Vol. 70 (1992) Suppl. 202
ACTA OPHTHALMOLOGICA
MOLECULAR AND CELLULAR RESPONSES OF THE CORNEAL EPITHELIUM TO WOUND HEALING Roger W. Beuerman and Hilary W. Thompson
Lions Eye Research Laboratories, and the Laboratory for the Molecular Biology of the Ocular Surface, LSU Eye Center, Louisiana State University Medical Center, School of Medicine, New Orleans, LA, USA
Abstract. The corneal epithelium responds rapidly to injury, repairing defects with a layer of cells that covers the denuded corneal surface and prevents infection and loss of vision. After a wound, reorganization of the remaining epithelial cells occurs over several hours, resulting in the formation of a migratory leading edge. However, expression of genes such as c-fos occurs within minutes of wounding. This early expression may be important for directing epithelial reorganization and the later mitotic burst. Our results show that receptors for epidermal growth factor are upregulated in the migratory cell population. Proliferation through a mitotic burst was observed in cells surrounding the original wound margin after 36 hours. The interaction between gene expression and cell surface receptors for growth factors and cell proliferation suggests that wound healing occurs in a complex, but tightly controlled process in the corneal epithelium.
- cornea - epidermal growth factor - gene expression - wound healing.
Key words: c-fos
Introduction The cell biology of corneal wound healing has been studied because of its clinical importance and potential impact on the development of new approaches to clinical therapy. The unique nature of the cornea has made possible fundamental observations of the wound healing process, such as the time course of production of specific classes of proteins associated with wound healing (Zeiske et al., 1987; Zeiske & Gibson, 1986). Models of the wound healing process developed in our laboratory have shown that morpho-
typic changes in the major cells types, the epithelial cells and the keratocytes, are associated with phases of wound healing (Crosson et al., 1986). Keratocytes transform into a fibroblastic state as part of the wound repair process and may continue to be biosynthetically active long after epithelial cell migration has covered the defect (Nakayasu, 1988; Crosson, 1989). Studies of corneal epithelium and stroma have generally involved end points such as cell movement or alteration of specific proteins or classes of proteins. There is little information, however, on the molecular control of these processes, in terms of the expression of various genes associated with the response to wounding. Our current investigations are focused on the impact of gene expression on the control of the cellular response and the effects of polypeptide growth factors on the healing processes following corneal wounding (Assouline et al., 1989; Stern et al., 1989; Brazzell et al., 1991). Cellular oncogenes are a class of genes first characterized as part of the early studies of tumor viruses. In many instances, viral tumors can be attributed to the expression of one or more genes within the viral genome which have been recognized as mutant versions of cellular genes involved in growth and differentiation (Bishop, 1985). These genes became known as the cellular oncogenes or proto-oncogenes. It is now clear that the expression of these genes and their products are related to the transfer of a stimulus from the cell surface to the intracellular machin7
Roger W. Beuerman and Hilary W. Thompson
ery and then to subsequent changes in long-term expression of other genes involved in the production of structural proteins (Bishop, 1983). The purpose of the studies described in this paper has been to explore the expression of genes for cell signaling, the cellular oncogenes, after wounding, as well as the role of these genes in the wound healing process. Evidence now suggests, for instance, that certain genes, such as cjun and c-fos, may regulate the transcription of other genes and that c-fos may be induced by a number of stress-related conditions in other tissues, such as kidney and heart (Fine & Norman, 1989; Izumo et al., 1988).
Materials and methods The procedures used in this study were consistent with the ARVO Resolution on the Use of Animals in Research. Albino rabbits, 5 to 6 weeks of age and 2 to 3 kg in weight, were anesthetized by intramuscular injection of a ketamine/xylazine mixture (1.5 ml/kg body weight). Eyes were wounded as previously described (Frantz et al., 1989). The eye was protected by a sterile rubber dam and two drops of a topical anesthetic (proparacaine hydrochloride, 0.5 Yo) were applied to the cornea. A 6-mm-diameter central keratectomy 150-200 pm deep was created and topical antibiotic ointment (gentamicin, 3 mg/kg) was applied once. The progress of wound closure was documented photographically by means of fluorescein staining of the defect. At various times, animals were killed by an overdose of pentobarbital, and the corneas were removed and processed for light microscopy (Crosson et al., 1986). Epithelium for gene probe analysis was removed and immediately frozen in liquid nitrogen. Northern blot analysis was performed as described previously (Thompson et al., 1989). The changes in DNA content of the corneal epithelial cells and EGF receptor expression were evaluated by flow cytometry. The eyes were enucleated and transferred to a flow cytometer in moist chambers on crushed ice. The epithelium was collected from the surface of the wound using a No. 15 scalpel blade for DNA content or urethane for EGF receptors and placed in Hank's balanced salt solution (HBSS). Epithelial cells 8
were dissociated by gentle aspiration through a 25-gauge needle after filtration through nylon mesh to remove dissociated cells. Finally, after additional stages of centrifugation and treatment with RNAse and hypo-osmotic shock, the cells were stained for nuclear DNA with propidium iodide (Thompson et al., 1991; Vindelov et al., 1982). Flow cytometry was performed on an EPICS PROFILE set to acquire a maximum of 25,000 events. The 2N and 4N peaks were established by analysis of nuclei from rabbit white blood cells stained under the same conditions as the epithelial cell nuclei. The numbers and relative percentages of nuclei in the GO/G1, S , and G2/M phases were calculated using the quadratic method with cytologic software. Specific tests were conducted to compare mean percentages of nuclei in each phase in unwounded control specimens with the percentages at each time point after wounding. Flow cytometry was used for fluorescence detection of antibodies to the EGF receptor (EGFR) to quantify EGFR on corneal epithelial cells after experimental wounds. EGFR-specific fluorescence on epithelial cells was determined at 24 hours after wounding. The epithelial tissue was collected from eyes immediately after sacrifice by the topical application of 100 pl of urethane (500 mg/ml in normal saline), which induced complete epithelial sloughing. The sloughed epithelium was fixed briefly in 1 '70paraformaldehyde, dispersed to single cells by syringing with 23-gauge needles, and filtered through 70-pm nylon mesh to remove debris and undissociated cells. The effect of this method of tissue collection and fixation on EGF receptors was determined to be negligible by experiments on A431 cells, which overexpress the EGF receptor. Flow cytometry was performed as before on a Coulter Epic 753 flow cytometer. Each epithelial sample was divided into two parts. Half was incubated with the secondary antibody only for 30 minutes at room temperature to provide background fluorescence values. The other half was incubated with mouse antihuman EGFR monoclonal antibody (Oncogene Science, Manhasset, NY) at a 1:1000 dilution. The fluoresceintagged secondary was incubated with the preparation for 30 minutes at room temperature.
Molecular and cellular responses of the corneal epithelium
Results Closure of a circular epithelial abrasion or keratectomy follows a well-known sequence of events in which the radial advance of the epithelial cells to the center of the wound resembles a “purse string” or “draw string” pattern. As seen in Fig. 1, the stages of wound healing are accompanied by’ a characteristic appearance of the epithelium. Wounding leaves a swathe of damaged cells at the periphery (Fig. 1 , O hr). During the so-called latent period, the cells at the margin dedifferentiate and reorganize, with a single cell layer at the leading edge (Fig. 1, 14 hr). This appearance is maintained until closure. Because of the two or three layers of cells just behind the leading edge, the epithelium has a multilayered appearance just after closure, although the thickness is still one or two cell layers less than that of normal epithelium (Fig. 1, 40 hr). Fig. 1 also confirms the role of the epithelium in the maintenance of the stromal keratocytes. After wounding, there are visible spaces in the swollen stroma just beneath the de-epithelialized surface formerly occupied by keratocyte cell bodies that were damaged by the anterior expansion of the stroma. As the wound closes and stromal swelling decreases, keratocytes move in from the periphery of the wound and repopulate the anterior stroma. The suggestion that changes in gene expression may underlie significant developments in the dedifferentiation, migration, and even potentially changes in membrane receptors is based on our examination of the expression of cellular oncogenes. Early experiments in our laboratory examined the response of wounded corneal epithelium and the cellular oncogene c-fos (Thompson et al., 1989). In these experiments, the corneal epithelium was collected directly from rabbits at various times following anterior keratectomy wounds. The importance of the cellular oncogenes is that they signal a membrane event and their protein products are involved in the transcriptional control of other genes. In the present studies, the corneas were wounded by an anterior keratectomy as before, and the epithelium was collected from the area outside the wound at 15 minutes, 30 minutes, and 1 hour. Epithelium was then collected from another group of six corneas 2 hours after wound-
Fig. 1. Cross-sectional view of the cornea, including
the wound margin, and the microphotographic appearance of the defect at the indicated times after wounding. The wound used in this sequence was a 4-mmdiameter abrasion.
ing with cycloheximide applied every 10 minutes during the 2-hour period. Cycloheximide interferes with synthesis of proteins that would be involved in the degradation of the product of the c-fos gene; the use of cycloheximide in these studies is designed to show that the gene was under transcriptional control through negative feedback. Expression of c-fos in the wounded cornea was determined (Fig. 2). The results represent the RNA from six rabbits at each time point. The fact that cycloheximide prolongs the expression of c-fos confirms that negative feedback is involved in the control of this gene, which has such global effects on the cell. Without cycloheximide
9
Roger W. Beuerman and Hilary W. Thompson
mxlRsAFTwWMMHG
Fig. 2. Autoradiograph of Northern blot. Rabbit epithelial total cellular RNA was collected at the indicated times after wounding. The blot was probed with the genomic clone of human c-fos. Lanes from left to
right represent increased time after wounding. Lane 1 represents unwounded. Lane 2 represents 15 minutes after wounding. Lane 3 represents an empty lane for orientation. Lane 4 represents 30 minutes after wounding. Lane 5 represents 1 hour after wounding. Lane 6 represents 2 hours after wounding with topical cycloheximide applied to the cornea before specimen collection.
treatment, c-fos expression was minimal at 60 minutes; thus, the transient nature of this expression was shown in vivo. Flow cytometric analysis of DNA content was carried out at each of the postwounding time points indicated in Figure 3. The actual cell counts of the number of nuclei in each phase are shown. The basal rate of mitosis seen in our previous study of 32 unwounded control specimens (Thompson et al., 1991) was characterized SEM) nuclei in S by 13.5 f 0.70 Yo (mean phase and 8.7 Yo f 0.42 Yo nuclei in G2/M (4N) phase. In that study, wounded corneas at 12 hrs had significantly greater numbers of nuclei in S phase, compared to unwounded corneas. However, there was no significant difference in G2/M nuclei, compared to the control specimens. Twenty four hours after wounding, the number of nuclei in S phase was not significantly different from the number in unwounded controls. There were, however, significantly more nuclei in the G2/M phase (p < 0.01), compared with the controls. In the present study, the number of nuclei in 10
Fig. 3. Means and standard errors of the total number of cell nuclei counted in each phase of mitosis by flow cytometry. The epithelium was wounded and tissue was collected at 12-hour intervals. There were 12 samples at each time interval. Asterisks indicate significant differences (p < 0.05) from 12-hour values. 2N indicates GO/Gl phase; 4N indicates G2/M phase.
G2/M (4N) was significantly greater than the number in S phase at later times (36 to 48 hours) (Fig. 3). This phenomenon indicates a burst of mitotic activity. However, in the earlier study (Thompson et al., 1991), it was found that 72 hours after wounding, the distribution of DNA in the regenerated epithelium at the center of the wound was not significantly different from that of unwounded control specimens, suggesting that this area is not a focus of cell proliferation after epithelial cell cover is complete. The current results show an overall decrease in cellularity, as indicated by the decline in the number of cells in the 2N phase. Additionally, at 48 hours, there is a marked difference in the total numbers of cells, compared to earlier times, as well as to control specimens. These results may reflect the loss of more superficial cells accompanying the movement of the epithelium as a thin sheet into the area of the wound (Fig. 1). As noted above, the regenerated epithelium forms a multilayered appearance as healing is completed, but is actually one or two cell layers thinner than normal epithelium. Flow cytograms show the EGF receptor density of normal corneal epithelial cells (Fig. 4A) and in corneal epithelial cells after wounding (Fig. 4B). It is possible that wounding results in an overall increase in cell surface antigens or that
Molecular and cellular responses
of t h e
corneal epithelium
epithelial cells increases significantly by 24 hours after wounding, compared to unwounded controls.
Discussion The present studies show that wounding of the cornea leads to broad changes in the physiology and behavior of the epithelial cells. These changes are induced by damage to the corneal epithelial cells; however, the precise stimulus at the cellular level is not clear at the present time (Crosson et al., 1986). One possibility may be a change in calcium within the cells, which is known to lead to expression of cellular,oncogenes. Another possibility involves the exposure of many cell membranes to growth factors such as epidermal growth factor in the tears following the wound. Recent studies have shown that EGF is present in the tears of patients with corneal injuries (Van Setten et al., 1990). The expression of c-fos within the first 15 minutes after wounding is an indication of the role the products of this gene may have in the regulation of the expression of subsequent genes, such as those for structural proteins (Miller et al., LOG FLUORESCENCE 1984). Interestingly, the expression of c-fos in the Fig. 4. Plots of flow cytograms demonstrating EGF corneal epithelium is very similar to that observed receptor density as determined by fluorescence antibody detection. The dark curve represents background in other in vivo systems, and as well in some cell fluorescence from anti-EGFR monoclonal primary culture systems: early transient expression, with antibodies. The grey, filled-in areas represent the sig- levels declining by 1 hour after stimulation nal from fluorescein-tagged secondary antibodies. (Fine & Norman, 1989). Although the expression Where the grey areas fall within the dark curve, the of c-myc has not yet been determined in the corsecondary signal is a false positive. Only where the grey areas are shifted to the right and outside of the dark neal epithelium, observations in other systems curve is true signal recorded. Thus, in the unwounded suggest that this gene would be expressed some control corneas (A), no EGF receptors are detectable time after the expression of c-fos. C-myc is in normal epithelial cells, whereas in healing corneas important since it is usually associated with the 24 hours after wounding (B), the arrow indicates an commitment of a cell to mitosis; however, there area of true signal and hence an increase in EGF recep- is expression in cell lines that do not go on to tor density on the epithelial cell population. proliferation. The results from the flow cytometry experiments suggest that the mitotic burst is probably induced by expression of the c-myc there is an increase in membrane-related recep- oncogene. Most likely this occurs at some time tors associated with epithelial cell migration con- within 2 to 5 hours following wounding. The tributing to this effect. However, the most sig- changes in the regulation of the EGFR are probnificant aspect of the postwounding cytogram is ably contingent upon the early expression of the shift of the filled curve to the right (Fig. 4B; other cellular oncogenes in addition to c-fos. arrow), indicating brighter cells and more However, it is this early expression that sets in fluorescent label, which represents an increase in motion a cascade of events leading to later EGF receptor-specific binding. The data demon- changes in cellular receptors which may be instrate that EGFR density on the surface of volved in the healing process.
11
Roger W. Beuerman and Hilary W. Thompson
These types of experiments may lead to new avenues for the development of mechanisms for pharmacological control of cell behavior in slowly healing and non-healing wounds. For example, if it could be demonstrated that epithelial cells do not express these cellular genes in certain clinical pathological states, then various approaches to the stimulation of this expression, such as changes in the ionic composition of the tear film, could be tested for the ability to enhance epithelial healing. Along this line, it is been shown clinically that removal of the tissue or debridement of the cornea around a slowly healing wound may spur the epithelium into successfully covering the defect. The reason for the success of this procedure may be that it sets in motion a series of genetic events which initiate the wound healing activity.
Acknowledgements This work was supported in part by U.S. Public Health Service grants EY04074 and EY02377 from the National Eye Institute, National Institutes of Health, Bethesda, Maryland.
References Assouline M, Montefiore G, Pouliquen Y & Courtois Y (1989): Fibroblast growth factor effects on corneal wound healing in the rabbit. In: Beuerman R W, Crosson C E & Kaufman H E (eds). Healing Processes in the Cornea, p. 79-98. Portfolio, The Woodlands, TX. Bishop J M (1983): Cellular oncogenes and retroviruses. Annu Rev Biochem 52: 301-354. Bishop J M (1985): Viral oncogenes. Cell 42: 23-38. Brazzell R K, Stern M E, Aquavella J V, Beuerman R W & Baird L (1991): Human recombinant epidermal growth factor in experimental corneal wound healing. Invest Ophthalmol Vis Sci 32: 336-340. Crosson C (1989): Cellular changes following epithelial abrasion. In: Beuerman R W, Crosson C E, Kaufman H E (eds). Healing Processes in the Cornea, p. 3-14. Portfolio, The Woodlands, TX. Crosson C E, Klyce S, & Beuerman R W (1986): Epithelial wound closure in the rabbit cornea. Invest Ophthalmol Vis Sci 27: 464-473.
12
Fine L G & Norman J (1989): Cellular events in renal hypertrophy. Annu Rev Physiol 51: 19-32. Frantz J M, Dupuy B M, Kaufman H E & Beuerman R W (1989): The effect of collagen shields on epithelial wound healing in rabbits. Am J Ophthalmol 108: 524-528. Izumo S, Nadal-Ginard B & Mahdavi V (1988): Protooncogene induction and reprogramming of cardiac gene expression produced by pressure overload. Proc Natl Acad Sci U S A 85: 339-343. Miller A D, Curran T & Verma I M (1984): C-fos protein can induce cellular transformation: a novel mechanism of activation of a cellular oncogene. Cell 36: 51-60. Nakayasu K (1988): Stromal changes following removal of epithelium in rat cornea. Jpn J Ophthalmol 32: 113-125. Stern M E, Brazzell R K, Beuerman R W, Aquavella J V & Kirschner S E (1989): The effects of human recombinant epidermal growth factor on epithelial wound healing. In: Beuerman R W, Crosson C E & Kaufman H E (eds). Healing Processes in the Cornea, p. 69-78. Portfolio, The Woodlands, TX. Thompson H, Malter J S, Steinemann T L & Beuerman R W (1991): Flow cytometry measurements of the DNA content of corneal epithelial cells during wound healing. Invest Ophthalmol Vis Sci 32: 433-436. Thompson H , Thompson J, Lockyer J & Beuerman R W (1989): Protooncogene expression during corneal wound healing. In: Beuerman R W, Crosson C E & Kaufman H E (eds). Healing Processes in the Cornea, p. 59-68. Portfolio, The Woodlands, TX. Van Setten G B, Viinikka L, Tervo T & Pesonen K (1990): Epidermal growth factor is a constant component of normal human tear fluid. Graefes Arch Clin Exp Ophthalmol 227: 184-187. Vindelov L L, Christensen J & Nissen N I (1982): A detergent-trypsin method for the preparation of nuclei for flow cytometric DNA analysis. Cytometry 3: 323-328. Zeiske J D & Gipson I K (1986): Protein synthesis during corneal epithelial wound healing. Invest Ophthalmol Vis Sci 27: 1-7. Zeiske J D, Higashijimo S C, Spurr-Michaud S J & Gipson I K (1987): Biosynthetic response of the rabbit cornea to a keratectomy wound. Invest Ophthalmol Vis Sci 28: 1668-1677.
Author’s address: Roger W. Beuerman, Ph.D LSU Eye Center 2020 Gravier Street, Suite B New Orleans, LA 70112, USA
ACTA OPHTHALMOLOGICA
MOLECULAR AND CELLULAR RESPONSES OF THE CORNEAL EPITHELIUM TO WOUND HEALING Roger W. Beuerman and Hilary W. Thompson
Lions Eye Research Laboratories, and the Laboratory for the Molecular Biology of the Ocular Surface, LSU Eye Center, Louisiana State University Medical Center, School of Medicine, New Orleans, LA, USA
Abstract. The corneal epithelium responds rapidly to injury, repairing defects with a layer of cells that covers the denuded corneal surface and prevents infection and loss of vision. After a wound, reorganization of the remaining epithelial cells occurs over several hours, resulting in the formation of a migratory leading edge. However, expression of genes such as c-fos occurs within minutes of wounding. This early expression may be important for directing epithelial reorganization and the later mitotic burst. Our results show that receptors for epidermal growth factor are upregulated in the migratory cell population. Proliferation through a mitotic burst was observed in cells surrounding the original wound margin after 36 hours. The interaction between gene expression and cell surface receptors for growth factors and cell proliferation suggests that wound healing occurs in a complex, but tightly controlled process in the corneal epithelium.
- cornea - epidermal growth factor - gene expression - wound healing.
Key words: c-fos
Introduction The cell biology of corneal wound healing has been studied because of its clinical importance and potential impact on the development of new approaches to clinical therapy. The unique nature of the cornea has made possible fundamental observations of the wound healing process, such as the time course of production of specific classes of proteins associated with wound healing (Zeiske et al., 1987; Zeiske & Gibson, 1986). Models of the wound healing process developed in our laboratory have shown that morpho-
typic changes in the major cells types, the epithelial cells and the keratocytes, are associated with phases of wound healing (Crosson et al., 1986). Keratocytes transform into a fibroblastic state as part of the wound repair process and may continue to be biosynthetically active long after epithelial cell migration has covered the defect (Nakayasu, 1988; Crosson, 1989). Studies of corneal epithelium and stroma have generally involved end points such as cell movement or alteration of specific proteins or classes of proteins. There is little information, however, on the molecular control of these processes, in terms of the expression of various genes associated with the response to wounding. Our current investigations are focused on the impact of gene expression on the control of the cellular response and the effects of polypeptide growth factors on the healing processes following corneal wounding (Assouline et al., 1989; Stern et al., 1989; Brazzell et al., 1991). Cellular oncogenes are a class of genes first characterized as part of the early studies of tumor viruses. In many instances, viral tumors can be attributed to the expression of one or more genes within the viral genome which have been recognized as mutant versions of cellular genes involved in growth and differentiation (Bishop, 1985). These genes became known as the cellular oncogenes or proto-oncogenes. It is now clear that the expression of these genes and their products are related to the transfer of a stimulus from the cell surface to the intracellular machin7
Roger W. Beuerman and Hilary W. Thompson
ery and then to subsequent changes in long-term expression of other genes involved in the production of structural proteins (Bishop, 1983). The purpose of the studies described in this paper has been to explore the expression of genes for cell signaling, the cellular oncogenes, after wounding, as well as the role of these genes in the wound healing process. Evidence now suggests, for instance, that certain genes, such as cjun and c-fos, may regulate the transcription of other genes and that c-fos may be induced by a number of stress-related conditions in other tissues, such as kidney and heart (Fine & Norman, 1989; Izumo et al., 1988).
Materials and methods The procedures used in this study were consistent with the ARVO Resolution on the Use of Animals in Research. Albino rabbits, 5 to 6 weeks of age and 2 to 3 kg in weight, were anesthetized by intramuscular injection of a ketamine/xylazine mixture (1.5 ml/kg body weight). Eyes were wounded as previously described (Frantz et al., 1989). The eye was protected by a sterile rubber dam and two drops of a topical anesthetic (proparacaine hydrochloride, 0.5 Yo) were applied to the cornea. A 6-mm-diameter central keratectomy 150-200 pm deep was created and topical antibiotic ointment (gentamicin, 3 mg/kg) was applied once. The progress of wound closure was documented photographically by means of fluorescein staining of the defect. At various times, animals were killed by an overdose of pentobarbital, and the corneas were removed and processed for light microscopy (Crosson et al., 1986). Epithelium for gene probe analysis was removed and immediately frozen in liquid nitrogen. Northern blot analysis was performed as described previously (Thompson et al., 1989). The changes in DNA content of the corneal epithelial cells and EGF receptor expression were evaluated by flow cytometry. The eyes were enucleated and transferred to a flow cytometer in moist chambers on crushed ice. The epithelium was collected from the surface of the wound using a No. 15 scalpel blade for DNA content or urethane for EGF receptors and placed in Hank's balanced salt solution (HBSS). Epithelial cells 8
were dissociated by gentle aspiration through a 25-gauge needle after filtration through nylon mesh to remove dissociated cells. Finally, after additional stages of centrifugation and treatment with RNAse and hypo-osmotic shock, the cells were stained for nuclear DNA with propidium iodide (Thompson et al., 1991; Vindelov et al., 1982). Flow cytometry was performed on an EPICS PROFILE set to acquire a maximum of 25,000 events. The 2N and 4N peaks were established by analysis of nuclei from rabbit white blood cells stained under the same conditions as the epithelial cell nuclei. The numbers and relative percentages of nuclei in the GO/G1, S , and G2/M phases were calculated using the quadratic method with cytologic software. Specific tests were conducted to compare mean percentages of nuclei in each phase in unwounded control specimens with the percentages at each time point after wounding. Flow cytometry was used for fluorescence detection of antibodies to the EGF receptor (EGFR) to quantify EGFR on corneal epithelial cells after experimental wounds. EGFR-specific fluorescence on epithelial cells was determined at 24 hours after wounding. The epithelial tissue was collected from eyes immediately after sacrifice by the topical application of 100 pl of urethane (500 mg/ml in normal saline), which induced complete epithelial sloughing. The sloughed epithelium was fixed briefly in 1 '70paraformaldehyde, dispersed to single cells by syringing with 23-gauge needles, and filtered through 70-pm nylon mesh to remove debris and undissociated cells. The effect of this method of tissue collection and fixation on EGF receptors was determined to be negligible by experiments on A431 cells, which overexpress the EGF receptor. Flow cytometry was performed as before on a Coulter Epic 753 flow cytometer. Each epithelial sample was divided into two parts. Half was incubated with the secondary antibody only for 30 minutes at room temperature to provide background fluorescence values. The other half was incubated with mouse antihuman EGFR monoclonal antibody (Oncogene Science, Manhasset, NY) at a 1:1000 dilution. The fluoresceintagged secondary was incubated with the preparation for 30 minutes at room temperature.
Molecular and cellular responses of the corneal epithelium
Results Closure of a circular epithelial abrasion or keratectomy follows a well-known sequence of events in which the radial advance of the epithelial cells to the center of the wound resembles a “purse string” or “draw string” pattern. As seen in Fig. 1, the stages of wound healing are accompanied by’ a characteristic appearance of the epithelium. Wounding leaves a swathe of damaged cells at the periphery (Fig. 1 , O hr). During the so-called latent period, the cells at the margin dedifferentiate and reorganize, with a single cell layer at the leading edge (Fig. 1, 14 hr). This appearance is maintained until closure. Because of the two or three layers of cells just behind the leading edge, the epithelium has a multilayered appearance just after closure, although the thickness is still one or two cell layers less than that of normal epithelium (Fig. 1, 40 hr). Fig. 1 also confirms the role of the epithelium in the maintenance of the stromal keratocytes. After wounding, there are visible spaces in the swollen stroma just beneath the de-epithelialized surface formerly occupied by keratocyte cell bodies that were damaged by the anterior expansion of the stroma. As the wound closes and stromal swelling decreases, keratocytes move in from the periphery of the wound and repopulate the anterior stroma. The suggestion that changes in gene expression may underlie significant developments in the dedifferentiation, migration, and even potentially changes in membrane receptors is based on our examination of the expression of cellular oncogenes. Early experiments in our laboratory examined the response of wounded corneal epithelium and the cellular oncogene c-fos (Thompson et al., 1989). In these experiments, the corneal epithelium was collected directly from rabbits at various times following anterior keratectomy wounds. The importance of the cellular oncogenes is that they signal a membrane event and their protein products are involved in the transcriptional control of other genes. In the present studies, the corneas were wounded by an anterior keratectomy as before, and the epithelium was collected from the area outside the wound at 15 minutes, 30 minutes, and 1 hour. Epithelium was then collected from another group of six corneas 2 hours after wound-
Fig. 1. Cross-sectional view of the cornea, including
the wound margin, and the microphotographic appearance of the defect at the indicated times after wounding. The wound used in this sequence was a 4-mmdiameter abrasion.
ing with cycloheximide applied every 10 minutes during the 2-hour period. Cycloheximide interferes with synthesis of proteins that would be involved in the degradation of the product of the c-fos gene; the use of cycloheximide in these studies is designed to show that the gene was under transcriptional control through negative feedback. Expression of c-fos in the wounded cornea was determined (Fig. 2). The results represent the RNA from six rabbits at each time point. The fact that cycloheximide prolongs the expression of c-fos confirms that negative feedback is involved in the control of this gene, which has such global effects on the cell. Without cycloheximide
9
Roger W. Beuerman and Hilary W. Thompson
mxlRsAFTwWMMHG
Fig. 2. Autoradiograph of Northern blot. Rabbit epithelial total cellular RNA was collected at the indicated times after wounding. The blot was probed with the genomic clone of human c-fos. Lanes from left to
right represent increased time after wounding. Lane 1 represents unwounded. Lane 2 represents 15 minutes after wounding. Lane 3 represents an empty lane for orientation. Lane 4 represents 30 minutes after wounding. Lane 5 represents 1 hour after wounding. Lane 6 represents 2 hours after wounding with topical cycloheximide applied to the cornea before specimen collection.
treatment, c-fos expression was minimal at 60 minutes; thus, the transient nature of this expression was shown in vivo. Flow cytometric analysis of DNA content was carried out at each of the postwounding time points indicated in Figure 3. The actual cell counts of the number of nuclei in each phase are shown. The basal rate of mitosis seen in our previous study of 32 unwounded control specimens (Thompson et al., 1991) was characterized SEM) nuclei in S by 13.5 f 0.70 Yo (mean phase and 8.7 Yo f 0.42 Yo nuclei in G2/M (4N) phase. In that study, wounded corneas at 12 hrs had significantly greater numbers of nuclei in S phase, compared to unwounded corneas. However, there was no significant difference in G2/M nuclei, compared to the control specimens. Twenty four hours after wounding, the number of nuclei in S phase was not significantly different from the number in unwounded controls. There were, however, significantly more nuclei in the G2/M phase (p < 0.01), compared with the controls. In the present study, the number of nuclei in 10
Fig. 3. Means and standard errors of the total number of cell nuclei counted in each phase of mitosis by flow cytometry. The epithelium was wounded and tissue was collected at 12-hour intervals. There were 12 samples at each time interval. Asterisks indicate significant differences (p < 0.05) from 12-hour values. 2N indicates GO/Gl phase; 4N indicates G2/M phase.
G2/M (4N) was significantly greater than the number in S phase at later times (36 to 48 hours) (Fig. 3). This phenomenon indicates a burst of mitotic activity. However, in the earlier study (Thompson et al., 1991), it was found that 72 hours after wounding, the distribution of DNA in the regenerated epithelium at the center of the wound was not significantly different from that of unwounded control specimens, suggesting that this area is not a focus of cell proliferation after epithelial cell cover is complete. The current results show an overall decrease in cellularity, as indicated by the decline in the number of cells in the 2N phase. Additionally, at 48 hours, there is a marked difference in the total numbers of cells, compared to earlier times, as well as to control specimens. These results may reflect the loss of more superficial cells accompanying the movement of the epithelium as a thin sheet into the area of the wound (Fig. 1). As noted above, the regenerated epithelium forms a multilayered appearance as healing is completed, but is actually one or two cell layers thinner than normal epithelium. Flow cytograms show the EGF receptor density of normal corneal epithelial cells (Fig. 4A) and in corneal epithelial cells after wounding (Fig. 4B). It is possible that wounding results in an overall increase in cell surface antigens or that
Molecular and cellular responses
of t h e
corneal epithelium
epithelial cells increases significantly by 24 hours after wounding, compared to unwounded controls.
Discussion The present studies show that wounding of the cornea leads to broad changes in the physiology and behavior of the epithelial cells. These changes are induced by damage to the corneal epithelial cells; however, the precise stimulus at the cellular level is not clear at the present time (Crosson et al., 1986). One possibility may be a change in calcium within the cells, which is known to lead to expression of cellular,oncogenes. Another possibility involves the exposure of many cell membranes to growth factors such as epidermal growth factor in the tears following the wound. Recent studies have shown that EGF is present in the tears of patients with corneal injuries (Van Setten et al., 1990). The expression of c-fos within the first 15 minutes after wounding is an indication of the role the products of this gene may have in the regulation of the expression of subsequent genes, such as those for structural proteins (Miller et al., LOG FLUORESCENCE 1984). Interestingly, the expression of c-fos in the Fig. 4. Plots of flow cytograms demonstrating EGF corneal epithelium is very similar to that observed receptor density as determined by fluorescence antibody detection. The dark curve represents background in other in vivo systems, and as well in some cell fluorescence from anti-EGFR monoclonal primary culture systems: early transient expression, with antibodies. The grey, filled-in areas represent the sig- levels declining by 1 hour after stimulation nal from fluorescein-tagged secondary antibodies. (Fine & Norman, 1989). Although the expression Where the grey areas fall within the dark curve, the of c-myc has not yet been determined in the corsecondary signal is a false positive. Only where the grey areas are shifted to the right and outside of the dark neal epithelium, observations in other systems curve is true signal recorded. Thus, in the unwounded suggest that this gene would be expressed some control corneas (A), no EGF receptors are detectable time after the expression of c-fos. C-myc is in normal epithelial cells, whereas in healing corneas important since it is usually associated with the 24 hours after wounding (B), the arrow indicates an commitment of a cell to mitosis; however, there area of true signal and hence an increase in EGF recep- is expression in cell lines that do not go on to tor density on the epithelial cell population. proliferation. The results from the flow cytometry experiments suggest that the mitotic burst is probably induced by expression of the c-myc there is an increase in membrane-related recep- oncogene. Most likely this occurs at some time tors associated with epithelial cell migration con- within 2 to 5 hours following wounding. The tributing to this effect. However, the most sig- changes in the regulation of the EGFR are probnificant aspect of the postwounding cytogram is ably contingent upon the early expression of the shift of the filled curve to the right (Fig. 4B; other cellular oncogenes in addition to c-fos. arrow), indicating brighter cells and more However, it is this early expression that sets in fluorescent label, which represents an increase in motion a cascade of events leading to later EGF receptor-specific binding. The data demon- changes in cellular receptors which may be instrate that EGFR density on the surface of volved in the healing process.
11
Roger W. Beuerman and Hilary W. Thompson
These types of experiments may lead to new avenues for the development of mechanisms for pharmacological control of cell behavior in slowly healing and non-healing wounds. For example, if it could be demonstrated that epithelial cells do not express these cellular genes in certain clinical pathological states, then various approaches to the stimulation of this expression, such as changes in the ionic composition of the tear film, could be tested for the ability to enhance epithelial healing. Along this line, it is been shown clinically that removal of the tissue or debridement of the cornea around a slowly healing wound may spur the epithelium into successfully covering the defect. The reason for the success of this procedure may be that it sets in motion a series of genetic events which initiate the wound healing activity.
Acknowledgements This work was supported in part by U.S. Public Health Service grants EY04074 and EY02377 from the National Eye Institute, National Institutes of Health, Bethesda, Maryland.
References Assouline M, Montefiore G, Pouliquen Y & Courtois Y (1989): Fibroblast growth factor effects on corneal wound healing in the rabbit. In: Beuerman R W, Crosson C E & Kaufman H E (eds). Healing Processes in the Cornea, p. 79-98. Portfolio, The Woodlands, TX. Bishop J M (1983): Cellular oncogenes and retroviruses. Annu Rev Biochem 52: 301-354. Bishop J M (1985): Viral oncogenes. Cell 42: 23-38. Brazzell R K, Stern M E, Aquavella J V, Beuerman R W & Baird L (1991): Human recombinant epidermal growth factor in experimental corneal wound healing. Invest Ophthalmol Vis Sci 32: 336-340. Crosson C (1989): Cellular changes following epithelial abrasion. In: Beuerman R W, Crosson C E, Kaufman H E (eds). Healing Processes in the Cornea, p. 3-14. Portfolio, The Woodlands, TX. Crosson C E, Klyce S, & Beuerman R W (1986): Epithelial wound closure in the rabbit cornea. Invest Ophthalmol Vis Sci 27: 464-473.
12
Fine L G & Norman J (1989): Cellular events in renal hypertrophy. Annu Rev Physiol 51: 19-32. Frantz J M, Dupuy B M, Kaufman H E & Beuerman R W (1989): The effect of collagen shields on epithelial wound healing in rabbits. Am J Ophthalmol 108: 524-528. Izumo S, Nadal-Ginard B & Mahdavi V (1988): Protooncogene induction and reprogramming of cardiac gene expression produced by pressure overload. Proc Natl Acad Sci U S A 85: 339-343. Miller A D, Curran T & Verma I M (1984): C-fos protein can induce cellular transformation: a novel mechanism of activation of a cellular oncogene. Cell 36: 51-60. Nakayasu K (1988): Stromal changes following removal of epithelium in rat cornea. Jpn J Ophthalmol 32: 113-125. Stern M E, Brazzell R K, Beuerman R W, Aquavella J V & Kirschner S E (1989): The effects of human recombinant epidermal growth factor on epithelial wound healing. In: Beuerman R W, Crosson C E & Kaufman H E (eds). Healing Processes in the Cornea, p. 69-78. Portfolio, The Woodlands, TX. Thompson H, Malter J S, Steinemann T L & Beuerman R W (1991): Flow cytometry measurements of the DNA content of corneal epithelial cells during wound healing. Invest Ophthalmol Vis Sci 32: 433-436. Thompson H , Thompson J, Lockyer J & Beuerman R W (1989): Protooncogene expression during corneal wound healing. In: Beuerman R W, Crosson C E & Kaufman H E (eds). Healing Processes in the Cornea, p. 59-68. Portfolio, The Woodlands, TX. Van Setten G B, Viinikka L, Tervo T & Pesonen K (1990): Epidermal growth factor is a constant component of normal human tear fluid. Graefes Arch Clin Exp Ophthalmol 227: 184-187. Vindelov L L, Christensen J & Nissen N I (1982): A detergent-trypsin method for the preparation of nuclei for flow cytometric DNA analysis. Cytometry 3: 323-328. Zeiske J D & Gipson I K (1986): Protein synthesis during corneal epithelial wound healing. Invest Ophthalmol Vis Sci 27: 1-7. Zeiske J D, Higashijimo S C, Spurr-Michaud S J & Gipson I K (1987): Biosynthetic response of the rabbit cornea to a keratectomy wound. Invest Ophthalmol Vis Sci 28: 1668-1677.
Author’s address: Roger W. Beuerman, Ph.D LSU Eye Center 2020 Gravier Street, Suite B New Orleans, LA 70112, USA
