Keratin-like Proteins in Corneal and Conjunctival Epithelium are Different Shigeru Kinoshiro,* Judith Friend, Timothy C. Kiorpes, and Richard A. Thoft Using SDS polyacrylamide slab-gel electrophoresis, water-insoluble (keratin-like) proteins in normal and regenerated ocular surface epithelium from rabbits were studied. The results indicated that keratin-like proteins from corneal and conjunctival epithelia in vivo were distinctly different. Regenerated epithelia from either source retained their original keratin characteristics for at least 10 days after healing over the cornea, but at very early stages of healing migrating and regenerated epithelia showed either an extra band or a prominent band in addition to the original keratin-like proteins. Three months after healing, however, regenerated conjunctival epithelium on the cornea had changed its keratin characteristics, and resembled, but was not identical to, corneal epithelium. Invest Ophthalmol Vis Sci 24:577-581, 1983 Keratins are a group of proteins that are insoluble in neutral aqueous buffers. 1 Some intermediate filaments in the cytoplasm of eukaryocytic cells are made up of keratin, and keratin is an important constituent of the intracellular cytoskeleton in almost all epithelia including nonkeratinizing epithelia such as those covering the cornea and conjunctiva, the ocular surface epithelium. 2 " 5 In addition, keratin is the major product associated with maturation of some epithelial tissues, so that various stages of cell differentiation as well as the types of epithelial cells present in a species are often demonstrable by keratin composition. 5-6 Since conjunctival epithelium covers the corneal surface in some clinical situations such as chemical injury and conjunctival transplantation, 7 it is important to understand to what degree conjunctival epithelium can become a true substitute for corneal epithelium. Possibly, keratin may be a more sensitive biochemical marker to demonstrate whether regenerating ocular surface epithelium is corneal or conjunctival than are glycolytic enzymes and glycogen content. 8 Therefore, we have studied the water-in- From the Department of Cornea Research, Eye Research Institute of Retina Foundation, and Department of Ophthalmology, Harvard Medical School, Boston, Mass. Supported in part by Research Grants EY-01830 and EY-03061 from the National Eye Institute, National Institute of Health, and in part by the Lions Eye Research Fund, Inc. Submitted for publication: June 1, 1982. Reprint requests: Judith Friend, Eye Research Institute of Retina Foundation, 20 Staniford Street, Boston, MA 02114. * Present address: Department of Ophthalmology, Osaka University Medical School, Osaka, Japan. soluble (keratin-like) proteins in epithelium from either cornea or conjunctiva as it grows over the cornea. Materials and Methods Animal Preparation Albino rabbits weighing 2-3 kg were anesthetized with chlorpromazine hydrochloride-ketamine hydrochloride, maintained with ether inhalation and topical proparacaine. The eyes then received epithelial debridements of either a 10-mm diameter central corneal area or the whole corneal surface plus a 2-3 mm zone of limbal and bulbar conjunctival area by n- heptanol according to the procedure of Shapiro et al. 9 Forty-eight wounded eyes, which showed no posthealing complications such as extensive vascularization or recurrent epithelial defects, and ten normal unwounded eyes were used in this experiment. At various times after wounding, animals were killed by an overdose of sodium pentobarbital, and the eyes were rinsed with balanced salt solution. Samples of normal corneal epithelium and epithelium regenerated from cornea or conjunctiva were scraped off the cornea, frozen in liquid nitrogen, and stored at 80 C. Normal conjunctival epithelial samples were obtained by gently scraping the bulbar conjunctiva and freezing the samples as before. Preparation of Keratin-like Proteins and Total Proteins from Epithelial Samples Insoluble proteins of the epithelial samples were obtained by a modification of the technique of Sun 0146-0404/83/0500/577/$ 1.05 Association for Research in Vision and Ophthalmology 577
578 INVESTIGATIVE OPHTHALMOLOGY 6 VISUAL SCIENCE / Moy 1983 Vol. 24 so that we have chosen to call them "keratin-like" protein. Briefly, epithelial samples from two eyes were homogenized in borosilicate glass tube (6 X 50 mm, Kimble, IL) with fitted glass rods and extracted in 300 /xl of 20 mm Tris-HCl buffer (ph 7.4) at 4 C. Keratin-like proteins were obtained by-centrifugation at 12,000 g for 30 min. The extraction procedure was performed twice for a total extraction time of 24 hrs. Then, the keratin-protein containing pellets were solubilized in 100 /A of 2% sodium dodecyl sulfate (SDS) in 10 mm sodium phosphate (ph 7.0) using Bonification and boiling for 10 min as described by Hassell et al.10 More than 90% of keratin-like proteins were solubilized by this treatment. For total protein of epithelial samples, epithelium from one eye was solubilized directly in 2% SDS in sodium phosphate using the method described above. The protein content in the solubilized samples was determined by the Lowry method" using bovine serum albumin as a standard. CO 'o Polyacrylamide Gel Electrophoresis (PAGE) 36- A B C D E Std Fig. 1. A 7.J-/O puiyaurymmiue siao-gei eiecxropnoresis oi Keratinlike proteins in regenerated and normal corneal epithelium. The direction of electrophoresis was from top to bottom. Samples in the tracks were: A, keratin-like proteins in migrating corneal epithelium from an eye that had a 2-mm diameter central epithelial defect remaining; B, keratin-like proteins in regenerated corneal epithelium one day after healing; C, keratin-like proteins in regenerated corneal epithelium 10 days after healing; D, keratin-like proteins in normal corneal epithelium; E, total proteins in normal corneal epithelium; Std-standard proteins including phosphorylase B (92,500 MW), bovine serum albumin (66,200 MW), ovalbumin (45,000 MW) and carbonic anhydrase (31,000 MW). and Green.' These proteins are denned as keratin-like proteins. Although the preparatory technique and electrophoresis patterns are those of the keratins, our quantities did not allow further biochemical analysis, Using a modification of the techniques of Laemmli,12 vertical polyacrylamide slab-gels were constructed in a Bio-Rad Model 220 apparatus to contain 3.0% acrylamide in the stacking gels and 7.5% or 8.0% acrylamide in the running gels. Samples for electrophoresis contained 40 ng of protein from keratin-like protein samples, and 70 fig from total protein samples. The samples were reduced with 500 mm dithiothreitol at 37 C for 10 min before application to the gel.10 Electrophoresis was achieved with a pulsed constant power supply (Ortec 4100) at 320 volts. After electrophoresis, the gel was fixed with 50% trichloroacetic acid for 30 min, stained overnight with 0.1 % Coomassie blue in water/methanol/acetic acid (53:40:7), and destained with the same solution without dye. The reproducibility of these epithelial proteins was established by analyzing a minimum of two different samples for each category. The migration of these proteins was analyzed by linear regression from standard proteins, and the result was used to estimate the molecular weights of each protein band. Results Normal and Regenerated Corneal Epithelium PAGE of keratin-like proteins in the regenerated epithelium derived from cornea at various times after wounding possessed the same bands as that of normal corneal epithelium. By referring to Figure 1, those keratin-like proteins were two major (63K and 55K)
No. 5 579 KERATIN IN OCULAR SURFACE EPITHELIUM / Kinoshiro, er ol. and three minor bands (59K, 5IK and 36K). Those bands were in close agreement with the results of Doran et al.13 There was, in addition, one extra-band (58K) in migrating epithelium and in regenerated epithelium one day after healing (Fig. 1, Tracks AD). Total proteins of normal corneal epithelium showed many bands that were absent in PAGE of keratin-like proteins of normal corneal epithelium (Fig. 1, Tracks D and E). There were no remarkable changes in the bands of total proteins from regenerated epithelium at various times after healing. Normal and Regenerated Conjunctiva! Epithelium PAGE of keratin-like proteins in normal conjunctival epithelium showed major bands such as 58K and 48K, but not 63K or 55K, quite different patterns from that found with normal corneal epithelium (Fig. 2, Tracks A and F). In contrast to regenerated corneal epithelium, PAGE of regenerated epithelium derived from bulbar conjunctiva showed some differences in bands of keratin-like proteins with time after wounding. There were two major (58K and 48K) and five minor bands (63K, 59K, 5IK, 42K and 36K) seen in regenerated epithelium during migration and 10 days after healing (Fig. 2, Tracks B-D). The pattern of those keratin-like proteins was similar to that of normal conjunctival epithelium (Track A) and different from that of normal corneal epithelium (Track F). In addition, one band (5IK) was very prominent in migrating and regenerated epithelium one day after healing. Since conjunctival epithelium already had a prominent 58K band, we could not detect whether there was a more intense 58K band in migrating and regenerated epithelium one day after healing (compare Fig. 1, Tracks A and B). Regenerated conjunctival epithelium on the cornea 3 months after healing had two major bands (63K and 55K) and resembled normal corneal epithelium. There were, however, still some different keratin-like proteins, eg, the 58K and prominent 49K bands (Fig. 2, Tracks E and F). The only unique feature of the PAGE pattern of total proteins as compared to keratin-like proteins of regenerated conjunctival epithelium after healing was the presence of a 68K soluble band for up to 3 months. Discussion This study demonstrates clearly that although rabbit corneal and conjunctival epithelium are both nonkeratinizing ocular surface epithelia, those two types of epithelium in vivo show a remarkable divergence IL -48-42 -36 Std A B C O E F Fig. 2. An 8.0% polyacrylamide slab-gel electrophoresis of keratin-like proteins in regenerated conjunctival epithelium. Samples in the tracks were: Std-standard proteins including myosin (200,000 MW), b-galactosidase (116,250 MW), phosphorylase B (92,500 MW), bovine serum albumin (66,200 MW) and ovalbumin (45,000 MW); A, keratin-like proteins in normal conjunctival epithelium; B, keratin-like proteins in migrating conjunctival epithelium from an eye that had a 2-mm diameter central epithelial defect remaining; C, keratin-like proteins in regenerated conjunctiva] epithelium 1 day after healing; D, keratin-like proteins in regenerated conjunctival epithelium 10 days after healing; E, keratin-like proteins in regenerated conjunctival epithelium 3 months after healing; F, keratin-like proteins in normal cornea! epithelium. in keratin-like proteins. For example, PAGE of keratin-like proteins from normal conjunctival epithelium scraped in vivo shows different patterns from that of normal corneal epithelium. That there is a
580 INVESTIGATIVE OPHTHALMOLOGY & VISUAL SCIENCE / Moy 1983 Vol. 24 difference is confirmed by the observation that epithelia regenerated from cornea and conjunctiva are also different during the initial healing stages. These data are in contrast to previous studies in vitro that show that human corneal and conjunctival epithelial cells grown in tissue culture have identical keratinlike protein patterns on PAGE. 14 It is, however, supported by other in vivo studies in which corneal and conjunctival epithelia showed different patterns and intensities of staining to monoclonal antibodies to keratin, 15 and those two epithelia possess different proteins bands stained with monoclonal antibody to keratin. 6 It has been reported that when conjunctival epithelium grows over the cornea, it undergoes a biochemical and morphologic transformation into corneal epithelium. 916 Within the three months studied, the pattern of keratin-like proteins in the conjunctival epithelium underwent a substantial change when compared to the original in vivo pattern and began to resemble the pattern seen for keratin-like proteins of corneal epithelium in vivo (see Fig. 2). This suggests that the conversion of conjunctival to cornea may involve metabolic changes over and above those seen in previous studies of glucose metabolism. 16 The work indicates that the intracellular keratin framework of the epithelial cells changes as part of the process of transformation of conjunctival to corneal epithelium. Within the period we studied, however, the keratin-like proteins of the conjunctival epithelial cells have not completely transformed into those of corneal epithelial cells. Even after 3 months, it is still possible on the basis of the keratin-like protein banding patterns to distinguish which epithelial cells are derived from which ocular surface epithelium, corneal or conjunctival. Other studies also support the contention that transformation may not be completed within three months. In such studies, for up to 3 months, regenerated epithelium of conjunctival origin can be returned to its original histologic characteristics by either mechanical or immunological wounding. 1718 Thus, it appears that after total corneal and limbal epithelial loss with retention of corneal basement membrane, conjunctival epithelium may not transform completely into corneal epithelium. Superimposed on the changes in keratin-like protein banding patterns that occur as a result of the metaplasia of conjunctival into corneal epithelium, there are changes in protein patterns that appear to be associated with cell migration. For example, changes in keratin-like proteins are seen during the migration of the corneal epithelium to cover the central defects. As reference to Figure 1 shows, there is a striking band at 58K in migrating and just healed corneal epithelium. This is not present in normal corneal epithelium, and disappears only a few days after the defect is healed. Since migration and metaplasia are occurring at the same time in case of healing from conjunctival epithelium, it is difficult to separate which changes in keratin-like protein banding patterns are caused by which phenomenon. During migration and immediately after healing, there is a prominent 5IK band (and possibly a 58K band) in the regenerating conjunctival epithelium. As the 58K band in the migrating corneal epithelium, this 5IK band in the conjunctiva diminishes in intensity soon after wound closure, suggesting it too is related to migration. While the significance of the appearance of new protein bands is not clear, it is apparent that the migration and healing process itself is a stimulus for rapid synthesis or alteration of keratin-like proteins. Since migrating rat corneal epithelium incorporates more than ten times the amount of 14-Cleucine that normal corneal epithelium does, 19 it seems that migrating cells may synthesize new keratin-like proteins, possibly to change cytoplasmic filaments. Although this study does not provide direct evidence that those bands are essential for cell movement, it is interesting to speculate that they may play an important role in epithelial cell migration. Thus, it has been suggested that the phenotypes of the epithelia in vivo are predominantly determined by intrinsic divergence but that external environment or growth conditions can also affect differentiation of epithelial cells. 13 This study shows clearly that for regenerated epithelium derived from the conjunctiva growing over the cornea, the phenotype can be greatly regulated by the environment and may not be strictly determined by the intrinsic divergence. Key Words: ocular surface epithelium, migrating and regenerated epithelium, keratin, SDS polyacrylamide gel electrophoresis, rabbit, cornea, conjunctiva Acknowledgment The authors thank Dr. Joanne Blondin for critical reading and valuable suggestions of the manuscript. References 1. Sun TT and Green H: Keratin filaments of cultured human epidermal cells. Formation of intermolecular disulfide bonds during terminal differentiation. J Biol Chem 253:2053, 1978. 2. Franke WW, Schmid E, Osborn M, and Weber K: Different intermediate-sized filaments distinguished by immunofluorescence microscopy. Proc Natl Acad Sci USA 75:5034, 1978.
No. 5 KERATIN IN OCULAR SURFACE EPITHELIUM / Kinoshiro, er ol. 581 3. Sun TT and Green H: Immunofluorescent staining of keratin fibers in cultured cells. Cell 14:469, 1978. 4. Gipson IK: Cytoplasmic filaments: their role in motility and cell shape. Invest Ophthalmol Vis Sci 16:1081, 1977. 5. Lazarides E: Intermediate filaments as mechanical integrators of cellular space. Nature 283:249, 1980. 6. Sun TT, Tseng SCG, Nelson W, Jarvinen M, Woodcock- Mitchell J, and Huang, JW: Tissue distribution of keratin antigens: studies using monoclonal antibodies. ARVO Abstracts. Invest Ophthalmol Vis Sci 22 (suppl):73, 1982. 7. Thoft RA: Conjunctival transplantation. Arch Ophthalmol 95:1425, 1977. 8. Kinoshita S, Kiorpes TC, Friend J, and Thoft RA: Limbal epithelium in ocular surface wound healing. Invest Ophthalmol Vis Sci 23:73, 1982. 9. Shapiro MS, Friend J, and Thoft RA: Corneal re-epithelialization from the conjunctiva. Invest Ophthalmol Vis Sci 21:135, 1981. 10. Hassell JR, Newsome DA, and De Luca LM: Increased biosynthesis of specific glycoconjugates in rat corneal epithelium following treatment with Vitamin A. Invest Ophthalmol Vis Sci 19:642, 1980. 11. Lowry OH, Rosebrough NJ, Fair AL, and Randall JR: Protein measurement with the Folin phenol reagent. J Biol Chem 193:265, 1951. 12. Laemmli UK: Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680, 1970. 13. Doran TI, Vidrich A, and Sun TT: Intrinsic and extrinsic regulation of the differentiation of skin, corneal, and esophageal epithelial cells. Cell 22:17, 1980. 14. Sun TT and Green H: Cultured epithelial cells of cornea, conjunctiva and skin: absence of marked intrinsic divergence of their differentiated states. Nature 269:489, 1977. 15. Woodcock-Mitchell J and Sun TT: Studies of keratinization using monoclonal antibody to keratins. ARVO Abstracts. Invest Ophthalmol Vis Sci 20 (suppl):37, 1981. 16. Thoft RA and Friend J: Biochemical transformation of regenerating ocular surface epithelium. Invest Ophthalmol Vis Sci 16:14, 1977. 17. Thoft RA, Friend J, and Murphy HS: Ocular surface epithelium and corneal vascularization in rabbits. I. The role of wounding. Invest Ophthalmol Vis Sci 18:85, 1979. 18. Liu SH, Tagawa Y, Prendergast RA, Franklin RM, and Silverstein AM: Secretory component of IgA: a marker for differentiation of ocular epithelium. Invest Ophthalmol Vis Sci 20:100, 1981. 19. Gipson IK and Kiorpes TC: Epithelial sheet movement: Protein and glycoprotein synthesis. Dev Biol 92:259, 1982.