by US Public Health Service Grant EY10056 from the National Eye Institute, National Institutes of Health; The University of Washington School of Medicine, Seattle; The Cleveland Clinic Foundation, Ohio; and an Unrestricted Grant from Research to Prevent Blindness. Submitted for publication December 2, 1998; revised April 14, 1999; accepted May 20, Proprietary interest category: N. Corresponding author: Steven E. Wilson, Department of Ophthalmology, University of Washington School of Medicine, Box , Seattle, WA sewilson@u.washington.edu HGF bioavailability in tears is also increased in response to corneal injury. 17 Although HGF protein is probably produced by lacrimal gland interalveolar connective tissue cells or myoepithelial cells, 13 it is not known whether HGF production in the lacrimal gland is upregulated in response to corneal injury, because both the lacrimal gland and keratocytes could be sources of increased tear HGF. 13,17 Nothing is known about KGF production in the lacrimal gland or the presence of KGF in tears. HGF, KGF, and EGF are growth factors that are thought to be important in modulating corneal epithelial homeostasis and wound healing. HGF and EGF have both been shown to stimulate corneal epithelial cell proliferation and migration while inhibiting corneal epithelial cell terminal differentiation. 18,19 KGF has also been found to trigger corneal epithelial cell proliferation, but no KGF effects on migration or terminal differentiation of these cells have been noted. 18,19 The present study was performed to examine lacrimal gland levels of HGF, KGF, and EGF mrnas after corneal epithelial wounding using a sensitive quantitative technique. We also examined expression of HGF, KGF, and EGF proteins in rabbit lacrimal gland tissue using immunoprecipitation and western blot analysis. METHODS Animals and Tissues Six-week-old New Zealand White rabbits were anesthetized with, ketamine (30 mg/kg weight intramuscular) and xylazine (2 mg/kg weight intramuscular). Corneal epithelial scrape Investigative Ophthalmology & Visual Science, September 1999, Vol. 40, No. 10 Copyright Association for Research in Vision and Ophthalmology 2185

2 2186 Wilson et al. IOVS, September 1999, Vol. 40, No. 10 TABLE 1. PCR Primers Effective for Generating Rabbit cdna Sequences used in RNAse Protection Assay Modulator Size (bp) Upstream Primer Downstream Primer Accession Number Actin 350 AGGCCAACCGCGAGAAGATGACC GAAGTCCAGGGCGACGTAGCAC X00351 HGF 177 GAGTACTGTGCAATTAAAACATGC CATGTCATGCTTGTGAGGATA X16323 KGF 322 GGGGATATAAGGGTGAGAAGA GATTTAAGGCAACGAACATTT Z22703 EGF 394 TATGTCTGCCGGTGCTCAGAA AGCGTGGCGCAGTTCCCACCA X04571 Rabbit sequences were generated using PCR primers designed from the human or mouse sequence using rabbit liver cdna as target. All sequences were confirmed by DNA sequencing. wounds were produced in one eye selected at random or in neither eye in the control animals. Wounding was performed using a scalpel blade to remove epithelium from the central cornea sparing approximately 1 mm at the limbus for 360. Three drops of ciprofloxacin (Ciloxan; Alcon, Fort Worth, TX) were instilled immediately after epithelial scrape. Tape was applied to close the lid for 2 hours to prevent exposure while the animal was anesthetized. For the welfare of animals, only one eye of each rabbit was used in this study. If animals showed signs of having pain on awakening after corneal scrape injury, they were treated with oral acetaminophen, and patching was continued for another 12 to 24 hours. Animals were anesthetized again with ketamine and xylazine at the appropriate time after wounding, the lacrimal gland removed through the inferonasal conjunctival sack with forceps and sharp Wescott scissors, and the rabbit euthanatized with intravenous pentobarbital (100 mg/kg). The lacrimal tissue was used for RNA extraction or protein isolation. All animals were treated in accordance with the tenets of the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. RNase Protection Assay Total cellular RNA was extracted from rabbit lacrimal gland tissue using TRIzol (Gibco BRL, Rockville, MD) according to the manufacturer s instructions. RNA was dissolved in diethyl procarbonate water, and the concentrations were determined using a spectrophotometer. A cdna probe for rabbit HGF, KGF, EGF, or -actin was amplified using polymerase chain reaction (PCR) with the primers listed in Table 1. A melting temperature of 98 C, an annealing temperature of 55 C, and an extension temperature of 72 C were used in PCR amplifications, which were continued for 35 cycles. The amplification products were cloned into the pcr2.1 TA clone vector (Invitrogen, San Diego, CA) and sequenced using standard methods to confirm that the sequence was the corresponding rabbit sequence for each cytokine or -actin. In some cases, sequences were subcloned into the Bluescript SK vector (Stratagene, San Diego, CA) so that the antisense RNA probe used in the RNase protection assay would be synthesized from the T7 promoter, whether the sequence was in the pcr2.1 TA or Bluescript vector. A 32 P-labeled RNA probe for each cytokine and -actin was prepared with an RNA transcription kit (Stratagene). The RNase protection assay was performed with a kit (Boehringer Mannheim, Indianapolis, IN), according to the manufacturer s protocol with the following modifications: Two microliters of RNA probe ( cpm/ l) and 50 g of rabbit lacrimal gland RNA (100 g RNA was used for the KGF experiment) were coprecipitated in the presence of 0.3 M sodium acetate with ice-cold ethanol. -Actin RNA probe ( cpm/ l) was included with each growth factor RNA as a control for quantitation. RNA was recovered by centrifugation at 15,000 rpm for 15 minutes at 4 C in a microcentrifuge and dissolved in 30 l hybridization buffer. After denaturation for 5 minutes at 90 C, the samples were incubated overnight at 42 C in an oven. Each hybridization mixture was digested with 40 units of RNase T1 and 10 units of RNase T2 for 50 minutes at 30 C, then digested with 3 l proteinase K (20 g/ l) in the presence of 0.5% sodium dodecyl sulfate (SDS) for 20 minutes at 37 C. The protected RNA fragments were precipitated by adding 5 g yeast trna and 1 ml ethanol and then extracted with 400 l phenol-chloroform (1:1). The RNA pellet was resuspended in 7 l loading buffer, heated for 5 minutes at 90 C, and subjected to electrophoresis in a 4% polyacrylamide 7 M urea gel at 300 V in 1 TBE (89 mm Tris-HCl, 89 mm sodium borate, 2 mm EDTA) buffer. The gel was fixed with 10% acetic acid and 10% methanol and dried with a vacuum gel dryer (Bio-Rad, Richmond, CA). Dried gels were exposed overnight to film (i; Eastman Kodak, Rochester, NY). Unless otherwise specified, all reagents were obtained from Sigma (St. Louis, MO). The actual sizes of the protected RNA fragments were confirmed using 32 P-labeled RNA ladder that was included on each gel. This ladder was prepared by using the full-length transcripts from the cloned cdna templates for each of the growth factors and -actin. The full-length RNA probe sizes (not protected sequence size) were 320 (HGF), 460 ( -actin), 480 (KGF), and 550 (EGF) nucleotides. These markers were revealed with a brief exposure, and then the marker lane was cut from the dried gel to prevent overexposure of adjacent RNase-protected lanes. Quantitation of bands was performed by scanning the gels (Photoshop 4.0; Adobe, Seattle, WA) and using NIH Image 1.57 software to determine the density of both the growth factor and -actin band in each lane. All lanes were run on gels simultaneously for a particular RNase-protection experiment. A uniform box size that enclosed the largest band in a series for a particular growth factor or -actin was used for determining all the growth factor or -actin band densities in a particular experiment. The relative intensity of each band was calculated as density units (density of growth factor band/density of -actin band) 100. There were four different animals at each time point after wounding and in the control. Statistical analyses were performed with analysis of variance using the Bonferroni Dunn adjustment. P 0.05 was considered statistically significant.

3 IOVS, September 1999, Vol. 40, No. 10 Lacrimal Gland Growth Factors 2187 Immunoprecipitation and Western Blot Analysis Rabbit lacrimal glands were homogenized and lysed in 5 ml lysis buffer (50 mm Tris Cl [ph 8.0], 0.5% Triton X-100, 10% glycerol, 0.2 mm EDTA, 150 mm NaCl, 1 mm dithiothreitol, 0.5 mm phenylmethylsulfonyl fluoride, 3 g/ml aprotinin, 2 g/ml pepstatin, and 1 g/ml leupeptin) on ice for 20 minutes. The extracts were centrifuged at 15,000 rpm in a microcentrifuge for 10 minutes in 4 C. Supernatants were decanted into a fresh tube, and the protein concentration of each extract was determined with a protein assay kit (Bio-Rad). Five hundred micrograms of lysate at 1 mg/ml was incubated with preimmune serum (2.5 l) containing protein A Sepharose 6MB (Pharmacia, Piscataway, NJ) for 1 hour, and the lysate was clarified by brief centrifugation in a microcentrifuge at 15,000 rpm. The lysate was incubated with 10 g antibody and 50 l protein A Sepharose 6MB overnight at 4 C with continuous mild agitation. Sepharose beads were washed three times in cell lysis buffer, and the bound proteins were eluted in SDS gel loading buffer by boiling. SDS-PAGE was performed as previously described. 20 Immunoblot analysis was performed by the chemiluminescence system (ECL; Amersham, Arlington Heights, IL) based on the recommended protocols supplied by the manufacturer, with minor modification described previously. 21 The anti-hgf 1A3.1.2 (Genentech, South San Francisco, CA) and anti-kgf AF-251MA (R&D Systems, Minneapolis, MN) antibodies used for immunoprecipitation and western blot analysis have been previously shown to recognize the rabbit HGF and KGF antigens. 13,22 The anti-egf AB236-NA antibody (R&D Systems) is characterized as useful in immunoprecipitation and western blot analysis by the manufacturer. RESULTS Corneal epithelial wounding resulted in increased levels of HGF mrna, KGF mrna, and EGF mrna expression measured by RNase protection assay in the rabbit model (Figs. 1A, 1B, 1C). For HGF and KGF the increases in mrna levels became significant at 8 hours after wounding (Fig. 1D). For KGF mrna the increase was small and did not reach statistical significance until 1 day after wounding. There was a trend toward an increase in HGF mrna levels as early as 1 hour after wounding (Fig. 1D). For each growth factor the peak in mrna levels occurred at 3 days after wounding. Approximately 3 days are required for the epithelial defect to heal when injury is created by the method used in this study ( [SEM] days; S. E. Wilson, unpublished data, 1992). Each of the growth factor mrnas had decreased at 7 days after wounding compared with the peak at 3 days after epithelial injury (Fig. 1D). In each case, however, the day 7 levels of mrna were still significantly increased compared with those in the unwounded control corneas. Proteins near the expected size for HGF, KGF, and EGF were detected in lacrimal glands from rabbits with normal control corneas using immunoprecipitation and western blot analysis (Fig. 2). The antibodies used for detection of HGF and KGF have been shown to detect the rabbit proteins. 13 The detected proteins were 70 kda for HGF, 28 kda for KGF, and 17 kda for EGF. The immunoprecipitation and western blot analysis technique was not used to attempt to measure changes in growth factor protein production after corneal epithelial wound healing, because, at best, the technique is semiquantitative. DISCUSSION The results of this study demonstrate that HGF mrna, KGF mrna, and EGF mrna levels increased in the lacrimal gland after corneal epithelial injury. Thus, we conclude that there is a regulatory communication between the ocular surface and the lacrimal gland tissue that modulates the levels of the growth factor mrnas. The route of communication is thought to be through the trigeminal sensory nerves of the cornea to the brain stem and subsequently to the lacrimal gland through parasympathetic nerves. 7,15 Studies by Meneray et al. 23 have demonstrated that secretory granules accumulate in the acinar cells after trigeminal denervation, providing evidence that trigeminal sensory nerve impulses contribute to the pathways mediating secretory release of components from the lacrimal gland. Yoshino et al. 24 demonstrated that cholinergic stimulation of human lacrimal gland in vitro increases secretion of lactoferrin and EGF proteins. The results of these studies also support the hypothesis that the lacrimal gland responds to ocular surface stimulation by increasing secretion of growth factors and other tear components. These results for EGF confirm previous results of other investigators reported in abstract form at the annual meeting of the Association for Research in Vision and Ophthalmology. 7 No studies on changes in HGF or KGF production in the lacrimal gland that occur in response to corneal epithelial wounding had been reported before this study. Presumably, HGF, KGF, and EGF released from the lacrimal gland after epithelial injury to the cornea modulate corneal epithelial wound healing. The timing of increased expression relative to corneal epithelial wound closure supports this hypothesis. Basal HGF, KGF, and EGF released into the tears from the lacrimal gland may also have an important role in homeostasis of the ocular surface. 6 This study determined that the levels of HGF mrna, KGF mrna, and EGF mrna increased in lacrimal gland after wounding but could not distinguish between increased transcription of mrna and decreased degradation of mrna. One or both of these processes could be involved in increases in tissue mrna levels. In any case, it is likely that increased mrna levels for each of these growth factors resulted in increased protein production within the lacrimal gland. HGF, KGF, and EGF proteins were detected in rabbit lacrimal gland tissue, but we could not quantitate changes in protein expression accurately because of the semiquantitative nature of the immunoprecipitation and western blot analysis technique. However, previous studies have shown that there is increased HGF in tears after corneal wounding. 17 It is likely that this increased HGF in tears is, at least in part, attributable to increased lacrimal gland production of this growth factor. Studies are needed to examine tear EGF protein and KGF protein bioavailability in response to corneal injury. The HGF and KGF proteins detected in the rabbit lacrimal gland are similar in size to those previously detected in rabbit tissues with the same antibodies. 22 We know of no studies in which the size of rabbit EGF has been monitored. We detected only a 17-kDa protein in rabbit tissue, whereas the human EGF precursor is 120 kda in size. 25 Because this was the only protein band detected in the EGF immunoprecipitation and

4 2188 Wilson et al. IOVS, September 1999, Vol. 40, No. 10 FIGURE 1. RNase protection assay detection of changes in HGF mrna (A), KGF mrna (B), and EGF mrna (C) in rabbit lacrimal gland after corneal epithelial wounding. Levels of mrna were monitored in lacrimal glands of unwounded control animals and in lacrimal glands of animals at 1 hour, 8 hours, 1 day, 3 days, and 7 days after corneal epithelial scrape injury to the cornea. -Actin mrna was also monitored as a control for RNA loading in each lane. Sizes of the protected RNA sequences based on the base pair (BP) lengths of the cdna sequences (Table 1) used to generate probes are indicated. Note that -actin mrna levels were similar in most samples within a particular experiment. HGF mrna, KGF mrna, and EGF mrna levels in the lacrimal gland increased after corneal epithelial wounding. The levels appeared to peak at 3 days for each growth factor. (D) Quantitation of RNase protection assay results. Density units were calculated as density of the growth factor band in a lane divided by the -actin band in the same lane 100. For each time point the average density units and SE were calculated from the results from four different rabbits. * or **, Mean was significantly different from the prewound control; * and **, significantly different from each other. Note that results cannot be interpreted to show differences in levels between the different growth factor mrnas (for example, level of HGF mrna relative to KGF mrna). western blot analysis, this is probably the size of the rabbit EGF precursor protein. Our previous studies have demonstrated that HGF protein is expressed in interalveolar connective tissue cells or myoepithelial cells and that HGF receptor protein is expressed within acinar and ductal epithelial cells within human lacrimal gland tissue. 13 In the present study HGF mrna and protein were expressed in rabbit lacrimal tissue. Thus, in addition to being secreted into tears, HGF probably has localized effects on epithelial cell functions within the lacrimal gland. Nothing is known about which lacrimal epithelial cell functions are regulated by HGF. This is the first study to examine KGF expression in lacrimal tissue. Studies have yet to be performed evaluating KGF receptor expression in lacrimal tissue and it is therefore not known whether KGF regulates local functions in the lacrimal gland. HGF and KGF are classic paracrine mediators of stromal epithelial interactions that are produced by fibroblastic cells to

5 IOVS, September 1999, Vol. 40, No. 10 Lacrimal Gland Growth Factors 2189 regulate the functions of epithelial cells in many organs. For example, HGF and KGF are produced by keratocytes 13,18 20,26,27 in the cornea, and results of in vitro studies suggest that these cytokines regulate corneal epithelial cell proliferation, motility, and differentiation. 18,19 This study shows that the lacrimal gland is also a probable source of HGF and KGF involved in homeostasis of the corneal epithelium and the wound-healing response after corneal epithelial injury. Acknowledgment The authors thank Ralph Schwall, of Genentech, Inc., South San Francisco, California, for providing the anti-hgf antibody. References FIGURE 2. Immunoprecipitation and western blot analysis to detect HGF, KGF, and EGF protein in rabbit lacrimal gland tissue. Four lacrimal gland (LG) samples 1 4 from rabbits with normal unwounded corneas were used for each growth factor. HGF, KGF, and EGF proteins of approximately 70 kda, 28 kda, and 17 kda, respectively, were detected in the rabbit lacrimal gland tissue. Ab indicates antibody alone samples for each of the experiments. One band for KGF (doubleheaded arrow) was detected in the Ab-alone sample and was therefore an artifact related to the antibody solution. Ig-L and Ig-H indicate light and heavy chains, respectively, of the immunoglobulins used in immunoprecipitation and western blot analysis. 1. Ohashi Y, Motokura M, Kinoshita Y, et al. Presence of epidermal growth factor in human tears. Invest Ophthalmol Vis Sci. 1989; 30: van Setten GB, Viinikka L, Tervo T, Pesonen K, Tarkkanen A, Perheentupa J. Epidermal growth factor is a constant component of normal human tear fluid. Graefes Arch Ophthalmol. 1989;227: Kasayama S, Ohba Y, Oka T. Expression of the epidermal growth factor gene in mouse lachrymal gland: comparison with that in the submandibular gland and kidney. J Mol Endocrinol. 1990;4: van Setten GB, Tervo K, Virtanen I, Tarkkanen A, Tervo T. Immunohistochemical demonstration of epidermal growth factor in the lacrimal and submandibular glands of rats. Acta Ophthalmol. 1990; 68: Wilson SE, Lloyd SA, Kennedy RH. Epidermal growth factor messenger RNA production in human lacrimal gland. Cornea. 1991; 10: Wilson SE. Lacrimal gland epidermal growth factor production and the ocular surface. Am J Ophthalmol. 1991;6: Steinemann TL, Thompson HW, Maroney KM, et al. Changes in epithelial epidermal growth factor receptor and lacrimal gland EGF concentration after corneal wounding [ARVO Abstract]. Invest Ophthalmol Vis Sci. 1990;31;(4)S55. Abstract nr Van Setten GB, Tervo T, Viinikka L, Pesonen K. Ocular disease leads to decreased concentrations of epidermal growth factor in the tear fluid. Curr Eye Res. 1991;10: Wilson SE, Lloyd SA, Kennedy RH. Basic fibroblast growth factor (FGFb) and epidermal growth factor (EGF) receptor messenger RNA production in human lacrimal gland. Invest Ophthalmol Vis Sci. 1991;32: Wilson SE, He Y-G, Lloyd SA. EGF, basic FGF, and TGF beta-1 messenger RNA production in rabbit corneal epithelial cells. Invest Ophthalmol Vis Sci. 1992;33: Yoshino K, Monroy D, Pflugfelder SC. Cholinergic stimulation of lactoferrin and epidermal growth factor secretion by the human lacrimal gland. Cornea. 1996;15: Yoshino K, Garg R, Monroy D, Ji Z, Pflugfelder SC. Production and secretion of transforming growth factor beta (TGF-beta) by the human lacrimal gland. Curr Eye Res. 1996;15: Li Q, Weng J, Mohan RR, et al. Hepatocyte growth factor (HGF) and HGF receptor protein in lacrimal gland, tears, and cornea. Invest Ophthalmol Vis Sci. 1996;37: van Setten GB, Macauley S, Humphreys Beher M, Chegini N, Schultz G. Detection of transforming growth factor-alpha mrna and protein in rat lacrimal glands and characterization of transforming growth factor-alpha in human tears. Invest Ophthalmol Vis Sci. 1996;37: Thompson HW, Beuerman RW, Cook J, Underwood LW, Nguyen DH. Transcription of message for tumor necrosis factor-alpha by lacrimal gland is regulated by corneal wounding. Adv Exp Med Biol. 1994;350: Wilson SE. Growth factor and receptor messenger RNA production in human lacrimal gland tissue. Adv Exp Med Biol. 1994;350:

6 2190 Wilson et al. IOVS, September 1999, Vol. 40, No Tervo T, Vesaluoma M, Bennett GL, et al. Tear hepatocyte growth factor (HGF) availability increases markedly after excimer laser surface ablation. Exp Eye Res. 1997;64: Wilson SE, He Y-G, Weng J, Zeiske JD, Jester JV, Schultz GS. Effect of epidermal growth factor, hepatocyte growth factor, and keratinocyte growth factor on proliferation, motility, and differentiation of human corneal epithelial cells. Exp Eye Res. 1994;59: Wilson SE, Lie JJ, Mohan RR. Stromal-epithelial interactions in the cornea. Prog Retinal Eye Res. 1999;18: Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature. 1970;227: Liang Q, Richardson T. Expression and characterization of human lactoferrin in yeast Saccaromyces cerevisiae. J Agri Food Chem. 1993;41: Weng J, Mohan RR, Li Q, Wilson SE. IL-1 upregulates keratinocyte growth factor and hepatocyte growth factor mrna and protein production by cultured stromal fibroblast cells: interleukin-1 beta expression in the cornea. Cornea. 1997;16: Meneray MA, Bennett DJ, Nguyen DH, Beuerman RW. Effect of sensory denervation on the structure and physiologic responsiveness of rabbit lacrimal gland. Cornea. 1998;17: Yoshino K, Monroy D, Pflugfelder SC. Cholinergic stimulation of lactoferrin and epidermal growth factor secretion by the human lacrimal gland. Cornea. 1997;16: Mohan RR, Wilson SE. Ex vivo human corneal epithelial cells express membrane-bound precursor and mature soluble epidermal growth factor (EGF) and transforming growth factor (TGF) alpha proteins. Exp Eye Res. 1999;68: Wilson SE, Walker JW, Chwang EL, He Y-G. Hepatocyte growth factor (HGF), keratinocyte growth factor (KGF), their receptors, FGF Receptor-2, and the cells of the cornea. Invest Ophthalmol Vis Sci. 1993;34: Wilson SE, Chen L, Mohan RR, Liang Q, Liu J. Expression of HGF, KGF, EGF, and receptor messenger RNAs following corneal epithelial wounding. Exp Eye Res. 1999;68: