The Corneal Wound Healing Response:

The Corneal Wound Healing Response:

PII: S1350-9462(01)00008-8 The Corneal Wound Healing Response: Cytokine-mediated Interaction of the Epithelium, Stroma, and Inflammatory Cells Steven ...

383KB Sizes 1 Downloads 65 Views

PII: S1350-9462(01)00008-8

The Corneal Wound Healing Response: Cytokine-mediated Interaction of the Epithelium, Stroma, and Inflammatory Cells Steven E. Wilson*, Rahul R. Mohan, Rajiv R. Mohan, Renato Ambro´sio Jr, JongWook Hong, and JongSoo Lee The Department of Ophthalmology, University of Washington School of Medicine, Box-356485 Seattle, WA 98195-6485, USA; The Department of Ophthalmology, Korea University, Seoul, Korea and Department of Ophthalmology, Research Institute of Medical Science, College of Medicine, Pusan National University, Pusan, South Korea CONTENTS

1. 2. 3. 4. 5. 6. 7. 8. 9. 10.

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Evolutionary context . . . . . . . . . . . . . . . . . . . . . . . . Overview of the wound healing response . . . . . . . . . . . . . Master regulators of the wound healing response . . . . . . . . . The corneal wound healing cascade . . . . . . . . . . . . . . . . Keratocyte apoptosis and necrosis . . . . . . . . . . . . . . . . . Lacrimal-gland-derived cytokine mediators and epithelial healing Keratocyte proliferation and migration-myofibroblasts . . . . . . Inflammatory cell infiltration and function . . . . . . . . . . . . Return to normalcy . . . . . . . . . . . . . . . . . . . . . . . . . Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. . . . . . . . . . . . .

. 625 625 626 626 626 628 628 630 631 . 634 . 635 . 636 . 636

AbstractFThe corneal wound healing cascade is complex and involves stromal–epithelial and stromal–epithelial–immune interactions mediated by cytokines. Interleukin-1 appears to be a master modulator of many of the events involved in this cascade. Keratocyte apoptosis is the earliest stromal event noted following epithelial injury and remains a likely target for modulation of the overall wound healing response. Other processes such as epithelial mitosis and migration, stromal cell necrosis, keratocyte proliferation, myofibroblast generation, collagen deposition, and inflammatory cell infiltration contribute to the wound healing cascade and are also likely modulated by cytokines derived from corneal cells, the lacrimal gland, and possibly immune cells. Many questions remain regarding the origin and fate of different cell types that contribute to stromal wound healing. Over a period of months to years the cornea returns to a state similar to that found in the unwounded normal cornea. r 2001 Elsevier Science Ltd. All rights reserved

1. INTRODUCTION

*Corresponding author. Tel.: +1-206-543-7250; fax: +1206-543-4414; e-mail: [email protected]

The corneal wound healing response is an exceedingly complex cascade involving cytokinemediated interactions between the epithelial cells,

Progress in Retinal and Eye Research Vol. 20, No. 5, pp. 625 to 637, 2001 r 2001 Elsevier Science Ltd. All rights reserved Printed in Great Britain 1350-9462/01/$ - see front matter

626

S. E. Wilson et al.

keratocytes of the stroma, corneal nerves, lacrimal glands, tear film, and cells of the immune system. As complex as the response is, it is relatively simple in contrast to wound healing that occurs in skin and organs that contain blood vessels and other components that participate in the process. This response is very similar in different species, although there appear to be some quantitative and qualitative variations in specific processes that comprise the cascade. Within a species there is variability depending on the inciting injury. For example, incisional, lamellar, and surface scrape injuries are followed by typical wound healing responses that are similar in some respects, but different in others. This review article will provide an overview of the cellular interactions that contribute to the corneal wound healing response with emphasis on cytokine regulation of these interactions.

2. EVOLUTIONARY CONTEXT It is important to view the corneal wound healing response in the context of the types of injuries that were most likely to place selective pressure during evolution. Clearly vision was essential to the survival of most animals and responses would likely have evolved to maintain vision following injury. It seems probable that abrasions from branches, projectiles, and other sources would have been the most common injuries to the vertebrate cornea. The wound healing responses to these variable injuries must retain integrity of the eye, restore the protective epithelial surface, while at the same time maintaining sufficient corneal clarity to provide vision. Infection from ubiquitous viral pathogens such as herpes simplex virus, smallpox virus, adenovirus, and their ancestors may have posed significant threats to the vision and survival of the evolving animals. Thus, pathogens that had the potential to permanently blind would likely have placed great selective pressure on animal species. Systems designed to impede the spread of viral pathogens until the immune response could eradicate the invader could have provided an advantage (Wilson et al., 1997). The responses that occur in the

cornea appear to be well designed to accomplish these objectives.

3. OVERVIEW OF THE WOUND HEALING RESPONSE The epithelium, stroma, and nerves participate in homeostasis of the anterior cornea and ocular surface. The lacrimal glands and tear film also contribute to the maintenance of surface smoothness and integrity important to function of the eye. Following injury, these components participate in an orchestrated response that efficiently restores corneal structure and function in most situations. Many cytokines and receptors modulate the process. Activation of these systems also attracts immune cells that function to eliminate debris and microbes that may breach the injured surface and gain entry to the corneal stroma. Only by considering the individual contributions of each of these components and their interactions can one begin to truly appreciate the beauty and efficiency of the overall response.

4. MASTER REGULATORS OF THE WOUND HEALING RESPONSE The cascade of responses to injury to the cornea is initiated very rapidly regardless of the type of injury. For example, the keratocyte apoptosis response detected by electron microscopy is so rapid that if one euthenizes a mouse, enucleates the eye, and performs a single scrape across the epithelium prior to plunging the eye into fixative, the anterior keratocytes already show chromatin condensation and other morphologic changes consistent with apoptosis. Thus, the cornea is primed and ready to respond immediately to injury. This would make sense if one of the functions of this response was to retard dissemination of viral pathogens until other defense mechanisms can be rallied. What are the key modulators that regulate the early events in the wound healing cascade? After working on these problems for many years we have come to believe that there are a few key cytokine modulators that act as ‘‘master

The corneal wound healing response

regulators’’ of the response. Some of these master regulators and their receptors are likely to be constitutively produced and, therefore, constantly available to initiate the wound healing response. These initiators are likely to be sequestered until needed so that their mischief is not unleashed unduly, at least in the normal cornea. Thus, these master regulators serve as ever-vigilant guardians, silently waiting and on the ready to unleash the cascade at the first sign of trouble. Interleukin (IL)-1 seems to fit the requirements of a master regulator with regards to expression and function. First the expression of this cytokine– receptor system will be discussed and then the many functions that are regulated in the keratocytes by IL-1 will be reviewed. Both IL-1 alpha (Fig. 1) and IL-1 beta mRNAs and proteins are expressed constitutively in the corneal epithelium (and endothelium) (Wilson et al., 1994b; Weng et al., 1996). IL-1 receptor (binds both IL-1 alpha and IL-1 beta) is also constitutively produced in keratocytes and corneal fibroblasts (cultured keratocytes will be referred to as corneal fibroblasts if they are cultured in serum and cultured keratocytes if they are cultured without serum) (Bereau et al., 1993; Fabre et al., 1991; Girard et al., 1991; Wilson et al., 1994c; Beales et al., 1999; Jester et al., 1999b). Neither IL-1 alpha or IL-1 beta are detectable by

Fig. 1. Interleukin 1-alpha is constitutively produced in the corneal epithelium. This immunocytochemical localization was performed in normal human cornea. Note there is no detectable expression of IL-1 alpha in the keratocytes in the unwounded cornea. Reprinted with permission from Wilson et al., (1994), Exp. Eye Res. 59, 63–72, r 1994 by permission of the publisher, Academic Press.

627

immunocytochemistry in keratocytes in the unwounded cornea. However, Fini and coworkers have demonstrated that exposure to IL-1 can induce keratocytes to produce IL-1 via an autocrine loop (Strissel et al., 1997a,b; West-Mays et al., 1995). Thus, IL-1 protein is detectable in keratocytes or myofibroblasts in the wounded cornea. IL-1 (alpha or beta) does not appear to be released from the epithelium into the stroma in significant amounts in the normal unwounded cornea. Both forms of IL-1 lack signal sequences for transport from the cell and released via cell injury or death (Dinarello, 1994). IL-1 is released from apical epithelial cells undergoing programmed cell death as a part of the normal maturation and turnover of the epithelium and may be present in tears at increased levels in conditions associated with ocular surface injury such as keratoconjunctivitis sicca. (Pflugfelder, 1998; Pflugfelder et al., 1999). However, tear IL1 probably does not pass into the anterior stroma in the absence of epithelial injury or death because of the barrier provided by the intact epithelium. IL-1 is dumped onto the exposed stroma immediately following epithelial injury that is of sufficient magnitude to break down epithelial barrier function or directly damage basal epithelial cells. In some cases, epithelial debris could be tracked into the stroma, for example with a microkeratome cut into the cornea. Once IL-1 penetrates the stroma, it can bind IL-1 receptors on the keratocyte cells and modulate the functions of these cells. Thus, IL-1 is sequestered within the epithelium separated from the stromal cells in the normal unwounded cornea until epithelial injury triggers its release. IL-1 has an array of important effects on keratocyte cells related to wound healing. It has been shown to modulate apoptosis of keratocytes and corneal fibroblasts (Wilson et al., 1996), although the effect appears to be mediated indirectly via the Fas/Fas ligand system through autocrine suicide (Mohan et al., 1997). Since IL-1 alpha also triggers NF-kappa B activation (Mohan et al., 2000) it also has negative apoptotic effects on the keratocyte cells and myofibroblast cells that appear during the stromal wound healing response. Thus, the overall effect of IL-1 could be related to the specific milieu of the cell

628

S. E. Wilson et al.

and the overall input by a number of different cytokines. IL-1 has negative chemotactic effects on keratocytes and corneal fibroblasts and may have a role in maintaining corneal tissue organization (the morphologic separation of epithelium from stroma in the normal cornea) through this effect (Wilson et al., 1996; Kim et al., 1999a,b; Wilson and Hong, 2000). IL-1 is the primary regulator or hepatocyte growth factor (HGF) and keratinocyte growth factor (KGF) production by keratocytes (Weng et al., 1996). HGF and KGF are classical mediators of stromal-epithelial interaction that are produced by keratocytes and myofibroblasts to regulate proliferation, motility, and differentiation of epithelial cells (Wilson et al., 1994a). Thus, IL-1 released by injured corneal epithelial cells triggers production of HGF and KGF by the keratocytes to regulate the repair process of the corneal epithelial cells. IL-1 upregulates expression of collagenases, metalloproteinases, and other enzymes by keratocytes (Strissel et al., 1997a,b; West-Mays et al., 1995). These enzymes have an important role in remodeling of collagen during corneal wound healing. IL-1 and TNF alpha also upregulate the expression of some chemokines such as IL-8, RANTES, and monocyte chemotactic protein (MCP)-1 in keratocytes and corneal epithelial cells (Tran et al., 1996; Takano et al., 1999). IL-1 also potentiates the chemotactic effect of plateletderived growth factor (PDGF) on corneal fibroblasts. Thus, IL-1 directly regulates several critical processes that contribute to wound healing. This, along with the distribution of expression of IL-1 and IL-1 receptor in the cornea and the sequestration of IL-1 in the absence of injury, suggest a role for IL-1 as a master regulator of the corneal wound healing response. There may be other master regulator cytokines that function to initiate the early wound healing response. For example, PDGF is expressed by corneal epithelial cells and the keratocytes express the PDGF receptors (Denk and Knorr, 1997; Andresen and Ehlers, 1998; Kamiyama et al., 1998; Kim et al., 1999a,b). PDGF is found at very high levels in the epithelial basement membrane.

This localization and sequestration is likely related to the heparin-binding property of PDGF. PDGF is released into the stroma after damage to the epithelium and underlying basement membrane. PDGF modulates corneal fibroblast proliferation, chemotaxis, and possibly differentiation (Denk and Knorr, 1997; Andresen and Ehlers, 1998; Kamiyama et al., 1998; Kim et al., 1999a,b). Tumor necrosis factor (TNF) alpha could also participate (Mohan et al., 2000). Less is known about the function of TNF alpha and other cytokines that could also serve as master regulators in the corneal wound healing response.

5. THE CORNEAL WOUND HEALING CASCADE Studies performed over the past decade have revealed numerous processes that comprise the overall wound healing cascade in the cornea. It is helpful to outline the steps in this response prior to discussing individual components of the response. It must be appreciated that many of these events occur simultaneously and, therefore, the ‘‘cascade’’ should be viewed as such only in rough terms. It is clear that some events proceed others. For example, keratocyte apoptosis is the earliest stromal response that can be detected following epithelial injury and other components of the cascade appear to follow. Figure 2 provides the rough framework of the cascade. It is not intended to be comprehensive and some processes that are clearly important to the overall wound healing response are not depicted in the figure. Subsequent sections will concentrate on individual processes in the wound healing cascade with emphasis on cytokine mediation.

6. KERATOCYTE APOPTOSIS AND NECROSIS Work in our laboratory demonstrated that the disappearance of keratocytes that followed epithelial injury was mediated by apoptosis (Wilson et al., 1996). Studies have suggested that this regulated cell death is mediated by cytokines released from the injured epithelium such as IL-1

The corneal wound healing response

629

Fig. 3. Keratocyte apoptosis detected with the TUNEL assay at 4 h after epithelial scrape injury in the human cornea. The scrape injury was produced 4 h prior to enucleation for intraocular melanoma. Note staining of the superficial keratocytes (small arrows) beneath Bowman’s layer (large arrows). This experiment was approved by the Institutional Review Board at the University of Washington, Seattle, WA. Magnification 200  .

Fig. 2. Schematic diagram indicating some of the events that occur in the corneal wound healing response that occurs following corneal epithelial injury or surgical procedures such as PRK or LASIK. Note that this is a simplified scheme and not all events that may be important are included. While there is some indication of sequence (for example keratocyte apoptosis is the first observable event following injury) many of the events occur simultaneously in the cornea.

and TNF alpha (Wilson et al., 1996; Mohan et al., 1998, 2000). It is largely unknown how these different pro-apoptotic cytokine pathways interact to determine whether signaling dictates cell death or some other response in individual cells. Likely there is redundancy of the systems and possibly a requirement for context before a death signal is recognized and acted upon. In some cases, more than one of these systems may be involved. For

example, the effect of the IL-1 system on keratocyte apoptosis may be mediated indirectly through the Fas–Fas ligand system via autocrine suicide (Mohan et al., 1997). Many types of epithelial injury will induce keratocyte apoptosis. Some of the triggers include mechanical scrape (Wilson et al., 1996), corneal surgical procedures like photorefractive keratectomy (PRK) and laser in situ keratomeliusis (LASIK) (Helena et al., 1998), viral infection (Wilson et al., 1997), incisions (Helena et al., 1998), and even pressure applied with an instrument on the epithelial surface (Wilson, 1997). Recent experiments have confirmed that apoptosis occurs in the keratocytes underlying Bowman’s layer in the human eye when the epithelium has a scrape injury (Fig. 3, Ambrosio et al., unpublished data). It was previously noted that keratocyte apoptosis is the first observable stromal response following epithelial injury (Wilson et al., 1996). The earliest changes are noted at the electron microscopic level. It takes a few minutes longer (up to 30 min) before apoptosis can be detected with the TUNEL assay. Thus, DNA fragmentation detected by the TUNEL assay takes somewhat longer to develop and has been found to be most prominent at approximately 4 h after the scrape injury in mice and rabbits (Wilson et al., 1996, 1998).

630

S. E. Wilson et al.

Keratocyte apoptosis appears to continue for a period of time extending for at least 1 week following injuries such as epithelial scrape, epithelial scrape followed by PRK, or a microkeratome cut into the cornea (Gao et al., 1997; Mohan et al., 2001). Apoptosis detected with the TUNEL assay appears, however, to diminish with time past 72 h. Thus, while many keratocytes near the surface die immediately after epithelial injury by apoptosis, other cells derived from keratocytes that were stimulated to proliferate and migrate may continue to undergo apoptosis for days. Other cell types that arrive later could also undergo apoptosis. Thus, it is likely that the inflammatory cells that begin to arrive approximately 12–24 h after the injury eventually undergo apoptosis, but the time course over which this occurs is presently unknown. The disappearance of myofibroblasts over time following injury could also be mediated by apoptosis, rather than through a transition back to a keratocyte cell. There is currently no data regarding the mechanism of myofibroblast disappearance as the cornea returns to morphology and function more like that in the prewounding cornea. Virtually all of the anterior keratocytes that disappear in the early post-wounding period seem to undergo apoptosis at the electron microscopic level. As the wound healing process continues, however, there appear to be some cells recognizable as keratocytes that have hallmarks of necrosis rather than apoptosis (Mohan et al., 2001). It can be difficult to ascertain with certainty that these necrotic appearing cells noted between 12 h and 1 week are keratocytes rather than inflammatory cells. If some of these cells are keratocytes, then it is possible that this cell death is mediated via the inflammatory cells that arrive at the site of stromal wound healing during this same time period following the initial epithelial injury. More investigation needs to be performed to elucidate the complex processes occurring at these later points in the wound healing process. Studies have demonstrated that keratocytes in the unwounded cornea are connected by cellular processes called gap junctions to form a synsytium (Watsky, 1995; Spanakis et al., 1998). It may be that cytokines released from the injured epithelium only bind to receptors on the most superficial keratocytes

and that the signal to undergo apoptosis is directed to deeper cells via these intercellular communication channels. Once keratocyte cells begin to proliferate and migrate into the area of the wound healing response they do so as individual cells. At some point, however, the synsytial connections of the normal cornea appear to be restored. The location of the keratocyte apoptosis and necrosis response varies with the type of corneal epithelial injury. This in turn tends to influence the location and effect of the subsequent events in the cascade. Thus, injuries such as scrape of the epithelium, mechanical pressure on the epithelium, and viral infection of the epithelium trigger keratocyte apoptosis and necrosis in the superficial stroma (Fig. 4A). In contrast, a lamellar cut across the cornea produced by a microkeratome induces keratocyte apoptosis and necrosis at the site of epithelial injury and anterior and posterior to the lamellar interface (Fig. 4B). This localization is thought to be attributable to tracking of epithelial debris, including pro-apoptotic cytokines, into the interface by the microkeratome blade. Cytokines from the injured peripheral epithelium could also diffuse along the lamellar interface and into the central stroma. This can be of considerable importance because it influences the localization of other events such as proximity between myofibroblasts and wound healing fibroblasts that produce increased HGF. HGF has effects on corneal epithelial cells that tend to promote epithelial hyperplasia (Wilson et al., 1994a). Thus, superficial keratocyte apoptosis and necrosis such as that triggered by PRK could be more likely to result in epithelial hyperplasia than deeper keratocyte apoptosis and necrosis noted in LASIK. This could be of clinical significance and may, at least in part, explain differences between the two procedures when they are used to correct high myopia.

7. LACRIMAL-GLAND-DERIVED CYTOKINE MEDIATORS AND EPITHELIAL HEALING Expression of several growth factors such as HGF and epidermal growth factor (EGF) that modulate epithelial healing increase in the lacrimal gland shortly after epithelial injury (Steinemann

The corneal wound healing response

631

Fig. 4. Localization of keratocyte apoptosis depends on the type of injury. (A) A lamellar cut produced by a microkeratome in preparation for lamellar in situ keratomeliusis (LASIK) triggers keratocyte apoptosis at the site of blade perforation of the epithelium (not shown) and along the lamellar interface (arrows). (B). Epithelial scrape injury in photorefractive keratectomy (PRK) triggers superficial stromal keratocyte apoptosis (arrows) detected by the TUNEL assay in a rabbit cornea. Magnification 200  .

et al., 1990; Wilson et al., 1999) (Fig. 5). This correlates with an increase in the bioavailability of HGF in the tears (Tervo et al., 1997). This upregulation is likely mediated via a reflex arc involving the trigeminal sensory nerves of the cornea connecting via the brainstem to the facial nerve fibers that innervate the lacrimal gland. Keratocyte cells, that are a source of growth factors such as HGF, undergo immediate apoptosis in the anterior stroma. Thus, the lacrimal gland could serve as the primary source of HGF and other epithelium-modulating cytokines that regulate proliferation, migration, and differentiation during the early wound healing period until myofibroblasts or corneal fibroblasts repopulate the anterior stroma. Lacrimal-gland-derived cytokines may continue to modulate differentiation and other functions in superficial epithelial cells once epithelial integrity and barrier function is restored. Thus, the lacrimal gland appears to have an important role in the corneal wound healing response and likely continues to modulate functions of the superficial epithelium once homeostasis is restored. The accessory lacrimal glands may also contribute to this response, but this has not been examined.

There are also likely to be localized effects of a variety of modulators on corneal epithelial wound healing. These include autocrine growth factor and receptor effects such as those that might be regulated by the EGF-transforming growth factor (TGF) alpha-EGF receptor systems within the epithelium itself (Wilson et al., 1994b, Mohan and Wilson, 1999). In addition, there are clearly important effects of other biomolecular modulators such as integrins, fibronectin, and other matrix components. The functions of these systems have been discussed in a recent review (Alio´ et al., 2000). Recently identified discoidin domain receptors expressed in the corneal epithelium that are activated by collagen could also have a role in epithelial healing (Mohan et al., 2001). Variations in proliferative and migration-related activity occur across the surface of the epithelium with some cells migrating to close the defect and others proliferating to provide additional cells. Cytokines are likely involved in regulating these processes (Zieske, 2000; Zieske et al., 2000). 8. KERATOCYTE PROLIFERATION AND MIGRATION-MYOFIBROBLASTS Apoptosis and necrosis result in an area of the cornea being relatively devoid of keratocytes with

632

S. E. Wilson et al.

Fig. 5. Growth factor expression increases in the lacrimal gland in response to corneal epithelial scrape injury. HGF mRNA (A), KGF mRNA (B), and EGF mRNA (C) levels were monitored in rabbit lacrimal glands following corneal epithelial wounding using an RNAse protection assay. Levels of mRNA were monitored in lacrimal glands of unwounded control animals and in the lacrimal glands of animals at 1 h, 8 h, 1, 3, and 7 days after corneal epithelial scrape injury. Beta actin mRNA was also monitored as a control for RNA loading in each lane. Note that beta actin mRNA levels are similar in most samples within an experiment. Sizes of the protected RNA sequences based on the base pair (BP) lengths of the cDNA sequences used to generate probes are indicated. HGF mRNA, KGF mRNA, and EGF mRNA levels increase in the lacrimal gland after corneal epithelial wounding. The levels appear to peak at 3 days for each growth factor. Reprinted with permission from Invest. Ophthalmol. Vis. Sci. 1999, 40, 2185–2190. r 1999 Association for Research in Vision and Opthalmology.

the region within the cornea being related to the type of wound. Zieske and coworkers (2001) have demonstrated that approximately 12–24 h following the original injury remaining keratocytes begin

to proliferate (Fig. 6). For example, with a corneal epithelial scrape injury, the anterior stroma is relatively devoid of keratocytes and the remaining keratocytes in the posterior and peripheral cornea

The corneal wound healing response

633

Fig. 6. Keratocyte proliferation monitored by detection of the mitosis specific Ki-67 antigen in rabbit corneas using immunohistochemistry. Cells undergoing mitosis stain green for the mitosis-associated antigen Ki-67 (arrows). Propidium iodide stains other cell nuclei orange/red: (A) is the unwounded cornea. Note no keratocytes stain for mitosis. (B) shows the cornea 8 h after epithelial scrape injury. Note there are no cells detected in the anterior stroma due to keratocyte apoptosis. No mitosis is detected. (C) shows the cornea 24 h after scrape injury. The epithelium has healed and there are some posterior keratocytes now beginning to undergo mitosis (arrows). (D) shows the cornea at 48 h after epithelial scrape. Note the broad band of keratocytes in the posterior stroma undergoing mitosis. Keratocyte or myofibroblast cells have repopulated the anterior stroma. (E). At 5 days after epithelial scrape injury the anterior stroma is fully repopulated. A few cells in the anterior stroma continue to undergo mitosis (arrows). (F) At 10 days after wounding no keratocytes undergoing mitosis are detected. Magnification 200  . Figure kindly provided by SR Guimaraes, AEK Hutcheon, and JD Zieske from their studies presented at the Association for Research in Vision and Ophthalmology annual meeting in Ft. Lauderdale, FL in 1999.

begin proliferation and migration. The factors modulating the beginning and ending of proliferation are not understood. We have hypothesized that platelet-derived growth factor (PDGF) and PDGF receptors have a role in this process (Kim et al., 1999b). PDGF is sequestered via its heparinbinding properties in high levels in the basement membrane in the unwounded cornea. Once epithelial injury occurs there is likely release of PDGF into the stroma. Vesaluoma and coworkers (1997) also reported that PDGF is detected in the tears following corneal injury. Binding of PDGF to PDGF receptors on viable keratocytes likely stimulates proliferation and chemotaxis of the keratocytes, based on studies of PDGF in vitro (Denk and Knorr, 1997; Andresen and Ehlers, 1998; Kamiyama et al., 1998; Kim et al., 1999a,b). There may be other cytokines that contribute to

this regulation of keratocyte apoptosis that have yet to be identified. Keratocytes proliferation continues for several days (Zieske et al., 2001; Mohan et al., 2001). Recent data has demonstrated that keratocyte apoptosis and necrosis also continue for at least a week after the initial injury and, therefore, it may be that some of the new cells derived from keratocyte proliferation themselves undergo apoptosis or necrosis (Mohan et al., 2001). After a few days, apoptosis, necrosis, and mitosis wind down and a relatively quiescent state is restored. It is unknown how this overall process is regulated. One possible signal is healing and restoration of the normal differentiated state of the epithelium leading to restoration of the homeostatic levels of key cytokines including IL-1 and PDGF, but this has not been tested.

634

S. E. Wilson et al.

What cell types are derived from the mitotic keratocytes? In vitro studies have suggested that myofibroblasts are an important cell type generated during the first few days following injury (Masur et al., 1996; Jester et al., 1990, 1999b; Moller-Pederson et al., 1998). These studies suggest that transforming growth factor beta has an important role in the generation of the myofibroblasts. Myofibroblast cells are characterized by alpha smooth muscle actin expression. They also have altered transparency related to corneal cystallin production (Jester et al., 1999a). They may have additional differences relative to keratocytes that include increased production of growth factors such as HGF and KGF (Weng et al., 1996), collagen (Kaji et al., 1998; Jester, et al., 1999b; El-Shabrawi et al., 1998), glycosaminoglycans, collagenases, gelatinases, and metalloproteinases associated with remodeling of the collagen and the stroma (Girard et al., 1991; Strissel et al., 1997a,b; West-Mays et al., 1995; Ye and Azar, 1998; Ye et al., 2000). A question that remains unanswered is whether all of the cells derived from keratocytes proliferation following corneal injury are myofibroblasts. This could be the case, but it is also possible that all of the cells are originally typical keratocytes and some are modulated to transform into myofibroblasts once they migrate to the area involved in the healing process. The eventual fate of the myofibroblast cells is also uncertain. Do they trans-differentiate into keratocytes as the wound healing process is completed or are they terminally differentiated and eliminated by apoptosis to be replaced by keratocytes? These are questions that future research should address.

9. INFLAMMATORY CELL INFILTRATION AND FUNCTION Beginning approximately 12–24 h after injury to the cornea there is an influx of inflammatory cells into the corneal stroma (O’Brien et al., 1998). These inflammatory cells are thought to move into the stroma from the limbal blood vessels and possibly from the tear film. Although some of these cells can be detected with traditional

histologic analyses such as hematoxylin and eosin staining, specific identification of the cell type is typically ambiguous when these methods are applied. Surprising numbers of inflammatory cells can be detected when immunocytochemistry is performed with monoclonal antibodies to specific cell types. Thus, many invading macrophage/ monocyte cells can be detected in a rabbit cornea at 24 h after epithelial scrape injury when an antiCD 11b antibody is utilized (Mohan et al., 2001). Other types of inflammatory cells such as T cells and polymorphonuclear cells can also be detected using electron microscopy and immunocytochemistry (Mohan et al., 2001). What are the signals that draw these inflammatory cells into the corneal stroma following the corneal injury? Most likely there are soluble cytokines or chemokines released following injury that attract the inflammatory cells to the site of injury. Tran and coworkers (1996) demonstrated that corneal fibroblasts produce monocyte chemotactic protein-1 (MCP-1, also called monocyte chemotactic and activating factor or MCAF) in vitro when stimulated by IL-1 or TNF alpha. Recent studies in our laboratory have confirmed that corneal fibroblasts stimulated with IL-1 or TNF alpha upregulate expression of MCP-1/ MCAF and that this protein is only expressed at detectable levels in keratocytes in vivo after epithelial scrape injury or some other injury to the overlying corneal epithelium. These studies also noted that genes coding for other cytokines and chemokines that are chemotactic and activating for inflammatory cells are upregulated in corneal stromal fibroblasts in response to IL-1 or TNF alpha. These include granulocyte colonystimulating factor (G-CSF), neutrophil-activating peptide ENA-78, and monocyte-derived neutrophil chemotactic factor MDNCF. These studies suggest that cytokines such as IL-1 or TNF alpha released from the injured corneal epithelium bind to IL-1 and TNF receptors, respectively, on keratocytes and stimulate production of cytokines such as G-CSF and MCAF that are chemotactic to inflammatory cells. Further work is needed to explore these epithelial–stromal–immune interactions. What is the function of inflammatory cells attracted to the site of wound healing in the

The corneal wound healing response

cornea? It seems likely that one important function is the scavenging of cellular components released from keratocytes that undergo programmed cell death. Thus, the apoptotic bodies that distribute throughout the stroma are engulfed by some of these inflammatory cells. If the corneal injury is associated with tissue invasion by microorganisms then these cells function to eliminate these pathogens. Recently, we have begun to explore another possible function served by the macrophage/ monocyte type cells. At present it is assumed that all newly generated myofibroblast, fibroblast, and keratocyte cells in the healing cornea are derived from the proliferation of residual keratocytes. However, we have noted that a small proportion of the fibroblastic appearing cells that repopulate the anterior stroma after epithelial injury stain with the anti-CD 11b antibody. These cells are not detected in the normal unwounded cornea. In bone, the osteoclasts were originally assumed to be fibroblastic cells. It is now clear that osteoclasts are derived from monocyte/macrophage cells through a cytokine-mediated process that involves direct cell interaction between the osteoblasts and macrophage/monocyte cells (Teitelbaum, 2000). This differentiation process involving the macrophage/monocyte cell is mediated by M-CSF, RANKL, and OPG. These cytokine–receptor systems are present in the cells detected in the corneal stroma following corneal epithelial injury (Mohan and Wilson, 2001). Further investigation is needed to conclusively demonstrate that monocyte/macrophage cell types give rise to fibroblast cells in the corneal stroma. If confirmed, these studies would suggest that there is heterogeneous population of myofibroblast/fibroblast cells in the healing corneal stroma.

10. RETURN TO NORMALCY During the months to years following corneal injury and wound healing there is a return to normalcy with elimination of inflammatory and myofibroblast/fibroblast cells and restoration of the quiescent state of the keratocytes. This process is associated with remodeling of any disordered

635

collagen that may be layed down during the wound healing process. Cintron and workers (Lee et al., 1982; Cintron et al., 1990) have demonstrated that this stromal remodeling process can continue for years and result in at least partial clearing of even the most severe stromal scar. The process through which the inflammatory and myofibroblast/fibroblast cells are eliminated from the cornea are poorly characterized. It seems likely that the majority of inflammatory cells eventually undergo apoptosis since this is the process through which these cells are eliminated in many other organs. Thus, some of the cells that are noted to be undergoing apoptosis using the TUNEL assay at 24 h and later following corneal injury are likely to be inflammatory cells. At the electron microscopic level, some of the cells that die in the corneal stroma at 24 h to 1 week after injury also appear to have morphology that resembles necrosis more than apoptosis (Mohan et al., 2001). This later necrosis could be mediated by the inflammatory cells that are initially detected in the stroma between 12 and 24 h after injury. Myofibroblast cells may undergo apoptosis over the weeks to months following the initial injury. Thus, these cells are not detectable via alpha smooth muscle actin immunocytochemical staining several months after the original injury. This would be the case if the myofibroblast/fibroblast cells are terminally differentiated. Recent studies have suggested that apoptosis of some stromal cells can detected at a very low rate as long as 3 months after the original injury (Mohan and Wilson, 2001). The rate of this late apoptosis may be greater than that detected in the unwounded corneal stroma, but it is difficult to quantitate. It remains a possibility that at least some of the myofibroblasts differentiate to quiescent keratocytes. This remains an uncertainty. The corneal epithelium may undergo epithelial hyperplasia following corneal injury (Gauthier et al., 1997; Lohmann and Guell, 1998; Lohmann et al., 1999; Kim et al., 1999a,b). This is one of the mechanisms of regression of the refractive effect of PRK or LASIK surgery. There is often a return to a normal epithelial thickness over a period of months to years. The regulatory mechanisms that regulate this return to normal corneal epithelial morphology have not been characterized.

636

S. E. Wilson et al.

AcknowledgementsFSupported by EY10056 and EYO1730 from the National Eye Institute and an unrestricted grant from Research to Prevent Blindness, New York, NY.

REFERENCES Alio´, J. L., Pe´rez-Santonja, J. J., Tervo, T., Tabbara, K. F., Vesaluoma, M., Smith, R. J., Maddox, B. and Maloney, R. K. (2000) Postoperative inflammation, microbial complications, and wound healing following laser in situ keratomileusis. J. Refract. Surg. 16, 523–538. Andresen, J. L. and Ehlers, N. (1998) Chemotaxis of human keratocytes is increased by platelet-derived growth factor-BB, epidermal growth factor, transforming growth factor-alpha, acidic fibroblast growth factor, insulin-like growth factor-I, and transforming growth factor-beta. Curr. Eye Res. 17, 79–87. Beales, M. P., Funderburgh, J. L., Jester, J. V. and Hassell, J. R. (1999) Proteoglycan synthesis by bovine keratocytes and corneal fibroblasts: maintenance of the keratocyte phenotype in culture. Invest Ophthalmol. Vis. Sci. 40, 1658–1663. Bereau, J., Fabre, E. J., Hecquet, C., Pouliquen, Y. and Lorans, G. (1993) Modification of prostiglandin E2 and collagen synthesis in keratoconus fibroblasts associated with an increase of interleukin-1 alpha receptor number. C. R. Acad. Sci. III. 316, 425–430. Cintron, C., Covington, H. I. and Kublin, C. L. (1990) Morphologic analyses of proteoglycans in rabbit corneal scars. Invest. Ophthalmol. Vis. Sci. 31, 1789–1798. Denk, P. O. and Knorr, M. (1997) The in vitro effect of plateletderived growth factor isoforms on the proliferation of bovine corneal stromal fibroblasts depends on cell density. Graefes Arch. Clin. Exp. Ophthalmol. 235, 530– 534. Dinarello, C. A. (1994) The interleukin-1 family: 10 years of discovery. FASEB J. 8, 1314–1325. El-Shabrawi, Y., Kublin, C. L. and Cintron, C. (1998) mRNA levels of alpha 1 (VI) collagen, alpha 1 (XII) collagen, and beta ig in rabbit cornea during normal development and healing. Invest. Ophthalmol. Vis. Sci. 39, 36–44. Fabre, E. J., Bureau, J., Pouliquen, Y. and Lorans, G. (1991) Binding sites for human interleukin-1 alpha, gamma interferon and tumor necrosis factor on cultured fibroblasts of normal cornea and keratoconus. Curr. Eye Res. 10, 585–592. Gao, J., Gelber-Schwalb, T. A., Addeo, J. V. and Stern, M. E. (1997) Apoptosis in the rabbit cornea after photorefractive keratectomy. Cornea 16, 200–208. Gauthier, C. A., Holden, B. A., Epstein, D., Tengroth, B., Fagerholm, P. and Hamberg-Nystrom, H. (1997) Factors affecting epithelial hyperplasia after photorefractive keratectomy. J. Cataract. Refract. Surg. 23, 1042–1050. Girard, M. T., Matsubara, M. and Fini, M. E. (1991) Transforming growth factor-beta and interleukin-1 modulate metalloproteinase expression by corneal stromal cells. Invest. Ophthalmol. Vis. Sci. 32, 2441–2454. Helena, M. C., Baerveldt, F., Kim, W-J. and Wilson, S. E. (1998) Keratocyte apoptosis after corneal surgery. Invest. Ophthalmol. Vis. Sci. 39, 276–283. Jester, J. V., Huang, J., Barry-Lane, P. A., Kao, W. W., Petroll, W. M. and Cavanagh, H. D. (1990) Transforming growth

factor(beta)-mediated corneal myofibroblast differentiation requires actin and fibronectin assembly. Invest. Ophthalmol. Vis. Sci. 40, 1959–1967. Jester, J. V., Moller-Pedersen, T., Huang, J., Sax, C. M., Kays, W. T., Cavangh, H. D., Petroll, W. M. and Piatigorsky, J. (1999a) The cellular basis of corneal transparency: evidence for ‘corneal crystallins’. J. Cell. Sci. 112, 613– 622. Jester, J. V., Petroll, W. M. and Cavanagh, H. D. (1999b) Corneal stromal wound healing in refractive surgery: the role of myofibroblasts. Prog. Retin. Eye. Res. 18, 311– 356. Kaji, Y., Obata, H., Usui, T., Soya, K., Machinami, R., Tsuru, T. and Yamashita, H. (1998) Three-dimensional organization of collagen fibrils during corneal stromal wound healing after excimer laser keratectomy. J. Cataract. Refract. Surg. 24, 1441–1446. Kamiyama, K., Iguchi, I., Wang, X. and Imanishi, J. (1998) Effects of PDGF on the migration of rabbit corneal fibroblasts and epithelial cells. Cornea 17, 315–325. Kim, W-J., Helena, M. C., Mohan, R. R. and Wilson, S. E. (1999a) Changes in corneal morphology associated with chronic epithelial injury. Invest. Ophthalmol. Vis. Sci. 40, 35–42. Kim, W-J., Mohan, R. R., Mohan, R. R. and Wilson, S. E. (1999b) Effect of PDGF, IL-1 alpha, and BMP 2/4 on corneal fibroblast chemotaxis: expression of the plateletderived growth factor system in the cornea. Invest. Ophthalmol. Vis. Sci. 40, 1364–1372. Lee, R. E., Davison, P. F. and Cintron, C. (1982) The healing of linear nonperforating wounds in rabbit corneas of different ages. Invest. Ophthalmol. Vis. Sci. 23, 660–665. Lohmann, C. P. and Guell, J. L. (1998) Regression after LASIK for the treatment of myopia: the role of the corneal epithelium. Semin. Ophthalmol. 13, 79–82. Lohmann, C. P., Reischl, U. and Marshall, J. (1999) Regression and epithelial hyperplasia after myopic photorefractive keratectomy in a human cornea. J. Cataract. Refract. Surg. 25, 712–715. Masur, S. K., Dewal, H. S., Dinh, T. T., Erenburg, I. and Petridou, S. (1996) Myofibroblasts differentiate from fibroblasts when plated at low density. Proc. Natl. Acad. Sci. 93, 4219–4223. Mohan, R. R., Kim, W-J., Mohan, R. R., Chen, L. and Wilson, S. E. (1998) Bone morphogenic proteins 2 and 4 and their receptors in the adult human cornea. Invest. Ophthalmol. Vis. Sci. 39, 2626–2636. Mohan, R. R., Liang, Q., Kim, W-J., Helena, M. C., Baerveldt, F. and Wilson, S. E. (1997) Apoptosis in the cornea: further characterization of Fas/Fas ligand system. Exp. Eye Res. 65, 575–589. Mohan, R. R., Mohan, R. R., Kim, W. J. and Wilson, S. E. (2000) Modulation of TNF-alpha-induced apoptosis in corneal fibroblasts by transcription factor NF-kb. Invest. Ophthalmol. Vis. Sci. 41, 1327–1336. Mohan, R. R., Mohan, R. R. and Wilson, S. E. (2001). Discoidin domain receptor (DDR) 1 and 2: collagenactivated tyrosine kinase receptors in the cornea. Exp. Eye Res. 72, 87–92. Mohan, R. R. and Wilson, S. E. (1999) Ex vivo human corneal epithelial cells express membrane-bound precursor and mature soluble epidermal growth factor (EGF) and transforming growth factor (TGF) alpha proteins. Exp. Eye. Res. 68, 129–131.

The corneal wound healing response Moller-Pederson, T., Cavanagh, H. D., Petroll, W. M. and Jester, J. V. (1998) Neutralizing antibody to TGFbeta modulates stromal fibrosis but not regression of photoablative effect following PRK. Curr. Eye Res. 17, 736–747. O’Brien, T. P., Li, Q., Ashraf, M. F., Matteson, D. M., Stark, W. J. and Chan, C. C. (1998) Inflammatory response in the early stages of wound healing after excimer laser keratectomy. Arch. Ophthalmol. 116, 1470–1474. Pflugfelder, S. C. (1998) Tear fluid influence on the ocular surface. Adv. Exp. Med. Biol. 438, 611–617. Pflugfelder, S. C., Jones, D., Ji, Z., Afonso, A. and Monroy, D. (1999) Altered cytokine balance in the tear fluid and conjunctiva of patients with Sjogren’s syndrome keratoconjunctivitis sicca. Curr. Eye Res. 19, 201–211. Spanakis, S. G., Petridou, S. and Masur, S. K. (1998) Functional gap junctions in corneal fibroblasts and myofibroblasts. Invest. Ophthalmol. Vis. Sci. 39, 1320–1328. Steinemann, T. L., Thompson, H. W., Maroney, K. M., Palmer, C. H., Henderson, L. A., Malter, J. S., Clarke, D., Bromberg, B., Kunkle, M. and Beuerman, R. W. (1990) Changes in epithelial epidermal growth factor receptor and lacrimal gland EGF concentration after corneal wounding. Invest. Ophthalmol. Vis. Sci. 31, S55. Strissel, K. J., Girard, M. T., West-Mays, J. A., Rinehart, W. B., Cook, J. R., Brinckerhoff, C. E. and Fini, M. E. (1997a) Role of serum amyloid A as an intermediate in the IL-1 and PMA-stimulated signaling pathways regulating expression of rabbit fibroblast collagenase. Exp. Cell. Res. 15(237), 275–287. Strissel, K. J., Rinehart, W. B. and Fini, M. E. (1997b) Regulation of paracrine cytokine balance controlling collagenase synthesis by corneal cells. Invest. Ophthalmol. Vis. Sci. 38, 546–552. Takano, Y., Fukagawa, K., Shimmura, S., Tsubota, K., Oguchi, Y. and Saito, H. (1999) IL-4 regulates chemokine production induced by TNF-alpha in keratocytes and corneal epithelial cells. Br. J. Ophthalmol. 83, 1074–1076. Teitelbaum, S. L. (2000) Bone resorption by osteoclasts. Science 289, 1504–1508. Tervo, T., Vesaluoma, M., Bennett, G. L., Schwall, R., Helena, M., Liang, Q. and Wilson, S. E. (1997) Tear hepatocyte growth factor (HGF) availability increases markedly after excimer laser surface ablation. Exp. Eye Res. 64, 501–504. Tran, M. T., Tellaetxe-Isusi, M., Elner, V., Strieter, R. M., Lausch, R. N. and Oakes, J. E. (1996) Proinflammatory cytokines induce RANTES and MCP-1 synthesis in human corneal keratocytes but not in corneal epithelial cells. Beta-chemokine synthesis in corneal cells. Invest. Ophthalmol. Vis. Sci. 37, 987–996. Vesaluoma, M., Teppo, A. M., Gronhagen-Riska, C. and Tervo, T. (1997) Platelet-derived growth factor-BB (PDGF-BB) in tear fluid: a potential modulator of corneal wound healing following photorefractive keratectomy. Curr. Eye Res. 16, 825–831. Watsky, M. A. (1995) Keratocyte gap junctional communication in normal and wounded rabbit corneas and human corneas. Invest. Ophthalmol. Vis. Sci. 36, 2568–2576.

637

Weng, J., Mohan, R. R., Li, Q. and Wilson, S. E. (1996) IL-1 upregulates keratinocyte growth factor and hepatocyte growth factor mRNA and protein production by cultured stromal fibroblast cells: Interleukin-1 beta expression in the cornea. Cornea. 16, 465–471. West-Mays, J. A., Strissel, K. J., Sadow, P. M. and Fini, M. E. (1995) Competence for collagenase gene expression by tissue fibroblasts requires activation of an interleukin 1 alpha autocrine loop. Proc. Natl. Acad. Sci. (U S A) 92, 6768–6772. Wilson, S. E. (1998) Keratocyte apoptosis in refractive surgery: Everett Kinsey Lecture. CLAO J. 24, 181–185. Wilson, S. E., He, Y-G., Weng, J., Li, Q., Vital, M. and Chwang, E. L. (1996) Epithelial injury induces keratocyte apoptosis: hypothesized role for the interleukin-1 system in the modulation of corneal tissue organization. Exp. Eye Res. 62, 325–338. Wilson, S. E., He, Y-G., Weng, J., Zieske, J. D., Jester, J. V. and Schultz, G. S. (1994a) Effect of epidermal growth factor, hepatocyte growth factor, and keratinocyte growth factor, on proliferation, motility, and differentiation of human corneal epithelial cells. Exp. Eye Res. 59, 665–678. Wilson, S. E. and Hong, J. W. (2000) Bowman’s layer structure and function: critical or dispensable to corneal function? A hypothesis. Cornea 19, 417–420. Wilson, S. E., Liang, Q. and Kim, W-J. (1999) Lacrimal gland HGF, KGF, and EGF mRNA levels increase after corneal epithelial wounding. Invest. Ophthalmol. Vis. Sci. 40, 2185–2190. Wilson, S. E., Lloyd, S. A. and He, Y-G. (1994c) Glucocorticoid receptor and interleukin-1 receptor messenger RNA expression in corneal cells. Cornea 13, 4–8. Wilson, S. E., Pedroza, L., Beuerman, R. and Hill, J. M. (1997) Herpes simplex virus type-1 infection of corneal epithelial cells induces apoptosis of the underlying keratocytes. Exp. Eye Res. 64, 775–779. Wilson, S. E., Schultz, G. S., Chegini, N., Weng, J. and He, YG. (1994b) Epidermal growth factor, transforming growth factor alpha, transforming growth factor beta, acidic fibroblast growth factor, basic fibroblast growth factor, and interleukin-1 proteins in the cornea. Exp. Eye Res. 59, 63–72. Ye, H. Q. and Azar, D. T. (1998) Expression of gelatinases A and B, and TIMPs 1 and 2 during corneal wound healing. Invest. Ophthalmol. Vis. Sci. 39, 913–921. Ye, H. Q., Maeda, M., Yu, F. S. and Azar, D. T. (2000) Differential expression of MT1-MMP (MMP-14) and collagenase III (MMP-13) genes in normal and wounded rat corneas. Invest. Ophthalmol. Vis. Sci. 41, 2894–2899. Zieske, J. D. (2000) Expression of cyclin-dependent kinase inhibitors during corneal wound repair. Prog. Retin. Eye Res. 19, 257–270. Zieske, J. D., Guimaraes, S. R. and Hutcheon, A. E. K. (2001) Kinetics of keratocyte proliferation in response to epithelial debridement. Exp. Eye Res. 72, 33–40. Zieske, J. D., Takahashi, H., Hutcheon, A. E. and Dalbone, A. C. (2000) Activation of epidermal growth factor receptor during corneal epithelial migration. Invest. Ophthalmol. Vis. Sci. 41, 1346–1355.