Corneal Wound Healing in an Osteopontin-Deficient Mouse

Abstract

purpose. To investigate the effects of loss of osteopontin (OPN) in the healing of the injured cornea in mice. Cell culture study was also conducted to clarify the effects of OPN in fibroblast behaviors.

methods. Ocular fibroblasts from wild-type (WT) and OPN-null (KO) mice were used to study the role of OPN on cell behavior. The effect of the lack of OPN on corneal wound healing was evaluated in mice.

results. In cell culture, OPN is involved in cell adhesion and in the migration of ocular fibroblasts. Adhesion of the corneal epithelial cell line was not affected by exogenous OPN. OPN was upregulated in a healing, injured mouse cornea. Loss of OPN did not affect epithelial healing after simple epithelial debridement. Healing of an incision injury in cornea was delayed, with less appearance of myofibroblasts and transforming growth factor β1 expression in a KO mouse than in a WT mouse. The absence of OPN promoted tissue destruction after an alkali burn, resulting in a higher incidence of corneal perforation in KO mice than in WT mice.

conclusions. OPN modulates wound healing-related fibroblast behavior and is required to restore the physiological structure of the cornea after wound healing.

The cornea is an avascular tissue of the eye and must remain transparent to refract light properly. An organized extracellular matrix structure is essential to the maintenance of transparency. Alkali burn to the cornea is a serious problem that may cause severe and permanent visual impairment by tissue scarring. ^{1 2} Activation of keratocytes (corneal fibroblasts) results in the generation of myofibroblasts. ^{3 4 5} During healing, myofibroblasts express and polymerize incorrect extracellular matrix components to remodel the damaged stromal connective tissue; unfortunately, excessive tissue fibrosis is also induced in certain conditions. Cell behavior during the process of wound healing in an injured ocular tissue is regulated in a complex way by various growth factors that play critical roles in profibrogenic and pro-inflammatory reactions. ^{6 7}  

Osteopontin (OPN) is a matrix and structural glycophosphoprotein that is also abundantly expressed in tissues during inflammation and repair. It functions as a cytokine that regulates the activities of macrophages, other immune cells, and resident tissue cells (epithelial cell types and mesenchymal cells) at sites of injury. ^{8 9 10 11 12 13 14 15 16 17 18 19} OPN modulates fibroblast proliferation, migration, and matrix remodeling in vitro, ^{8 10 13 17} all of which are critical components of wound healing processes in vivo. OPN function in tissue fibrosis has been investigated. ^{20 21 22 23 24 25 26 27} It has been reported that the loss of OPN results in dilated airspace rather than in alveolar obstruction (fibrosis), with reduction of type I collagen deposition after intratracheal administration of bleomycin sulfate. ^{22 25 27} Other examples of impaired wound healing in the absence of OPN were observed in heart, skin, or other tissues in mice. ^{20 21 22 23 24 25 26 27} These phenomena might be caused, at least in part, by defective matrix deposition/reconstruction. OPN is involved in cancer progression, ^{28 29 30 31} which also suggests its role in cell behavior modulation. As for the eye, we reported that OPN is expressed in injured lens epithelial cells in association with fibrotic scar formation. ³¹ Our preliminary experiment showed that OPN expression is upregulated in a healing, injured cornea in mice. However, the effects of this molecule on the regulation of inflammation and tissue fibrosis in the healing process in ocular surface tissues are still unknown. To address this, we took advantage of the availability of OPN-null (KO) mice. In the present study, cell culture studies first showed that OPN is involved in cell adhesion and migration. ^{32 33} We then examined the effect of absent OPN on the healing process in corneal stroma after an incision injury or an alkali exposure. The present results demonstrated that OPN newly expressed is required for normal healing of the corneal stroma. 

Materials and Methods

In vivo experiments were approved by the DNA Recombination Experiment Committee and the Animal Care and Use Committee of Wakayama Medical University and were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

In Vitro Experiments

Adhesion Assay of Cells to OPN-Coated Glass Slides.

Glass chamber slides (Laboratory-Tec; Nalge Nunc International, Rochester, NY) were coated with human recombinant OPN (0.25 μg/mL, catalog no. CC1074; Chemicon, Temecula, CA) in phosphate-buffered saline (PBS) for 24 hours at 4°C. Fibroblasts were obtained from sclera and cornea of postnatal day (P) 1 mice, as previously reported. ³³ In brief, the eyeball shells (including cornea and sclera) of P1 wild-type (WT) mice were carefully separated from other intraocular tissues (lens, iris, vitreous, retina) and were minced and explanted in collagen-coated 60-mm culture dishes (Corning-Iwaki Glass, Tokyo, Japan) for the outgrowth of ocular fibroblasts in Eagle minimum essential medium supplemented with 10% fetal calf serum, antibiotics (100 U/mL penicillin, 1000 μg/mL streptomycin), and an antimycotic (2.50 μg/mL amphotericin B). The cells were negative for keratocan, as revealed by Western blotting and as expected (data not shown). ^{34 35} Ocular fibroblasts (5.1 × 10⁴/μL/well) or an SV40-immortalized cultured human corneal epithelial cell line ³⁶ (2.0 × 10⁴/μL/well), kindly provided by Kaoru Araki-Sasaki, were seeded in the wells of chamber slides. After 20-hour incubation, the slides were fixed with formalin and the cells were stained with hematoxylin and eosin (HE). The number of cells that had adhered to and spread on the chamber slides was counted. Three wells were prepared, and the mean value of the cells from three independent areas was determined. 

Role of Endogenous OPN in Cell Adhesion.

KO mice of C57BL/6 background ³⁷ were used, and WT littermates were used as control WT mice. Ocular fibroblasts, obtained from WT or KO mice as described, were used. Cells (1.1 × 10⁴/μL/well) were seeded onto glass chamber slides and allowed to adhere for 18 and 30 hours. The cells were then fixed with formalin and stained with HE. 

To further examine the role of OPN on cell adhesion, WT cells (2.3 × 10⁴/mL/well) or corneal epithelial cell lines (1.4 × 10⁴/μL/well) were seeded onto glass chamber slides in the presence or absence of rabbit polyclonal anti–OPN neutralizing antibody ³⁷ (40 μg/mL) in 5% fetal calf serum-plus medium. After 13-hour incubation, the slides were fixed with formalin and the cells were stained with HE. The number of cells that had adhered to and spread onto the chamber slides was counted. Three wells were prepared, and the mean value of the cells from three independent areas was determined. 

Cell Migration as Examined by Using Scratch Assay.

Cell migration was assayed by evaluating closure of a linear wound produced by scratching the cell monolayer on a glass chamber slide in 1% fetal calf serum plus medium, as previously reported. ³⁸ WT and KO cells were used. KO cells were cultured on noncoated chamber slides or on slides coated with human recombinant OPN (0.25 μg/mL; Chemicon), as described. 

In Vivo Wound Healing Experiments

Healing of Epithelial Debridement in Mouse Corneas.

Four-week-old KO mice or WT mice (n = 4 each) were given general and topical anesthesia, as previously reported. ^{39 40} A circular defect (1.6 mm in diameter) was produced in the central corneal epithelium by using an ear punch for a mouse (1.6 mm in diameter) and a dull microsurgery blade, as previously reported. ⁴⁰ The cornea was photographed with fluorescein staining, and the mice were killed 20 hours after debridement. ⁴⁰ Enucleated eyes were fixed in 4% paraformaldehyde, embedded in paraffin, and stained with HE, as previously reported. ^{39 40}  

Incision Injury in the Central Cornea.

WT and KO mice were given general and topical anesthesia, as previously reported. ^{6 7} A full-thickness penetrating incision injury (1.5 mm in length) was produced in the central cornea with a surgical blade (Satin Crescent; Alcon, Fort Worth, TX). Ofloxacin ointment was topically administered immediately after injury, and the affected eye was allowed to heal at days 1, 3, 7, 14, and 21. The eye was enucleated and fixed in 4% paraformaldehyde for 48 hours. Four WT and 4 KO eyes were prepared for histology in each experimental condition at 3, 7, and 21 days. Fixed specimens were routinely embedded in paraffin. 

Alkali Burn in Mouse Ocular Surface.

Mice were given general and topical anesthesia, as previously reported. ^{6 7} Three microliters of 1 N sodium hydroxide solution was applied to each right eye of 4-week-old KO mice or WT mice to produce an ocular surface alkali burn. ^{6 7} Ofloxacin ointment was topically administered twice a week in the first 2 weeks and then once a week until week 8 to reduce the risk for bacterial contamination. Eyes with obvious bacterial infection were excluded from the study. Eyes of KO and WT mice underwent histologic examination at days 5, 10, and 20 after alkali burn (n = 6 for paraffin sections; n = 3 for cryosections for each experimental condition). The eye was enucleated and fixed in 4% paraformaldehyde for 48 hours. Unfixed eyes were embedded in OCT compound. Six cornea specimens of WT or KO mice at each time point were used for RNA extraction. Two corneas served as one RNA specimen at each experimental condition. 

Immunohistochemistry/Cytochemistry.

Deparaffinized sections and fixed cultured cells were processed for immunohistochemistry, as previously reported. ^{6 7} Antibodies used were goat polyclonal anti–OPN (N-terminal peptide) antibody (1:100 dilution in PBS; Santa Cruz Biotechnology, Santa Cruz, CA), rat monoclonal anti–macrophage F4/80 antibody (clone A3-1; 1:400 dilution in PBS; BMA Biomedicals, August, Switzerland), and mouse monoclonal anti–α-smooth muscle actin (αSMA antibody; Neomarker, Fremont, CA). 

Western Blotting for OPN.

Six uninjured corneas and six corneas excised from healing eyes at day 5 or day 10 after alkali burn were lysed in tissue lysis buffer (Sigma, St. Louis, MO), as previously reported. Six corneas at each condition represented one cell lysate sample. Each sample was processed for SDS-polyacrylamide gel electrophoresis and Western blotting, as previously reported. ^{7 33}  

RNA Extraction and Real-Time Reverse-Transcription–Polymerase Chain Reaction.

Expression of cytokine mRNAs was assayed by real-time RT-PCR. Alkali-burned corneas were processed for RNA extraction and real-time RT-PCR, as previously reported (Table 1) . ^{6 7} RNA from two corneas was combined; three samples (six corneas) were prepared for each experimental condition. Six uninjured corneas of three KO and three WT mice were included to obtain the basal expression level of each cytokine. 

Statistical Analysis

Results are expressed as mean ± SD. Mean ± SE is used when results from several experiments are reported together. Student’s t-test was used to compare two groups of animals, whereas ANOVA was used in multiple-group comparisons. P < 0.05 was considered significant. Degrees of significant difference are indicated by asterisks in the figures. 

Results

Effect of Exogenous and Endogenous OPN on Ocular Fibroblast Adhesion and Migration

Coating glass slides with exogenous OPN markedly facilitated the adhesion of cultured mouse ocular fibroblasts (Fig. 1a 1a ′). The absence of endogenous OPN in fibroblasts decreased the number of cells that adhered to chamber slides; this reduction was restored by coating the surface with exogenous OPN (Fig. 1b 1b ′). Corneal epithelial cells are also a critical component for resurfacing an injured cornea. The adhesion of the Araki-Sasaki human corneal epithelial cell line was minimally affected by coating the glass slides with OPN (data not shown). We used a neutralizing anti–OPN antibody to block endogenous OPN on the cell surface. Neutralizing antibody against OPN attenuated the adhesion of the ocular fibroblasts, but not of the Araki-Sasaki corneal epithelial cells, to glass slides (Fig. 2) . 

Fibroblast migration is essential to the repopulation of cells in injured connective tissue. The role of OPN in ocular fibroblast migration was evaluated by use of a scratch assay (Fig. 3) . Closure of the linear defect was delayed in the KO fibroblast monolayer compared with WT cells. Twenty-four hours after wounding, the closure of the defect was delayed in a KO cell culture (Fig. 3F)compared with a WT culture (Fig. 3B) . The difference was then seen at 44 and 59 hours (Figs. 3C 3D 3G 3H) . Such delayed rate of defect closure was recovered by coating the glass slides with exogenous OPN (Figs. 3I 3J 3K 3L) . 

Roles of OPN in In Vivo Wound Healing of the Corneal Epithelium

Our previous study showed that the circular defect in corneal epithelium used here was resurfaced within 24 hours. ^{39 40} Healing of epithelial debridement was similar between KO and WT mice at 20 hours, with a 20% defect remaining (data not shown). This indicates that OPN deficiency minimally affects epithelial healing in the mouse cornea. 

Expression Pattern of OPN in an Alkali-Burned WT Cornea

Using dual immunohistochemistry, we showed abundant expression of OPN protein in a healing cornea after alkali burn. At day 5, OPN was detected in cells that labeled with the F4/80 antibody (macrophages) in the peripheral zone of the healing cornea (Fig. 4A) . In the central area of the cornea, there were almost no mesenchymal cells, but such cells were observed in the corneal periphery close to the sclera. At this time point, some of the cells labeled for both αSMA (myofibroblasts) and OPN, but other myofibroblasts were negative for OPN (Fig. 4B)in this peripheral cornea. At day 10, most OPN-positive cells were F4/80 negative and αSMA positive (Figs. 4C 4D) . At day 20, only a small amount of OPN was detected in the stromal matrix (Figs. 4E 4F) . 

Western blotting with anti–C-terminal OPN antibody showed that very faint or almost no OPN protein was detected in an injured mouse cornea. At days 5 and 10, OPN bands were observed at 65 and 25 kDa (Fig. 5) . 

Healing of KO and WT Mouse Eyes after an Incision Injury in Cornea

We showed that OPN is expressed in myofibroblasts, which are essential for stromal healing. We therefore first examined the role of OPN in the healing of corneal stroma after an incision injury. At day 1 after incision, the injury was filled with fibrin deposition (Figs. 6A 6B) , and at day 3 the injury was not sealed in WT or KO mice (not shown). At day 7, the injured stroma was sealed with granulation tissue in a WT mouse (Fig. 6C) , but it was not closed beneath the regenerated epithelium in a KO mouse (Fig. 6D) . At day 14, the incision was packed with fibrous tissue in a WT mouse (Fig. 6E) . In a KO mouse, the injury was occupied with sparse fibroblastic cells, but there was almost no extracellular matrix deposition among such cells beneath the healing epithelium (Fig. 6F) . At day 21, the wound was healed, showing a fibrotic appearance in a WT mouse (Fig. 6G) , but it was not healed and was still open beneath the regenerated epithelium in a KO mouse (Fig. 6H) . 

Immunohistochemistry showed that myofibroblast generation, as revealed by αSMA expression, was detected at day 3 after incision (Fig. 7a) . It was more frequently observed in WT corneas at days 3 (Fig. 7aC) , 7 (Fig. 7aE) , and 14 (not shown) than in KO corneas (Figs. 7aD 7F) . At day 21, expression of αSMA decreased in a WT cornea, concomitant with stromal healing (Fig. 7aG) . On the other hand, more myofibroblasts were observed adjacent to the stromal wound in a KO cornea (Fig. 7aH) . TGFβ1 and TGFβ2 were detected in the stroma adjacent to the incision wound. Deposition of TGFβ1 and TGFβ2 was also delayed in healing corneas of KO mice compared with WT mice (Fig. 7b) . 

Healing of KO and WT Mouse Corneas after Alkali Burn

We then examined the role of endogenous OPN in burn recovery by using a cornea alkali burn model in mice. In this model, myofibroblasts and inflammatory cells (i.e., macrophages) are considered to be involved in the healing process. There seemed to be no difference in transparency and curvature in uninjured corneas between WT and KO mice (Fig. 8aA B) . Alkali exposure to the ocular surface resulted in total epithelial defect; during healing, stromal edema, ulceration, and neovascularization developed (Fig. 8a) . At day 5, 64% or 73% of the corneas exhibited an abnormal epithelial appearance (including minor punctuate, minor erosion, epithelial defect, and ulceration), but none of them developed perforation in WT or KO mice, respectively (Fig. 8b) . At day 10, the healing stroma was almost entirely resurfaced with remaining stromal opacification and hemorrhage in the anterior chamber in a WT cornea (Fig. 8b 8A) , whereas the KO cornea showed epithelial defects (Fig. 8b 8B) . At this time, 64% (7 of 11) of the corneas in WT mice were abnormal because of epithelial appearance (6 of 11) or perforation (1 of 11). In KO mice, 91% (10 of 11) of the corneas were abnormal because of epithelial appearance (3 of 11) or perforation (7 of 11; Fig. 8b ). At day 20, the corneas of KO mice (Fig. 8a 8F)still exhibited more marked neovascularization than did those of WT mice (Fig. 8a 8E) . At this time, 27% (3 of 11) of WT mice had epithelial abnormalities and none of them exhibited perforation, whereas 64% of KO corneas still had epithelial abnormalities (4 of 11) or perforation (4 of 11). The incidence of perforation was prominently higher in KO mice than in WT mice, as examined by χ² testing. In some mice, the perforation site was closed with regenerated epithelium at day 20 (Fig. 8b) . 

Fig 8cshows the typical histology of healing corneas. Marked inflammation was seen in the stroma in WT (Fig. 8c 8A)and KO (Fig. 8c 8B)corneas at day 5. Regardless of the mouse genotype, myofibroblast generation and macrophage infiltration were more marked in a cornea with stromal perforation than in a nonperforated cornea. 

Real-time RT-PCR was conducted to evaluate the expression of matrix components and cytokines (Fig. 9) . Real-time RT-PCR showed that collagen Iα2 mRNA was less in a KO cornea than in a WT cornea at day 20. In addition, the expression of mRNA of TGFβ1 was less in KO mice at days 10 and 20 compared with expression in WT mice. The expression level of connective tissue growth factor (CTGF) was overall less in KO tissue in two of three experiments between WT and KO corneas. Vascular endothelial growth factor (VEGF) mRNA expression seemed more marked in KO corneas than in WT corneas. Expression of monocyte/macrophage chemoattractant protein-1 (MCP-1) was more prominent in WT corneas than in KO corneas throughout the healing interval. 

Discussion

In the present study, we first showed that loss of OPN impaired adhesion and migration in cultured mouse ocular fibroblasts. Exogenous OPN recovered impaired adhesion of KO fibroblasts, indicating that the loss of OPN does not affect OPN receptor expression. On the other hand, cell culture experiments and in vivo epithelium debridement experiments in corneal epithelium showed that OPN might not have significant roles in the adhesion and healing of corneal epithelium. We then examined whether absent OPN has a significant effect on stromal healing in injured corneas in mice. 

After an incision injury in a WT cornea, keratocytes transform to myofibroblasts, express matrix components, and accelerate wound closure. The present study showed that OPN of 65-kDa and of 25-kDa fragment upregulation in a healing, alkali-burned cornea and that loss of OPN impaired myofibroblast generation and wound closure in association. Generation of myofibroblasts from fibroblasts is a hallmark of granulation tissue formation and tissue scarring. ⁵ Our unpublished data (2007) showed that the loss of OPN reduced the phosphorylation of Smad3, but not Smad2, and that it did not affect the expression of αSMA, mediated by Smad2 signal in each cell, on exposure to TGFβ1 in vitro. ⁴¹ Nevertheless, in the in vivo incision-injured corneal stroma, myofibroblast generation was much reduced. Although the reason for this discrepancy is still unknown, the explanation includes the following. Blocking of phosphorylation of Smad3 by absent OPN might decrease TGFβ in the injured tissue, which might inhibit keratocyte–myofibroblast conversion. Similar impaired Smad signal and tissue fibrosis in a KO mouse were observed in the lens epithelium. ⁴¹  

We then showed that corneas of KO mice exhibited more prominent tissue destruction of the cornea, including a higher incidence of ulceration and perforation, than those of WT mice in a healing alkali-burned eye. Because we showed that the loss of OPN did not affect epithelial healing, the defect in post-alkali burn corneal healing in KO mice is considered to have resulted from the effects of absent OPN on the stromal healing process. Cell adhesion, migration, and proliferation are critical for cell repopulation in an injured tissue. It is not evident from our studies whether the impairment of cell adhesion and migration in OPN-deficient ocular fibroblasts was responsible for the underlying higher incidence of ulceration and perforation in an alkali-burned cornea compared with that in WT mice. 

Results of the in vivo experiment also showed that expression of collagen Iα2 and TGFβ1 was delayed at day 10 or day 20, or both, in KO healing cornea compared with WT healing cornea. Although the whole mechanism of stromal destruction in the KO cornea after alkali burn is still unknown, one possible scenario may be related to this impaired matrix expression in healing tissue. ^{42 43 44} A cytokine expression profile also showed that the loss of OPN markedly decreased mRNA expression of MCP-1 in a healing corneal tissue, though CTGF expression was also mildly suppressed. Decreased TGFβ1 and MCP-1 in tissue might reduce the recruitment of macrophages that are the main source of TGFβ1 in healing tissue. On the other hand, VEGF expression was more prominent in KO corneas than in WT corneas. Although the reason for this phenomenon is still to be determined, marked neovascularization in a KO cornea that healed from perforation might be related to this marked VEGF expression. 

Similar incomplete reconstruction of matrix structure in KO mice was observed in experimental fibrosis models. For example, the tensile strength of healed skin was reduced in KO mice compared with WT mice. ²¹ Similarly, KO mice reportedly develop altered bleomycin-induced lung fibrosis characterized by cystic dilated air spaces, decreased type I collagen expression, and reduced levels of active TGFβ1 and MMP-2. ^{22 25} Another finding to support the idea that loss of OPN impairs reconstruction of the extracellular matrix framework during tissue repair was found in ventricular remodeling after experimental myocardial infarction. ²⁷ In KO mice, the left ventriculum exhibited dilated cardiomyopathy in association with reduced collagen expression and deposition compared with WT mice. However, in models with liver injury by bile duct ligation or CCl₄ treatment, posttreatment liver fibrosis was reportedly more prominent in KO mice than in WT mice, even though activated hepatic stellate cells are capable of OPN upregulation, ^{45 46 47} suggesting that the role of OPN in profibrogenic tissue reaction depends on each organ or tissue. 

 Supported by Ministry of Education, Science, Sports and Culture of Japan Grants C11591871 (SS) and C16590150 (KI), National Institutes of Health Grant EY13755, Research to Prevent Blindness, and Ohio Lions Eye Research Foundation (WW-YK).

 Submitted for publication August 5, 2007; revised October 11, November 10, and December 5, 2007; accepted February 28, 2008.

 Disclosure: K. Miyazaki, None; Y. Okada, None; O. Yamanaka, None; A. Kitano, None; K. Ikeda, None; S. Kon, None; T. Uede, None; S.R. Rittling, None; D.T. Denhardt, None; W.W.-Y. Kao, None; S. Saika, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Table 1.

 

Sequence of Primers for RT-PCR

Transcript	Sequence
mTGFβ1	F: 5′-gca aca tgt gga act cta cca gaa-3′
	R: 5′-gac gtc aaa aga cag cca ctc-3′
	P: 5′-acc ttg gta acc ggc tgc tga ccc-3″
mMCP-1	F: 5′-tgg ctc agc cag atg cag t-3
	R: 5′-cca gcc tac tca ttg gga tca-3′
	P: 5′-ccc cac tca cct gct gct act cat tca-3′
mVEGF	F: 5′-agc gga gaa agc att tgt ttg-3′
	R: 5′-caa cgc gag tct gtg ttt ttg-3′
	P: 5′-cca aga tcc gca gac gtg taa atg ttc c-3′
mCol 1α2	F: 5′-aag ggt ccc tct gga gaa cc-3′
	R: 5′-tct aga gcc agg gag acc ca-3′
	P: 5′-cag ggt ctt ctt ggt gct ccc ggt at-3′

 F, forward primer; R, reverse primer; P, probe.

Figure 1.

 

Effects of loss of OPN in cell adhesion. (a) Adhesion of WT mouse ocular fibroblasts on glass slides (A) is promoted by coating the glass with OPN (B). (a′) Number of adhered cells in 10⁴ μm² in each condition. (b) Number of adhered OPN-null (KO) mouse ocular fibroblasts (B) is much less than that of WT mouse ocular fibroblasts (A). Impaired adhesion of OPN-null (KO) fibroblasts is recovered by coating the glass slide with exogenous OPN (C). (a, b) HE staining. Scale bar, 100 μm. (a′, b′) Graphs indicate the results of counting the numbers of adherent cells in 10⁶ μm² in each experiment. ***Statistically significant difference, P < 0.005.

Figure 2.

 

Effects of anti–OPN neutralizing antibody on adhesion of ocular fibroblasts and a corneal epithelial cell line. (a) Anti–OPN neutralizing antibody blocks adhesion of cultured mouse ocular fibroblasts ([A] control culture with nonimmune IgG vs. [B] anti-OPN antibody culture). (a′) Number of adhered cells counted, as described in Materials and Methods. (b, b′) Anti–OPN antibody does not affect the adhesion of the Araki-Sasaki corneal epithelial cell line to a glass slide ([A] control culture with nonimmune IgG vs. [B] anti–OPN antibody culture). HE staining. Scale bar, 100 μm.

Figure 3.

 

Effects of OPN on fibroblast migration. Cell migration of mouse ocular fibroblasts as evaluated by scratch assay. Migration of KO mouse ocular fibroblasts (E–H) into the cell layer defect was delayed compared with the WT mouse ocular fibroblasts (A–D). Such delay is recovered by coating the glass with exogenous OPN (I–L). Upper layer of hours indicates the time after wounding. HE staining. Scale bar, 200 μm.

Figure 4.

 

Expression pattern of OPN in an alkali-burned WT cornea. With the use of dual immunohistochemistry, we showed abundant expression of OPN protein in a healing cornea after alkali burn. (A) At day 5, OPN was detected in cells labeled with F4/80 antibody (i.e., macrophages) in the peripheral zone of the healing cornea (yellow). In the central area of the cornea, there were almost no mesenchymal cells, but such cells were observed in the corneal periphery close to the sclera. (B) At this time point, αSMA-labeled myofibroblasts were positive for OPN (arrows). (C, D) At day 10, most OPN-positive cells were F4/80 negative and αSMA positive. (E, F) At day 20, only a small amount of OPN was detected in the stromal matrix, but cells were negative for OPN. Nuclei; DAPI nuclear staining.

Figure 5.

 

Detection of OPN in uninjured and healing mouse corneas by Western blotting. Western blotting with anti–C-terminal OPN antibody showed that faint or almost no OPN protein was detected in an injured mouse cornea. At days 5 and 10, OPN bands were observed at 65 and 25 kDa.

Figure 6.

 

Healing of an incision-injured cornea of WT or KO mouse. At day 1 after incision, the wound was occupied with epithelial cells and fibrin in WT (A) and KO (B) mice. At days 7 and 14, the wound was sealed with fibrotic scar tissue in a WT mouse (C, E), whereas the stromal wound was not closed in a KO mouse (D, F). At day 21, the scar tissue formed in the incision wound seemed more packed in a WT cornea (G), whereas the injury was partially healed in a KO cornea (H). Scale bar, 100 μm.

Figure 7.

 

Healing of an incision-injured cornea of WT or KO mouse as evaluated by using immunohistochemistry. (a) Myofibroblast distribution by FITC-immunofluorescence detection of αSMA expression. At day 1 after incision, almost no myofibroblasts are observed in the injured stroma in WT (A) or KO (B) mice. At day 3 after incision, myofibroblasts labeled with anti–αSMA antibody are detected in a WT stroma but not in a KO stroma. At day 7, αSMA-positive cells are more frequently seen in a WT stroma than in a KO stroma. At day 21, however, more myofibroblasts are observed in a KO stroma than in a WT stroma. Nuclear staining; DAPI. (b) Immunohistochemical detection of extracellular, secreted type of transforming growth factor β1 (TGFβ1) and TGFβ2 in the injured stroma at day 7. Deposition of TGFβ1 and TGFβ2 proteins adjacent to the injury is more marked in WT mice than in KO mice. Scale bar, 100 μm.

Figure 8.

 

Healing of an alkali-burned cornea of WT or KO mouse. (a) Stromal opacification and neovascularization in healing corneas. There seems no difference of corneal transparency and curvature in an uninjured cornea between a WT (A) and a KO (B) mouse. (C) WT cornea at day 10 after burn that is free from epithelial defect. KO cornea at day 10 (D) exhibits a central ulceration. At day 20, a WT cornea is opaque but with less neovascularization (E), whereas a KO cornea exhibits dense neovascularization (F). (b) Incidence of epithelial defect/ulceration (A) or stromal perforation (B). The incidence of both defect/ulceration and stromal perforation were lower in WT mice than in KO mice. *P < 0.05; **P < 0.01. (c) Typical histology of the healing corneas at each interval stained with hematoxylin. At day 5, the central corneal epithelium is abnormal and the stroma shows hypercellularity, primarily marked inflammation. At this time point, there was no difference in the findings between WT and KO mice. At day 10, a WT cornea was resurfaced with stromal hypercellularity (C). (D) Granulation tissue formed at the site of perforation in the KO central cornea, where many inflammatory cells accumulated. At day 20, a WT cornea was resurfaced with reduction of cells in the stroma (E). (F) The perforation site was packed with fibrotic tissue and covered with a regenerated epithelium in the center of the cornea in a KO mouse. Scale bar, 100 μm.

Figure 9.

 

Expression pattern of wound healing-related components in an injured stroma. Real-time RT-PCR shows that collagen I α2 mRNA was less abundant in a KO cornea than a WT cornea at day 20. Expression of mRNA of TGFβ1 was lower in KO mice at days 10 and 20 than in WT mice. Expression level of CTGF was similar between WT and KO corneas at each time point in one series, but it was lower at days 10 and 20 in other two series of samples. Expression of VEGF mRNA was higher at day 10 but not at day 20. Expression level of MCP was markedly lower in KO tissues at days 10 and 20. All data shown are from three time repeats from an independent series of samples.