October 2003
Volume 44, Issue 10
Free
Cornea  |   October 2003
Molecular Mechanisms Controlling the Fibrotic Repair Phenotype in Cornea: Implications for Surgical Outcomes
Author Affiliations
  • Brian M. Stramer
    From the Evelyn F. and William L. McKnight Vision Research Center, Bascom Palmer Eye Institute, University of Miami School of Medicine, Miami Florida; the
  • James D. Zieske
    Schepens Eye Research Institute and Departments of Ophthalmology and Cell Biology, Harvard Medical School, Boston Massachusetts; and the
  • Jae-Chang Jung
    Department of Biology, College of Natural Sciences, Kyungpook National University, Taegu, Korea.
  • Jeffrey S. Austin
    From the Evelyn F. and William L. McKnight Vision Research Center, Bascom Palmer Eye Institute, University of Miami School of Medicine, Miami Florida; the
  • M. Elizabeth Fini
    From the Evelyn F. and William L. McKnight Vision Research Center, Bascom Palmer Eye Institute, University of Miami School of Medicine, Miami Florida; the
Investigative Ophthalmology & Visual Science October 2003, Vol.44, 4237-4246. doi:10.1167/iovs.02-1188
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to Subscribers Only
      Sign In or Create an Account ×
    • Get Citation

      Brian M. Stramer, James D. Zieske, Jae-Chang Jung, Jeffrey S. Austin, M. Elizabeth Fini; Molecular Mechanisms Controlling the Fibrotic Repair Phenotype in Cornea: Implications for Surgical Outcomes. Invest. Ophthalmol. Vis. Sci. 2003;44(10):4237-4246. doi: 10.1167/iovs.02-1188.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

purpose. Incisional or ablation injury to the corneal stroma is repaired by deposition of a fibrotic tissue produced by activated keratocytes, whereas cells lost from the underlying stroma after epithelial abrasion are simply replaced by keratocyte replication without expression of fibrotic markers. The purpose of this study was to investigate mechanisms that determine this differential keratocyte response.

methods. A penetrating keratectomy rabbit model was adapted for mice to study the fibrotic repair response. A mouse epithelial abrasion model was applied to study the stromal cell replacement response. A primary rabbit corneal cell culture model and an organotypic culture model were also used.

results. When the epithelium was prevented from resurfacing the cornea after penetrating keratectomy, expression of fibrotic markers was considerably reduced. TGF-β2 was determined to be a major substance produced by corneal epithelial cells capable of inducing the fibrotic phenotype. In the intact mouse cornea, TGF-β2 was confined to the uninjured epithelium, but was released into the stroma during fibrotic repair. By contrast, TGF-β1 was never found in the epithelium. When epithelial cells were cultured on a basement-membrane–like gel or allowed to deposit their own basement membrane in organotypic culture, TGF-β2 production was reduced. Return of a basement membrane after wounding in vivo correlated with loss of the fibrotic phenotype. In the epithelial debridement injury model in which the basement membrane was left intact, TGF-β2 remained confined to the corneal epithelium, consistent with the absence of a fibrotic phenotype.

conclusions. These data suggest that integrity of the basement membrane is a deciding factor in determining the regenerative character of corneal repair.

In response to injury, resident cells in tissues adjacent to the damaged area initiate a cascade of events necessary to restore tissue function. The typical result in the adult mammal is deposition of a fibrotic repair tissue that matures to a scar. However, there are examples in which repair can have a more regenerative character. Organs such as liver are capable of mounting a regenerative healing response, but repair becomes fibrotic during chronic injury. 1 Similarly, skin repair during embryogenesis, unlike in the adult, results in no significant scarring. 2 Therefore, responding resident cells are capable of choosing between different repair pathways. Understanding the mechanisms that determine which repair pathway is activated is crucial to regulating the healing response. 
The cornea’s simple cellular organization provides a particularly useful model for studying wound-healing events. Like skin, the cornea comprises a collagenous stroma surfaced by a squamous epithelium. Penetrating incision or ablation injury to the corneal stroma stimulates a typical fibrotic repair response involving hypercellularity, expression of smooth muscle actin, and deposition of a disorganized extracellular matrix (ECM). 3 4 5 This repair tissue, which is comparable to a spot weld, is remodeled over time into a mature scar. Its deposition is undesirable in the cornea, because it causes opacity and because the contraction associated with remodeling alters corneal curvature and the capacity to focus light on the retina. For this reason, surgical procedures to correct refractive error in the cornea must be designed to minimize the fibrotic repair response. 
The fibrotic response in vascularized tissues is controlled by bioactive substances released from platelets at the wound site, in particular, the cytokines platelet-derived growth factor (PDGF) and transforming growth factor (TGF)-β. Topical application of a panspecific neutralizing antibody to TGF-β after corneal injury reduces stromal fibrosis, indicating that similar agents are active in corneal repair. 6 However, the cornea contains no blood vessels and heals avascularly. Results of historic investigations have suggested that the epithelium may produce substances that substitute for those produced by platelets. 7 Recent clinical observation supports this hypothesis and suggests further that the key to limiting fibrosis is to reduce the potential for epithelial interaction with stromal cells. Thus, debridement of the corneal epithelium from its basement membrane causes underlying stromal cells to undergo apoptosis; however, these cells are simply replaced by mitosis, with no obvious hypercellularity or deposition of matrix. 8 It is only if epithelial debridement is followed by penetration of the basement membrane and stromal ablation, as in photorefractive keratectomy (PRK), that the fibrotic response is stimulated. In a newer corrective procedure, laser in situ keratomileusis (LASIK), which involves laser ablation beneath a stromal flap created by microkeratome, basement membrane is penetrated only around the edges of the flap. In this case, keratocytes that die around the ablation site are replaced without expression of fibrotic markers. 9 The outcome is a clearer cornea without the haze or remodeling associated with fibrotic repair. 
In the current study we investigated the nature of the molecular mediator of epithelial–stromal interaction controlling corneal fibrosis and the role of the epithelial basement membrane in limiting this response. 
Materials and Methods
Surgical Procedures
Surgical procedures were performed in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. CD1 mice 6 to 8 weeks old were anesthetized with an intraperitoneal injection of tribromoethanol (Avertin; Aldrich, Milwaukee, WI). Animals were then given a topical anesthetic of Aclaine (Alcon, Fort Worth, TX), and pupils were dilated with drops of 0.5% cyclopentolate (Cyclogyl; Alcon). 
A rabbit penetrating keratectomy model was adapted for mice to study fibrotic repair. 10 11 Briefly, a 1-mm full-thickness button of central corneal tissue including all three corneal tissue layers—epithelium, stroma, and endothelium—was ablated. The area was scored by insertion and removal of a 26-gauge needle into the central cornea. The full-thickness tissue button that was demarcated by the needle insertion was then removed with microdissecting scissors. After approximately 15 minutes, a fibrin clot formed in the space, enabling the anterior chamber to reform. 
A previously characterized mouse epithelial debridement model was used to investigate the stromal cell replacement response. The surgical procedure, as adapted for mice, was previously described. 12 Briefly, the central epithelium was debrided using an Alger brush within a 1.5-mm region, leaving the basement membrane intact. Within 1 day, stromal cells beneath the debrided area are lost by apoptosis. 8  
Immunohistochemical Analysis
After surgery, eyes were embedded in optimal cutting temperature (OCT) embedding medium (Tissue Tek, Elkhart, IN) for cryosectioning. Cryosections (10 μm) were prepared from frozen eyes. For visualizing smooth muscle actin (α-sm actin) and filamentous actin, sections were fixed in 4% paraformaldehyde (PFA) for 10 minutes. Direct immunofluorescence for α-sm actin was performed with a primary FITC-conjugated monoclonal anti-α-sm actin antibody (Sigma-Aldrich, St. Louis, MO). Filamentous actin was visualized with rhodamine-conjugated phalloidin (Molecular Probes, Eugene, OR). For TGF-β1 and -β2 indirect immunohistochemistry, tissue was fixed in ice-cold acetone for 10 minutes and incubated with rabbit polyclonal antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) overnight. Secondary detection used an anti-rabbit Vectastain ABC kit (Vector Laboratories, Burlingame, CA). Basement membranes were detected by indirect immunofluorescence. For detection of entactin, sections were fixed in ice-cold acetone and incubated with a monoclonal anti-entactin antibody (Chemicon, Temecula, CA) followed by secondary anti-rat rhodamine (Chemicon). For detection of laminin, sections were fixed in 3% PFA and incubated with a sheep polyclonal antiserum raised against mouse laminin (a gift from Ilene Gipson, Schepens Eye Research Institute, Boston, MA), followed by a FITC-conjugated donkey anti-sheep antibody (Jackson ImmunoResearch, Inc., West Grove, PA). Before mounting immunofluorescent sections, nuclei were localized by staining with 4′, 6-diamidino-2-phenylindole (DAPI; 10 ng/mL). Images of sections were captured on a microscope (Nikon, Tokyo, Japan). Images for each experiment were obtained using the same camera settings. Quantification of immunofluorescence was performed on computer (Image Pro-Plus; MediaCybernetics, Silver Spring, MD). 
Cell Culture and Organotypic Culture
Eyes of New Zealand White rabbits were obtained from Pel Freez Co. (Rogers, AR), and primary stromal and epithelial cell cultures were prepared as previously described. 13 Corneal organotypic cocultures were also made as previously described. 14 Briefly, a mouse corneal endothelial cell line 15 was seeded atop a polycarbonate membrane. A type I collagen solution was seeded with rabbit corneal stromal cells and placed over the endothelial cells. Rabbit corneal epithelial cells were then plated atop the construct, thus mimicking corneal structure. Some cultures were treated with TGF-β2 (Sigma-Aldrich). 
Epithelial cell cultures were plated on plastic tissue culture dishes, synthetic matrix-coated plates (Matrigel), or plates coated with collagen (BD Biosciences, Bedford, MA). For epithelial cell–conditioned media, epithelial cells were plated in 10-cm tissue culture dishes in serum-free medium and allowed to condition the medium for 72 hours. Medium was then collected and cell debris was spun down. Media were stored at −80°C for further use. For neutralizing antibody experiments anti-TGF-β1 (AF-101-NA; R&D Systems, Minneapolis, MN) and anti-TGF-β2 (AB-112-NA; R&D Systems) was used to treat conditioned epithelial medium before use. Anti-TGF-β1 was used at 0.2 μg/mL and anti-TGF-β2 was used at 2 μg/mL. Antibody concentrations were chosen to give 50% maximum inhibition of cell replication when tested on an indicator cell line. 
Western Blot Analysis
Cells were plated in six-well dishes at 1 × 106 cells per well and allowed to condition the media for 48 hours. Media were then collected, and secreted proteins were precipitated with 10% ice-cold trichloroacetic acid (TCA). Proteins were redissolved in radioimmunoprecipitation assay (RIPA) buffer (1% NP-40, 0.5% deoxycholate, 0.1% SDS). Total secreted protein from equivalent cell numbers was then run on a 12% polyacrylamide gel for Western analysis. For cell lysates, stromal cells or epithelial cells were digested in RIPA with a protease inhibitor cocktail (Complete mini; Roche Diagnostics, Indianapolis, IN). Epithelial cell lysates were then TCA precipitated before Western blot analysis. For Western analysis, anti-TGF-β1 (AF-101-NA; R&D Systems) was used at 1:1000, anti-TGF-β2 (AB-112-NA; R&D Systems) at 1:500, and anti α-sm actin (Sigma-Aldrich) at 1:3000. 
Results
Characterization of the Fibrotic Response in Cornea
To investigate the mechanisms controlling corneal fibrosis in vivo, we adapted a rabbit model of penetrating keratectomy 11 to the mouse by reducing the size of the central button removed from the cornea to 1 mm. Representative results are shown in Figure 1A . Histologic analyses at days 5, 7, and 14 after surgery revealed a hypercellular repair tissue deposited in the ablated region of the cornea. The wound region at days 5 and 7 after surgery consisted of both fibroblasts and inflammatory cells, as examined by immunostaining for Mac1, a marker of both monocytes and polymorphonuclear cells (data not shown). Mac1-positive cells had essentially disappeared by day 14; however, the wound region remained hypercellular compared with control corneas throughout the time course. The character of repair was examined by staining with rhodamine phalloidin (Fig. 1 , red) to identify filamentous actin stress fibers characteristic of active fibroblasts, 16 and by immunostaining for α-sm actin (Fig. 1 , green), a marker of fibrotic repair. Control corneas showed no staining for either marker, nor were the uninjured areas of tissue stained in debrided corneas (not shown). Corneal stromas were still negative for the repair markers at day 5 after injury. However, by day 7, there was robust staining for both markers in the hypercellular repair tissue and stress fibers strongly colocalized with α-sm actin. By day 14 after injury, staining for both markers had regressed and appeared only beneath the regenerated epithelium, even though the tissue remained hypercellular. These data show that the fibrotic phenotype is temporally and spatially regulated during corneal wound repair. 
To examine the contribution of the epithelium to the fibrotic phenotype, we limited the exposure of the repairing stromal region to the overlying epithelium by debriding the regenerating epithelium daily beginning at day 3. Representative results are shown in Figure 1B . Staining for α-sm actin was significantly reduced by this procedure. Quantification of mean fluorescence intensity revealed a 60% reduction in α-sm actin staining (P < 0.01, Student’s paired t-test). Stress fiber organization was also reduced in debrided corneas but to a lesser extent, a 40% reduction was observed (P < 0.02, Student’s paired t-test). DAPI staining revealed an abundance of cells in both debrided and control corneas. Immunostaining for the Mac1 marker indicated that debridement causes an increase in inflammatory cell infiltration in the anterior aspect of the stroma. Nevertheless, there were many DAPI staining nuclei outside the region of Mac1 staining, indicating that loss of fibrotic markers is not due to a loss of repair cells. These data suggest that the repair cell phenotype is controlled by exposure to the regenerating epithelium. 
In contrast to skin, the corneal stroma contains a homogenous population of stromal cells (keratocytes) without specialized functions, providing a relatively pure population of cells for isolation into culture. Using this cell culture model, epithelial factors have been identified that control aspects of stromal cell activation in vitro. 17 In an attempt to identify the epithelial factors controlling expression of fibrotic markers, we used an in vitro assay. Rabbit corneal stromal cells freshly isolated from the corneal stroma and cultured in the absence of serum have a dendritic appearance, true to their in vivo phenotype in the uninjured cornea 16 (Fig. 2A) . Treatment of cells with TGF-β2 stimulated cellular spreading to a fibroblast spindle shape (representative results shown in Fig. 2A ) and development of stress fibers as examined by rhodamine phalloidin staining (data not shown). Treatment of stromal cells with conditioned media from cultured rabbit corneal epithelial cells induced a similar change in phenotype as treatment with TGF-β2. Addition of a neutralizing antibody to TGF-β2 but not TGF-β1 inhibited the phenotypic change. Treatment of stromal cells with conditioned media induced a level of α-sm actin expression similar to treatment with TGF-β (Fig. 2B) . Neutralizing antisera to TGF-β2 but not the TGF-β1 isoform, inhibited the potential of conditioned epithelial media to induce α-sm actin, suggesting TGF-β2 was at least one of the active agents. 
To determine whether the epithelium was directly responsible for the TGF-β2 production, corneal epithelial– and stromal cell–conditioned media were examined by Western analysis for TGF-β isoforms. TGF-β is secreted as an inactive high molecular weight complex that is processed to a 25-kDa active form. The TGF-β detected by Western analysis was 25 kDa representing active TGF-β. Low pH is capable of releasing the active form of TGF-β from the latent complex 18 and therefore it is likely that the TCA precipitation used to concentrate the protein artificially activated TGF-β. Other methods of protein concentration did not allow us to see TGF-β by Western analysis using these antibodies, and therefore it is probable that most of the TGF-β present in the culture media was actually the inactive protein (data not shown). Epithelial cultures produced approximately 10 ng of TGF-β2 per 106 cells during the 2 days of culture compared with stromal cells, which produced no detectable TGF-β2 (Fig. 2C) . In contrast, both epithelial and stromal cultures produced little TGF-β1, which could be detected only after long exposure (Fig. 2C) . These data show that, TGF-β2 is produced by corneal epithelium and is the predominant isoform produced by either corneal stromal or epithelial cells. 
Spatial and Temporal Patterns of TGF-β Isoforms during Corneal Wound Healing
To determine the sources of TGF-βs during fibrotic corneal wound repair in vivo, we used immunohistochemistry. Immunoreactive TGF-β2 was present in the unwounded corneal epithelium and was expressed in this tissue throughout healing (Fig. 3A) . By day 3 after surgery, TGF-β2 localized to the fibrin clot within the wound region. Day 7 showed strong TGF-β2 immunolocalization to the region directly beneath the epithelium within the wound area. Positive immunostaining remained within the wound region through day 14 and subsequently decreased until no TGF-β2 was detected by day 24. 
In contrast to TGF-β2 localization, TGF-β1 was never detected in the unwounded or wounded epithelium (Fig. 3B) . TGF-β1 first localized to the cornea at day 3 after wounding, associated with the fibrin clot and its infiltrated inflammatory cells. Day 7 showed strong TGF-β1 localization extending from the posterior region of the stroma and along the endothelium outside the wound region. By day 14, TGF-β1 staining was absent from the central corneal stroma, but remained associated with the posterior region along the wounded endothelial region. This strong endothelial staining remained through the 24 days of the study. In contrast to TGF-β1, the presence of TGF-β2 in the epithelium and the strong localization beneath the wounded epithelium was further evidence that the epithelial–stromal interaction involved mainly the TGF-β2 isoform. 
Effect of Basement Membrane Regulation of Epithelial TGF-β2 on Fibrotic Activation
We questioned whether a proper epithelial substratum was capable of regulating release of TGF-β2 into the stroma. Corneal epithelial cells were plated either on plastic culture dishes or basement membrane matrix (Matrigel; BD Bioscience)–coated tissue culture dishes and conditioned media were examined for the presence of TGF-β isoforms (Fig. 4A) . TGF-β1 release from corneal epithelial cells was variable and not regulated by plating on the matrix. In contrast, TGF-β2 was consistently downregulated in corneal epithelial cell–conditioned media when plated on synthetic matrix. Plating epithelial cells on collagen I, collagen IV, or laminin did not show a similar effect (data not shown). We then compared internal and external cellular pools of TGF-β2 by Western analysis to determine whether the decrease in detectable TGF-β2 in media of epithelial cells plated on synthetic matrix reflected a decrease in protein expression or a decrease in protein secretion. Epithelial cells plated on matrix decreased TGF-β2 levels intracellularly and extracellularly to a similar level (Fig. 4B) . This suggested that the decrease in secreted protein was due to a decrease in cellular production of the TGF-β2 isoform. 
We then investigated whether endogenous basement membrane production regulates the epithelial–stromal interaction. For these experiments we used an in vitro organotypic coculture system in which basement membrane synthesis can be controlled by manipulating the culture conditions. 14 Representative results are shown in Figure 4 . Epithelial cells produce a complete basement membrane when plated on a floating type I collagen substratum embedded with fibroblasts and backed with a lining of endothelial cells, mimicking corneal structure. As we have reported, 14 a complete basement membrane fails to form in the absence of endothelial cells. This was assessed by both electron microscopy and immunofluorescent localization of basement membrane components (Fig. 4C) . When organotypic cultures were created without endothelial cells and thus no basement membrane, stromal cells beneath the epithelium within the collagen matrix stained strongly for α-sm actin (Fig. 4C) . In the presence of endothelial cells, and therefore a complete basement membrane, stromal cells showed no staining for α-sm actin. These two types of organotypic cultures were immunostained for the TGF-β2 isoform (Fig. 4C) . In the absence of basement membrane, the stroma stained strongly for TGF-β2. In the presence of basement membrane, TGF-β2 was absent in the stroma, and only the epithelium stained positive, mimicking normal unwounded cornea. Therefore, endogenous basement membrane formation in organotypic coculture prevents fibrotic marker expression through inhibition of TGF-β2 release into the stroma. 
We then examined fibrotic corneal repair in vivo to determine whether return of the basement membrane after wounding correlates with a reduction in α-sm actin–positive stromal cells. Basement membrane was examined by immunostaining for entactin, a known corneal epithelial basement membrane component. Entactin localized to the unwounded cornea as a band directly beneath the epithelium, with no staining for α-sm actin within the stromal region (Fig. 5) . Seven days after wounding, entactin was absent from the area beneath the epithelium; however, a diffuse haze was present in the corneal stroma. The stromal staining of entactin was not surprising, as entactin is reported to be produced predominantly by mesenchymal cells in other organs during development 19 and is expressed in granulation tissue during skin repair until healing is complete. 20 During the complete absence of entactin localization to the basement membrane region at day 7, α-sm actin staining was strongest within the corneal stroma. By 10 days, entactin staining was strong throughout the stromal region, with a slightly more intense staining beneath the epithelium. At the 10-day time point, α-sm actin expression was still strong within the stroma. Finally, by 14 days, entactin relocalized as a band of intense staining directly beneath the epithelium with a marked reduction of α-sm actin in the stroma. These data show that, α-sm actin expression in the corneal stroma during fibrotic wound repair was inversely correlated with the return of basement membrane, as examined by immunolocalization of entactin. Furthermore, basement membrane return after wounding at approximately day 14 in the mouse model correlated with the beginning decline in TGF-β2 levels within the wound as examined by immunohistochemistry (Fig. 3A)
As discussed in the introduction, not all corneal wounds result in a fibrotic healing response. Epithelial debridement in the absence of basement membrane damage, causes a loss of corneal stromal cells beneath the wound area due to apoptosis. 21 Stromal cells temporarily become active and enter the cell cycle; however, rather than the hypercellular stroma and the accumulation of ECM that develops during fibrotic repair, the result is a complete cellular replacement, with maintenance of corneal clarity. 8 We hypothesized that this regenerative stromal cell response was due to maintenance of the basement membrane. To test this hypothesis, we performed epithelial debridement surgery and observed repair markers over time by staining with rhodamine phalloidin (Fig. 6) . Epithelial debridement wounds maintain the basement membrane as examined by entactin staining after epithelial removal. One hour after debridement, stromal cells were still present within the corneal stroma. By 18 hours, stromal cells were absent beneath the epithelium and adjacent stromal cells stained strongly for filamentous actin by rhodamine phalloidin staining. Actin rearrangement was transient, in that phalloidin staining was absent by day 3. Despite staining strongly for filamentous actin, stromal cells did not express α-sm actin at any time during the period when cells were undergoing replacement. Immunohistochemistry for TGF-β2 revealed that unlike in fibrotic repair, this isoform remains localized to the epithelium and was never detected in the stroma during this more regenerative form of repair. The lack of TGF-β2 in the stroma probably explains the failure of stromal cells to express α-sm actin after epithelial debridement. 
Discussion
Experimental investigation and clinical observation have suggested that an epithelial–stromal interaction mediates fibrotic repair in the cornea, where healing occurs completely avascularly. Several secreted factors produced by the epithelium that are capable of controlling fibrotic phenotypes in cell culture have been identified. 7 17 However, functionality of these factors has not been shown in vivo. Herein, we report the first direct evidence that an epithelial–stromal interaction controls aspects of the fibrotic phenotype, and we identify a major mediator as TGF-β2. We provide important evidence that this tissue interaction is regulated by integrity of the epithelial basement membrane and that this determines whether the fate of the repair process is fibrotic or regenerative in nature. 
Basement Membrane Regulation of Corneal Fibrosis
Corneal epithelial cell release of TGF-β2 was downregulated by plating cells on reconstituted basement membrane, as well as when epithelial cells were allowed to produce endogenous basement membrane. Furthermore, reappearance of an intact basement membrane during fibrotic wound repair in vivo correlated with loss of α-sm actin expression and TGF-β2 localization within the wound region. Therefore, in our current model, epithelial migration over a non–basement-membrane matrix signals an increase in production and release of the TGF-β2 isoform. The return of the basement membrane during healing then signals a decrease and possibly creates a physical barrier to TGF-β2. The fibrotic phenotype is unstable, because removal of TGF-β from cultures results in reduced expression of α-sm actin and stress fibers. 22 Therefore, during corneal repair the loss of TGF-β in the stroma due to return of the basement membrane would result in a loss of the fibrotic phenotype. 
Not all types of corneal injuries result in a fibrotic stromal cell response. Epithelial damage without basement membrane loss results in activation of adjacent stromal cells, the result is a complete cellular replacement without fibrotic marker expression and no further matrix remodeling. 8 During this regenerative response, the basement membrane is left intact, which explains the absence of TGF-β2 and α-sm actin expression in the stroma. Therefore, basement membrane integrity may be a deciding factor in the extent of fibrosis that develops during corneal wound repair. 
The role of TGF-β1 in corneal repair was not addressed by the experiments in this study, because our findings did not implicate this isoform in controlling the epithelial–stromal interaction under study. The localization of TGF-β1 within the early fibrin clot suggests that it is produced by infiltrating inflammatory cells. It is likely that it is involved in the resolution of the inflammatory response, as has been shown in other wound models. At later times during the repair process, TGF-β1 was localized to the endothelium, where it may be involved in endothelial repair. 23 24 25  
Developmental Correlation
Despite seemingly redundant effects, TGF-β1 and TGF-β2 have different expression patterns during embryogenesis and wound repair in a number of organs. 26 27 28 29 Furthermore, knockout mice have no overlapping phenotypes. Unlike TGF-β1 deficiency which results in 50% prenatal lethality due to defective hematopoiesis and vasculogenesis, TGF-β2 deficiency results in multiple developmental defects in numerous organs, including heart, craniofacial, inner ear, and cornea. 30 31 32 33 The developmental defects affected by TGF-β2 deficiency have led to the hypothesis that this isoform is involved in epithelial–mesenchymal interactions in cell growth, extracellular matrix production, and tissue remodeling during development. 30 This interaction is probably necessary during corneal stroma development where TGF-β2 knockout mice show a failure to form the corneal stroma due to a significant decrease in stromal cell synthesis of matrix components. 31 Furthermore TGF-β2, unlike TGF-β1, is expressed constitutively in the epithelium. 31 34 Therefore, our data suggest that the epithelial–stromal interaction during corneal repair in the adult is a recapitulation of a developmental event with an altered stromal cell response. Understanding why developmental exposure to TGF-β2 leads to proper corneal stromal development when exposure in the adult leads to fibrosis may shed light on the ability of fetuses to heal scar free. Whether mesenchymal stromal cells have an innate response to TGF-β different from the adult, or some environmental cofactor present or absent during development, such as an ECM component, decides the fate of the TGF-β response needs further elucidation. 
Basement membrane regulation of stromal phenotypes during wound repair may similarly occur during corneal stromal developmental. TGF-β2 production, presumably from the corneal epithelium, is necessary for mesenchymal cells to form the corneal stroma properly, beginning at approximately embryonic day (E)12.5 of mouse development. 31 Basement membrane during this time is not complete, as examined by immunolocalization of matrix components. Laminin 5, an ECM molecule in hemidesmosomes, does not localize to the corneal basement membrane until E16. 35 Therefore, it is possible that an incomplete epithelial basement membrane during embryogenesis allows for the release of TGF-β2 into the stroma, signaling its development while formation of a proper basement membrane at a later time point signals cessation. 
Clinical Significance
Historic studies have demonstrated that changes in corneal stromal cells associated with fibrotic activation fail to occur in the absence of an overlying epithelium. 7 The clinical impression has supported the concept that the corneal epithelium controls fibrotic activation. 5 9 Our data suggest that the mechanism for this difference in refractive surgeries is the maintenance of the basement membrane, preventing release of the TGF-β2 isoform. This would explain the lower degree of surgical complication due to scarring in LASIK compared with other procedures such as PRK. 36 An even newer procedure, laser subepithelial keratomileusis (LASEK), may reduce haze by the same mechanism. This procedure uses alcohol treatment of the ocular surface to displace the epithelium as an intact flap, which is then folded back into place after laser ablation is performed. A recent study indicates that the basement membrane splits from the underlying stroma with alcohol treatment and stays with the epithelium. 37 Overall, our findings have implications for further refinement of corneal refractive surgical procedures. 
The importance of a proper epithelial substratum for maintaining stromal homeostasis is supported by the results of studies suggesting that persistent corneal epithelial defects can be healed by application of amniotic membrane, which contains a prominent basement membrane. Amniotic membrane aids re-epithelialization and reduces inflammation and scarring. 38 Plating epithelial cells on amniotic membrane matrix is reported to reduce expression of IL-1α, a cytokine hypothesized to be involved in epithelial–stromal interactions in the cornea. 39 It would be interesting to determine whether amniotic membrane has a similar effect on epithelial TGF-β2 production, which could play a role in the membrane’s antiscarring properties. 
Investigations into epithelial–stromal interactions during wound repair are not specific to the cornea and have been shown to correlate with pathologic healing in other organs. Bronchial changes during pathogenesis of asthma, cystic fibrosis, and chronic obstructive pulmonary disease (COPD) have been postulated to involve a similar tissue-inductive event. 40 41 Bronchial epithelial cells, similar to corneal epithelium, express factors capable of regulating aspects of bronchial stromal cell wound activation. 42 43 44 Bronchial epithelial cell cultures also spontaneously produce predominantly the TGF-β2 isoform in culture. 44 Furthermore, bronchial stromal cell induction of α-sm actin expression 45 and an abnormal basement membrane 46 47 correlate with disease. Epithelial abnormalities have similarly been shown in hypertrophic scarring in skin where the epithelium is reported “activated” and expressing an increase in cell cycle markers and keratins. 48 49 Furthermore, in a porcine model of skin repair, TGF-β2 is the first isoform expressed in the keratinocytes, whereas in mice, this isoform was reported to be abundantly expressed beneath the hyperproliferative epithelium. 28 29 Therefore a similar epithelial/stromal interaction involving the TGF-β2 isoform may occur in skin as well. 
Conclusion
The cellular choice between a fibrotic and regenerative healing response has been investigated in other systems such as skin, where repair during embryogenesis, unlike in the adult, occurs scar free. In this case the choice between a regenerative and fibrotic response is hypothesized to involve inflammation, because there is a correlation between the onset of scarring and the age during development at which the organism is capable of mounting an inflammatory response. 50 51 This is unlikely to be a factor in the cornea models compared herein, where both regenerative and fibrotic repair involves an influx of inflammatory cells. Here we have shown the novel finding that one fibrotic switch during corneal repair involves loss of the epithelial basement membrane, which induces stromal exposure to TGF-β2. This suggests that fibrotic mechanisms may be tissue specific, and their control will require tissue-specific studies to identify the inducing paracrine factors and reasons for their production. 
 
Figure 1.
 
Reduction of stromal exposure to epithelium after penetrating keratectomy decreases the appearance of fibrotic markers. The quality of repair after penetrating keratectomy was examined by immunofluorescent localization of α-sm actin (green), which is characteristic of fibrotic repair and by using rhodamine phalloidin (red) to stain filamentous actin stress fibers, which are characteristic of active fibroblasts. Monocyte cell populations were localized by immunostaining for Mac1 (red). Cell nuclei were stained with DAPI (blue) to delineate the location of cells. (A) Time course of marker localization in corneas repairing after penetrating keratectomy. Bracket: epithelial tissue layer. (B) Effect of epithelial removal on marker localization in corneas repairing after penetrating keratectomy. The epithelium of repairing corneas was removed daily beginning at day 3 after surgery. Mean fluorescence intensity for markers was quantified and compared with a control group that did not undergo epithelial removal at day 7 after penetrating keratectomy. Only the stromal tissue layer is visible in these images. Scale bar, 50 μm.
Figure 1.
 
Reduction of stromal exposure to epithelium after penetrating keratectomy decreases the appearance of fibrotic markers. The quality of repair after penetrating keratectomy was examined by immunofluorescent localization of α-sm actin (green), which is characteristic of fibrotic repair and by using rhodamine phalloidin (red) to stain filamentous actin stress fibers, which are characteristic of active fibroblasts. Monocyte cell populations were localized by immunostaining for Mac1 (red). Cell nuclei were stained with DAPI (blue) to delineate the location of cells. (A) Time course of marker localization in corneas repairing after penetrating keratectomy. Bracket: epithelial tissue layer. (B) Effect of epithelial removal on marker localization in corneas repairing after penetrating keratectomy. The epithelium of repairing corneas was removed daily beginning at day 3 after surgery. Mean fluorescence intensity for markers was quantified and compared with a control group that did not undergo epithelial removal at day 7 after penetrating keratectomy. Only the stromal tissue layer is visible in these images. Scale bar, 50 μm.
Figure 2.
 
Conditioned epithelial media were capable of inducing the fibrotic phenotype in cultured corneal stromal cells through TGF-β2. (A) Subcultured rabbit corneal stromal cells were serum starved for 48 hours, and then treated with either 2 ng/mL TGF-β2, conditioned media from corneal epithelial cells, or conditioned media containing neutralizing antibodies specific for TGF-β1 and TGF-β2 isoforms. Arrows: indicate cells with normal keratocyte phenotype; (✶): cells with fibrotic phenotype. (B) Stromal cell lysates from the experiment were assayed for α-sm actin content by Western analysis. (C) Conditioned media prepared from 1 × 106 epithelial cells or stromal cells were harvested, and proteins were concentrated and examined by Western analysis for TGF-β expression. Purified TGF-β1 and -β2 (10 ng) were run as a positive control.
Figure 2.
 
Conditioned epithelial media were capable of inducing the fibrotic phenotype in cultured corneal stromal cells through TGF-β2. (A) Subcultured rabbit corneal stromal cells were serum starved for 48 hours, and then treated with either 2 ng/mL TGF-β2, conditioned media from corneal epithelial cells, or conditioned media containing neutralizing antibodies specific for TGF-β1 and TGF-β2 isoforms. Arrows: indicate cells with normal keratocyte phenotype; (✶): cells with fibrotic phenotype. (B) Stromal cell lysates from the experiment were assayed for α-sm actin content by Western analysis. (C) Conditioned media prepared from 1 × 106 epithelial cells or stromal cells were harvested, and proteins were concentrated and examined by Western analysis for TGF-β expression. Purified TGF-β1 and -β2 (10 ng) were run as a positive control.
Figure 3.
 
TGF-β isoforms were differentially localized during fibrotic corneal repair. TGF-β2 (A) and -β1 (B) isoforms were localized in repairing mouse corneal wounds by substrate immunohistochemistry (brown). Sections were counterstained with hematoxylin. Arrows: wound margin; brackets: epithelium. Scale bar, 50 μm.
Figure 3.
 
TGF-β isoforms were differentially localized during fibrotic corneal repair. TGF-β2 (A) and -β1 (B) isoforms were localized in repairing mouse corneal wounds by substrate immunohistochemistry (brown). Sections were counterstained with hematoxylin. Arrows: wound margin; brackets: epithelium. Scale bar, 50 μm.
Figure 4.
 
Epithelial–stromal interaction was inhibited by basement membrane. (A) Epithelial cells were plated in duplicate on either plastic or synthetic basement membrane, and their ability to release TGF-β isoforms was examined by Western analysis. (B) TGF-β2 from conditioned media was compared with cellular levels by Western analysis. (C) Corneal organotypic cultures were created without an endothelial layer or with an endothelial layer to stimulate formation of an epithelial basement membrane. The presence of basement membrane was confirmed by indirect immunofluorescence localization of laminin. Stromal cell myofibroblast transformation was examined by immunofluorescence for α-sm actin, and TGF-β2 was localized by substrate immunohistochemistry. TGF-β2–stained sections were counterstained with hematoxylin. Brackets: epithelial layer. Scale bar, 50 μm.
Figure 4.
 
Epithelial–stromal interaction was inhibited by basement membrane. (A) Epithelial cells were plated in duplicate on either plastic or synthetic basement membrane, and their ability to release TGF-β isoforms was examined by Western analysis. (B) TGF-β2 from conditioned media was compared with cellular levels by Western analysis. (C) Corneal organotypic cultures were created without an endothelial layer or with an endothelial layer to stimulate formation of an epithelial basement membrane. The presence of basement membrane was confirmed by indirect immunofluorescence localization of laminin. Stromal cell myofibroblast transformation was examined by immunofluorescence for α-sm actin, and TGF-β2 was localized by substrate immunohistochemistry. TGF-β2–stained sections were counterstained with hematoxylin. Brackets: epithelial layer. Scale bar, 50 μm.
Figure 5.
 
Lack of basement membrane in vivo during corneal repair correlated with myofibroblast transformation. Basement membrane return after corneal wounding was examined by immunolocalization of entactin (red) and correlated with myofibroblast transformation, as examined by α-sm actin immunofluorescence (green). Nuclei were localized by staining with DAPI (blue). Arrows: basement membrane region. Scale bar, 50 μm.
Figure 5.
 
Lack of basement membrane in vivo during corneal repair correlated with myofibroblast transformation. Basement membrane return after corneal wounding was examined by immunolocalization of entactin (red) and correlated with myofibroblast transformation, as examined by α-sm actin immunofluorescence (green). Nuclei were localized by staining with DAPI (blue). Arrows: basement membrane region. Scale bar, 50 μm.
Figure 6.
 
Stromal cell activation in the absence of basement membrane destruction occurred without myofibroblast transformation. The central corneal epithelium was debrided and the time course of stromal activation was examined by staining for filamentous actin with rhodamine phalloidin. Myofibroblast induction was examined by immunofluorescence for α-sm actin. The iris, which contains smooth muscle cells, was used as an internal positive control for α-sm actin. TGF-β2 was localized by substrate immunohistochemistry. Nuclei were localized by staining with DAPI (blue) in phalloidin staining. TGF-β2 immunohistochemistry was counterstained with hematoxylin. Arrows: basement membrane region. Scale bar, 50 μm.
Figure 6.
 
Stromal cell activation in the absence of basement membrane destruction occurred without myofibroblast transformation. The central corneal epithelium was debrided and the time course of stromal activation was examined by staining for filamentous actin with rhodamine phalloidin. Myofibroblast induction was examined by immunofluorescence for α-sm actin. The iris, which contains smooth muscle cells, was used as an internal positive control for α-sm actin. TGF-β2 was localized by substrate immunohistochemistry. Nuclei were localized by staining with DAPI (blue) in phalloidin staining. TGF-β2 immunohistochemistry was counterstained with hematoxylin. Arrows: basement membrane region. Scale bar, 50 μm.
The authors thank Brad Coyle for editorial assistance. 
Bisgaard, HC, Thorgeirsson, SS. (1996) Hepatic regeneration. The role of regeneration in pathogenesis of chronic liver diseases Clin Lab Med 16,325-339 [PubMed]
McCallion, RL, Ferguson, MWJ. (1996) Fetal Wound Repair Clark, RAF eds. The Molecular and Cellular Biology of Wound Repair ,561-600 Plenum Press New York.
Garana, RM, Petroll, WM, Chen, WT, Herman, IM, Barry, P, Andrews, P, Cavanagh, HD, Jester, JV. (1992) Radial keratotomy. II. Role of the myofibroblast in corneal wound contraction Invest Ophthalmol Vis Sci 33,3271-3282 [PubMed]
Cintron, C, Szanier, BR, Hassinger, LC, Kublin, C. (1982) Scanning electron microscopy of rabbit corneal scars Invest Ophthalmol Vis Sci 23,50-63 [PubMed]
Moller-Pedersen, T, Li, HF, Petroll, WM, Cavanagh, HD, Jester, JV. (1998) Confocal microscopic characterization of wound repair after photorefractive keratectomy Invest Ophthalmol Vis Sci 39,487-501 [PubMed]
Moller-Pedersen, T, Cavanagh, HD, Petroll, WM, Jester, JV. (1998) Neutralizing antibody to TGFbeta modulates stromal fibrosis but not regression of photoablative effect following PRK Curr Eye Res 17,736-747 [CrossRef] [PubMed]
Weimar, V. (1960) Healing processes in the cornea Duke-Elder, S Perkins, ES eds. The Transparency of the Cornea ,111-124 Blackwell Scientific Publications Oxford, UK.
Zieske, JD, Guimaraes, SR, Hutcheon, AE. (2001) Kinetics of keratocyte proliferation in response to epithelial debridement Exp Eye Res 72,33-39 [CrossRef] [PubMed]
Nakamura, K, Kurosaka, D, Bissen-Miyajima, H, Tsubota, K. (2001) Intact corneal epithelium is essential for the prevention of stromal haze after laser assisted in situ keratomileusis Br J Ophthalmol 85,209-213 [CrossRef] [PubMed]
West-Mays, JA, Sadow, PM, Tobin, TW, Strissel, KJ, Cintron, C, Fini, ME. (1997) Repair phenotype in corneal fibroblasts is controlled by an interleukin-1 alpha autocrine feedback loop Invest Ophthalmol Vis Sci 38,1367-1379 [PubMed]
Cintron, C, Schneider, H, Kublin, C. (1973) Corneal scar formation Exp Eye Res 17,251-259 [CrossRef] [PubMed]
Mohan, R, Rinehart, WB, Bargagna-Mohan, P, Fini, ME. (1998) Gelatinase B/lacZ transgenic mice, a model for mapping gelatinase B expression during developmental and injury-related tissue remodeling J Biol Chem 273,25903-25914 [CrossRef] [PubMed]
Strissel, KJ, Rinehart, WB, Fini, ME. (1995) A corneal epithelial inhibitor of stromal cell collagenase synthesis identified as TGF-beta 2 Invest Ophthalmol Vis Sci 36,151-162 [PubMed]
Zieske, JD, Mason, VS, Wasson, ME, et al (1994) Basement membrane assembly and differentiation of cultured corneal cells: importance of culture environment and endothelial cell interaction Exp Cell Res 214,621-633 [CrossRef] [PubMed]
Muragaki, Y, Shiota, C, Inoue, M, Ooshima, A, Olsen, BR, Ninomiya, Y. (1992) Alpha 1 (VIII)-collagen gene transcripts encode a short chain collagen polypeptide and are expressed by various epithelial, endothelial and mesenchymal cells in a new born mouse tissues Eur J Biochem 207,895-902 [CrossRef] [PubMed]
Jester, JV, Barry, PA, Lind, GJ, Petroll, WM, Garana, R, Cavanagh, HD. (1994) Corneal keratocytes: in situ and in vitro organization of cytoskeletal contractile proteins Invest Ophthalmol Vis Sci 35,730-743 [PubMed]
Wilson, SE, Mohan, RR, Ambrosio, R, Jr, Hong, J, Lee, J. (2001) The corneal wound healing response: cytokine-mediated interaction of the epithelium, stroma, and inflammatory cells Prog Retinal Eye Res 20,625-637 [CrossRef]
Pircher, R, Lawrence, DA, Jullien, P. (1984) Latent beta-transforming growth factor in nontransformed and Kirsten sarcoma virus-transformed normal rat kidney cells, clone 49F Cancer Res 44,5538-5543 [PubMed]
Ekblom, P, Ekblom, M, Fecker, L, et al (1994) Role of mesenchymal nidogen for epithelial morphogenesis in vitro Development 120,2003-2014 [PubMed]
Fassler, R, Sasaki, T, Timpl, R, Chu, ML, Werner, S. (1996) Differential regulation of fibulin, tenascin-C, and nidogen expression during wound healing of normal and glucocorticoid-treated mice Exp Cell Res 222,111-116 [CrossRef] [PubMed]
Wilson, SE, Kim, WJ. (1998) Keratocyte apoptosis: implications on corneal wound healing, tissue organization, and disease Invest Ophthalmol Vis Sci 39,220-226 [PubMed]
Vaughan, MB, Howard, EW, Tomasek, JJ. (2000) Transforming growth factor-beta1 promotes the morphological and functional differentiation of the myofibroblast Exp Cell Res 257,180-189 [CrossRef] [PubMed]
Joyce, NC, Harris, DL, Mello, DM. (2002) Mechanisms of mitotic inhibition in corneal endothelium: contact inhibition and TGF-beta2 Invest Ophthalmol Vis Sci 43,2152-2159 [PubMed]
Kim, TY, Kim, WI, Smith, RE, Kay, ED. (2001) Role of p27(Kip1) in cAMP-and TGF-beta2-mediated antiproliferation in rabbit corneal endothelial cells Invest Ophthalmol Vis Sci 42,3142-3149 [PubMed]
Petroll, WM, Jester, JV, Bean, JJ, Cavanagh, HD. (1998) Myofibroblast transformation of cat corneal endothelium by transforming growth factor-beta1, -beta2, and -beta3 Invest Ophthalmol Vis Sci 39,2018-2032 [PubMed]
Schmid, P, Cox, D, Bilbe, G, Maier, R, McMaster, GK. (1991) Differential expression of TGF beta 1, beta 2 and beta 3 genes during mouse embryogenesis Development 111,117-130 [PubMed]
Pelton, RW, Saxena, B, Jones, M, Moses, HL, Gold, LI. (1991) Immunohistochemical localization of TGF beta 1, TGF beta 2, and TGF beta 3 in the mouse embryo: expression patterns suggest multiple roles during embryonic development J Cell Biol 115,1091-1105 [CrossRef] [PubMed]
Frank, S, Madlener, M, Werner, S. (1996) Transforming growth factors beta1, beta2, and beta3 and their receptors are differentially regulated during normal and impaired wound healing J Biol Chem 271,10188-10193 [CrossRef] [PubMed]
Levine, JH, Moses, HL, Gold, LI, Nanney, LB. (1993) Spatial and temporal patterns of immunoreactive transforming growth factor beta 1, beta 2, and beta 3 during excisional wound repair Am J Pathol 143,368-380 [PubMed]
Sanford, LP, Ormsby, I, Gittenberger-de Groot, AC, et al (1997) TGFbeta2 knockout mice have multiple developmental defects that are non-overlapping with other TGFbeta knockout phenotypes Development 124,2659-2670 [PubMed]
Saika, S, Liu, CY, Azhar, M, et al (2001) TGFbeta2 in Corneal Morphogenesis during mouse embryonic development Dev Biol 240,419-432 [CrossRef] [PubMed]
Kulkarni, AB, Huh, CG, Becker, D, et al (1993) Transforming growth factor beta 1 null mutation in mice causes excessive inflammatory response and early death Proc Natl Acad Sci USA 90,770-774 [CrossRef] [PubMed]
Dickson, MC, Martin, JS, Cousins, FM, Kulkarni, AB, Karlsson, S, Akhurst, RJ. (1995) Defective haematopoiesis and vasculogenesis in transforming growth factor-beta 1 knock out mice Development 121,1845-1854 [PubMed]
Nishida, K, Kinoshita, S, Yokoi, N, Kaneda, M, Hashimoto, K, Yamamoto, S. (1994) Immunohistochemical localization of transforming growth factor-beta 1, - beta 2, and -beta 3 latency-associated peptide in human cornea Invest Ophthalmol Vis Sci 35,3289-3294 [PubMed]
Qin, P, Piechocki, M, Lu, S, Kurpakus, MA. (1997) Localization of basement membrane-associated protein isoforms during development of the ocular surface of mouse eye Dev Dyn 209,367-376 [CrossRef] [PubMed]
Pallikaris, IG, Siganos, DS. (1994) Excimer laser in situ keratomileusis and photorefractive keratectomy for correction of high myopia J Refract Corneal Surg 10,498-510 [PubMed]
Chen, CC, Chang, JH, Lee, JB, Javier, J, Azar, DT. (2002) Human corneal epithelial cell viability and morphology after dilute alcohol exposure Invest Ophthalmol Vis Sci 43,2593-2602 [PubMed]
Shimazaki, J, Yang, H-Y, Tsubota, K. (1997) Amniotic membrane transplantation for ocular surface reconstruction in patients with chemical burns Ophthalmology 104,2068-2076 [CrossRef] [PubMed]
Strissel, KJ, Rinehart, WB, Fini, ME. (1997) Regulation of paracrine cytokine balance controlling collagenase synthesis by corneal cells Invest Ophthalmol Vis Sci 38,546-552 [PubMed]
Holgate, ST, Davies, DE, Lackie, PM, Wilson, SJ, Puddicombe, SM, Lordan, JL. (2000) Epithelial-mesenchymal interactions in the pathogenesis of asthma J Allergy Clin Immunol 105,193-204 [CrossRef] [PubMed]
Knight, D. (2001) Epithelium-fibroblast interactions in response to airway inflammation Immunol Cell Biol 79,160-164 [CrossRef] [PubMed]
Mio, T, Liu, XD, Adachi, Y, et al (1998) Human bronchial epithelial cells modulate collagen gel contraction by fibroblasts Am J Physiol 274,L119-L126 [PubMed]
Zhang, S, Smartt, H, Holgate, ST, Roche, WR. (1999) Growth factors secreted by bronchial epithelial cells control myofibroblast proliferation: an in vitro co-culture model of airway remodeling in asthma Lab Invest 79,395-405 [PubMed]
Sacco, O, Romberger, D, Rizzino, A, Beckmann, JD, Rennard, SI, Spurzem, JR. (1992) Spontaneous production of transforming growth factor-beta 2 by primary cultures of bronchial epithelial cells: effects on cell behavior in vitro J Clin Invest 90,1379-1385 [CrossRef] [PubMed]
Gizycki, MJ, Adelroth, E, Rogers, AV, O’Byrne, PM, Jeffery, PK. (1997) Myofibroblast involvement in the allergen-induced late response in mild atopic asthma Am J Respir Cell Mol Biol 16,664-673 [CrossRef] [PubMed]
Laitinen, LA, Laitinen, A, Altraja, A, et al (1996) Bronchial biopsy findings in intermittent or “early” asthma J Allergy Clin Immunol 98,S3-S6discussion S33–S40 [CrossRef] [PubMed]
Laitinen, A, Altraja, A, Kampe, M, Linden, M, Virtanen, I, Laitinen, LA. (1997) Tenascin is increased in airway basement membrane of asthmatics and decreased by an inhaled steroid Am J Respir Crit Care Med 156,951-958 [CrossRef] [PubMed]
Andriessen, MP, Niessen, FB, Van de Kerkhof, PC, Schalkwijk, J. (1998) Hypertrophic scarring is associated with epidermal abnormalities: an immunohistochemical study J Pathol 186,192-200 [CrossRef] [PubMed]
Machesney, M, Tidman, N, Waseem, A, Kirby, L, Leigh, I. (1998) Activated keratinocytes in the epidermis of hypertrophic scars Am J Pathol 152,1133-1141 [PubMed]
Whitby, DJ, Ferguson, MW. (1991) The extracellular matrix of lip wounds in fetal, neonatal and adult mice Development 112,651-668 [PubMed]
Hopkinson-Woolley, J, Hughes, D, Gordon, S, Martin, P. (1994) Macrophage recruitment during limb development and wound healing in the embryonic and foetal mouse J Cell Sci 107,1159-1167 [PubMed]
Figure 1.
 
Reduction of stromal exposure to epithelium after penetrating keratectomy decreases the appearance of fibrotic markers. The quality of repair after penetrating keratectomy was examined by immunofluorescent localization of α-sm actin (green), which is characteristic of fibrotic repair and by using rhodamine phalloidin (red) to stain filamentous actin stress fibers, which are characteristic of active fibroblasts. Monocyte cell populations were localized by immunostaining for Mac1 (red). Cell nuclei were stained with DAPI (blue) to delineate the location of cells. (A) Time course of marker localization in corneas repairing after penetrating keratectomy. Bracket: epithelial tissue layer. (B) Effect of epithelial removal on marker localization in corneas repairing after penetrating keratectomy. The epithelium of repairing corneas was removed daily beginning at day 3 after surgery. Mean fluorescence intensity for markers was quantified and compared with a control group that did not undergo epithelial removal at day 7 after penetrating keratectomy. Only the stromal tissue layer is visible in these images. Scale bar, 50 μm.
Figure 1.
 
Reduction of stromal exposure to epithelium after penetrating keratectomy decreases the appearance of fibrotic markers. The quality of repair after penetrating keratectomy was examined by immunofluorescent localization of α-sm actin (green), which is characteristic of fibrotic repair and by using rhodamine phalloidin (red) to stain filamentous actin stress fibers, which are characteristic of active fibroblasts. Monocyte cell populations were localized by immunostaining for Mac1 (red). Cell nuclei were stained with DAPI (blue) to delineate the location of cells. (A) Time course of marker localization in corneas repairing after penetrating keratectomy. Bracket: epithelial tissue layer. (B) Effect of epithelial removal on marker localization in corneas repairing after penetrating keratectomy. The epithelium of repairing corneas was removed daily beginning at day 3 after surgery. Mean fluorescence intensity for markers was quantified and compared with a control group that did not undergo epithelial removal at day 7 after penetrating keratectomy. Only the stromal tissue layer is visible in these images. Scale bar, 50 μm.
Figure 2.
 
Conditioned epithelial media were capable of inducing the fibrotic phenotype in cultured corneal stromal cells through TGF-β2. (A) Subcultured rabbit corneal stromal cells were serum starved for 48 hours, and then treated with either 2 ng/mL TGF-β2, conditioned media from corneal epithelial cells, or conditioned media containing neutralizing antibodies specific for TGF-β1 and TGF-β2 isoforms. Arrows: indicate cells with normal keratocyte phenotype; (✶): cells with fibrotic phenotype. (B) Stromal cell lysates from the experiment were assayed for α-sm actin content by Western analysis. (C) Conditioned media prepared from 1 × 106 epithelial cells or stromal cells were harvested, and proteins were concentrated and examined by Western analysis for TGF-β expression. Purified TGF-β1 and -β2 (10 ng) were run as a positive control.
Figure 2.
 
Conditioned epithelial media were capable of inducing the fibrotic phenotype in cultured corneal stromal cells through TGF-β2. (A) Subcultured rabbit corneal stromal cells were serum starved for 48 hours, and then treated with either 2 ng/mL TGF-β2, conditioned media from corneal epithelial cells, or conditioned media containing neutralizing antibodies specific for TGF-β1 and TGF-β2 isoforms. Arrows: indicate cells with normal keratocyte phenotype; (✶): cells with fibrotic phenotype. (B) Stromal cell lysates from the experiment were assayed for α-sm actin content by Western analysis. (C) Conditioned media prepared from 1 × 106 epithelial cells or stromal cells were harvested, and proteins were concentrated and examined by Western analysis for TGF-β expression. Purified TGF-β1 and -β2 (10 ng) were run as a positive control.
Figure 3.
 
TGF-β isoforms were differentially localized during fibrotic corneal repair. TGF-β2 (A) and -β1 (B) isoforms were localized in repairing mouse corneal wounds by substrate immunohistochemistry (brown). Sections were counterstained with hematoxylin. Arrows: wound margin; brackets: epithelium. Scale bar, 50 μm.
Figure 3.
 
TGF-β isoforms were differentially localized during fibrotic corneal repair. TGF-β2 (A) and -β1 (B) isoforms were localized in repairing mouse corneal wounds by substrate immunohistochemistry (brown). Sections were counterstained with hematoxylin. Arrows: wound margin; brackets: epithelium. Scale bar, 50 μm.
Figure 4.
 
Epithelial–stromal interaction was inhibited by basement membrane. (A) Epithelial cells were plated in duplicate on either plastic or synthetic basement membrane, and their ability to release TGF-β isoforms was examined by Western analysis. (B) TGF-β2 from conditioned media was compared with cellular levels by Western analysis. (C) Corneal organotypic cultures were created without an endothelial layer or with an endothelial layer to stimulate formation of an epithelial basement membrane. The presence of basement membrane was confirmed by indirect immunofluorescence localization of laminin. Stromal cell myofibroblast transformation was examined by immunofluorescence for α-sm actin, and TGF-β2 was localized by substrate immunohistochemistry. TGF-β2–stained sections were counterstained with hematoxylin. Brackets: epithelial layer. Scale bar, 50 μm.
Figure 4.
 
Epithelial–stromal interaction was inhibited by basement membrane. (A) Epithelial cells were plated in duplicate on either plastic or synthetic basement membrane, and their ability to release TGF-β isoforms was examined by Western analysis. (B) TGF-β2 from conditioned media was compared with cellular levels by Western analysis. (C) Corneal organotypic cultures were created without an endothelial layer or with an endothelial layer to stimulate formation of an epithelial basement membrane. The presence of basement membrane was confirmed by indirect immunofluorescence localization of laminin. Stromal cell myofibroblast transformation was examined by immunofluorescence for α-sm actin, and TGF-β2 was localized by substrate immunohistochemistry. TGF-β2–stained sections were counterstained with hematoxylin. Brackets: epithelial layer. Scale bar, 50 μm.
Figure 5.
 
Lack of basement membrane in vivo during corneal repair correlated with myofibroblast transformation. Basement membrane return after corneal wounding was examined by immunolocalization of entactin (red) and correlated with myofibroblast transformation, as examined by α-sm actin immunofluorescence (green). Nuclei were localized by staining with DAPI (blue). Arrows: basement membrane region. Scale bar, 50 μm.
Figure 5.
 
Lack of basement membrane in vivo during corneal repair correlated with myofibroblast transformation. Basement membrane return after corneal wounding was examined by immunolocalization of entactin (red) and correlated with myofibroblast transformation, as examined by α-sm actin immunofluorescence (green). Nuclei were localized by staining with DAPI (blue). Arrows: basement membrane region. Scale bar, 50 μm.
Figure 6.
 
Stromal cell activation in the absence of basement membrane destruction occurred without myofibroblast transformation. The central corneal epithelium was debrided and the time course of stromal activation was examined by staining for filamentous actin with rhodamine phalloidin. Myofibroblast induction was examined by immunofluorescence for α-sm actin. The iris, which contains smooth muscle cells, was used as an internal positive control for α-sm actin. TGF-β2 was localized by substrate immunohistochemistry. Nuclei were localized by staining with DAPI (blue) in phalloidin staining. TGF-β2 immunohistochemistry was counterstained with hematoxylin. Arrows: basement membrane region. Scale bar, 50 μm.
Figure 6.
 
Stromal cell activation in the absence of basement membrane destruction occurred without myofibroblast transformation. The central corneal epithelium was debrided and the time course of stromal activation was examined by staining for filamentous actin with rhodamine phalloidin. Myofibroblast induction was examined by immunofluorescence for α-sm actin. The iris, which contains smooth muscle cells, was used as an internal positive control for α-sm actin. TGF-β2 was localized by substrate immunohistochemistry. Nuclei were localized by staining with DAPI (blue) in phalloidin staining. TGF-β2 immunohistochemistry was counterstained with hematoxylin. Arrows: basement membrane region. Scale bar, 50 μm.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×