Rho-Mediated Regulation of TGF-β1– and FGF-2–Induced Activation of Corneal Stromal Keratocytes

Abstract

purpose. To investigate the role of Rho GTPase signaling in FGF-2– and TGF-β1–induced activation of corneal keratocytes.

methods. Keratocytes isolated from rabbit corneal stroma and plated in a serum-free medium were treated with FGF-2/heparin or TGF-β1 in the presence or absence of Rho inhibitor (C3 exoenzyme) or ROCK (Rho kinase) inhibitor (Y27632). Specific phenotypic changes were analyzed by immunocytochemistry and Western blot analysis, and the relative abundance of specific mRNAs was estimated by quantitative RT-PCR.

results. TGF-β1–induced expression of α-SMA and transcription of α-SMA mRNA in activated keratocytes were reduced by Rho or ROCK inhibition during the activation. In nonactivated keratocytes, the expression of α3(IV) collagen was downregulated by Rho-inhibition. TGF-β1– or FGF-2–induced downregulation of the expression of α3(IV) collagen and its mRNA was not significantly altered by Rho or ROCK inhibition. TGF-β1– and FGF-2–induced decreases in cell-associated and secreted KS, and lumican mRNA levels were prevented by Rho or ROCK inhibition. However, FGF-2–induced decreases in keratocan mRNA levels were prevented by Rho inhibition but not by ROCK inhibition. Whereas Rho inhibition downregulated both TGF-β1– and FGF-2–induced tenascin-C expression, ROCK inhibition was found to downregulate only TGF-β1–induced expression.

conclusions. Rho signaling has a significant role in the activation of keratocytes. Rho, via ROCK-independent and/or -dependent pathways differentially regulates the TGF-β1–induced expression of α-SMA and TGF-β1– and FGF-2–induced de novo expression of tenascin-C and the downregulation of α3(IV) collagen and KSPGs, lumican and keratocan.

Corneal stroma cells (keratocytes) responsible for the development of a well-organized transparent extracellular matrix become relatively quiescent in the adult cornea. However, after an injury to the cornea, growth factors and cytokines originating from corneal epithelial cells, inflammatory cells, and tear fluid activate keratocytes to transdifferentiate into fibroblasts and myofibroblasts. ^{1 2 3 4 5} On activation, keratocytes lose their dendritic morphology and assemble a dynamic network of actomyosin filaments that regulate important cellular processes including migration and proliferation. Assembly and organization of actomyosin filaments are regulated by Rho GTPases. ^{6 7 8 9} Phenotypic characteristics of the activated cells in healing wounds differ from those exhibited by the stromal cells involved in the fetal and postnatal development of the normal cornea. These differences include not only increased expression of normal extracellular matrix (ECM) components but also de novo expression of ECM components that are not present in the normal cornea such as type III collagen, ^{10 11 12 13} tenascin-C (Tervo K, et al. IOVS 1994;45:ARVO Abstract 2012), ^{12 14 15} and matrix metalloproteinases. ^{16 17 18 19 20 21} Conversely, the changes also include downregulation or loss in the expression of certain normal stromal components, including keratan sulfate proteoglycans (KSPGs), ^{4 22 23 24} prostaglandin D synthase, ²⁵ α3(IV) collagen, ²⁶ and the corneal crystallins, aldehyde dehydrogenase I and III and α-trans ketolase. ²⁷  

When cultured in serum-free (SF) medium corneal keratocytes retain several of their phenotypic characteristics including dendritic morphology. ²⁸ Activation by serum or specific growth factors induces cultured keratocytes to attain morphologic and phenotypic characteristics exhibited by activated keratocytes (fibroblasts or myofibroblasts) in vivo. For example TGF-β1 and FGF-2 can induce increased synthesis of fibronectin, assembly of focal adhesions, ^{28 29 30} and loss in the expression of ALDHI and -III ³¹ and α3(IV) collagen. ²⁶ TGF-β1 also promotes the expression of α-smooth muscle actin (SMA). ²⁹ Although specific growth factors and cytokines that induce these phenotypic changes have been identified, current knowledge of the signaling pathways that induce these changes is limited. In recent years it has become evident that Rho signaling not only regulates the assembly of actomyosin filaments and focal adhesions but also influences gene expression, cell proliferation, endocytic and exocytic pathways, and cellular transformation. ^{32 33 34 35} To determine the role of Rho signaling in the activation of keratocytes, we evaluated the regulation of growth factor–mediated phenotypic changes, including the expression of α-SMA and tenascin-C and loss of the expression of α3(IV) collagen and KSPGs (lumican and keratocan) in cornea keratocytes. 

Materials and Methods

Cell Culture

Keratocytes were isolated and cultured according to a modified procedure of Jester et al. ³⁰ The corneas were excised from whole rabbit eyes obtained from Pel Freez Biological (Rogers, AR), cut into two halves, and the endothelium, along with the Descemet’s membrane, and epithelium, along with a thin layer of underlying stroma, were removed. Stromal pieces were then digested in 0.25% collagenase (Sigma-Aldrich, Inc., St. Louis, MO) at 37°C for 16 to 18 hours. After centrifugation at 1200 rpm for 7 minutes, the pellets containing keratocytes were resuspended in DMEM/F12 (with 0.021% GlutaMax [Invitrogen/Gibco, Carlsbad, CA] and 0.011% pyruvate), and penicillin/streptomycin (Invitrogen/Gibco), and the suspension was filtered through a cell strainer (70 μm; BD Biosciences-Falcon, Bedford, MA). Keratocytes were plated onto 60-mm dishes (Falcon Primaria; BD Biosciences) in SF DMEM/F12 containing 0.1 mM l-ascorbic acid 2-phosphate. 

Treatments with C3 Transferase and Y-27632 (Rho and ROCK Inhibitors)

Rho or ROCK activity was inhibited by treating the keratocytes with C3 or Y27632, respectively. The optimum concentrations of these inhibitors to prevent assembly of stress fibers in the keratocytes activated with FGF-2 were determined by testing a range of the inhibitor concentrations. C3 transferase inhibits Rho activity by ADP-ribosylation on asparagine 41 and ribosylated RhoA can be distinguished by SDS-PAGE as a slower migrating band than nonribosylated RhoA. Western blot analysis showed that in the extracts of cells treated with 0.5 μg C3 only one RhoA-reactive band was detectable, which migrated slower than the RhoA band in the control nontreated cells. C3 and Y27632 were not cytotoxic to the cells at the selected concentrations as determined by staining with trypan blue (dead cells stain blue, and live cells exclude the stain). To treat the cells with the inhibitors, we replaced the culture medium covering the keratocytes replaced after 24 hours with either fresh SF DMEM/F12 or DMEM/F12, with or without 0.5 μg/mL of cell-permeable C3 transferase (Cytoskeleton, Inc., Denver, CO) or 10 nM of Y27632 (EMD Biosciences Inc., Gibbstown, NJ). After 4 hours of incubation, the medium covering the keratocytes to be activated were supplemented with either TGF-β1 (8 ng/mL) or FGF-2 (40 ng/mL)+HS (5 μg/mL). 

Immunostaining

For immunocytochemical analyses the dishes with cells were rinsed with phosphate-buffered saline (PBS), the cells were fixed with 4% paraformaldehyde and permeabilized with PBS buffer containing 0.2% Triton X-100 as described previously. ²⁶ The cells were then reacted with 10% heat-inactivated goat serum in PBS (pH 7.5) for 45 minutes, rinsed with PBS, and treated with the primary and secondary antibodies, as described previously. ³⁰ Primary antibodies included monoclonal rat anti-human α3(IV) collagen (H31+H32), ³⁶ monoclonal mouse anti-KS (J19 or J10), ³⁶ rat anti–tenascin-C (R&D Systems, Minneapolis, MN) and anti–α-SMA antibody (Sigma-Aldrich). The secondary antibodies were Alexa 488- or Alexa 546-conjugated goat anti-rat, anti-rabbit, or anti-mouse IgG (Molecular Probes Inc./Invitrogen) at 1:2500 concentrations. For double-immunofluorescence staining, actin filaments were stained with Alexa 546 phalloidin (Molecular Probes/Invitrogen) included at 1:50 dilutions with the secondary antibody. Coverslips were mounted on the top of the cells (Immunomount; Shandon, Pittsburgh, PA). Fluorescent z-stack images were collected at 0.25-μm intervals with a confocal scanning laser system (Radiance 200; Bio-Rad, Hercules, CA) attached to an inverted microscope (IX70; Olympus, Tokyo, Japan) with the same settings used to facilitate comparisons. The immunofluorescent images presented were projected from z-stacks. 

Western Blot Analyses

Keratocytes grown under different culture conditions were extracted in RIPA buffer (9.1 mM dibasic sodium phosphate, 1.7 mM monobasic sodium phosphate, 150 mM NaCl, [pH, 7.4], 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 0.03 TIU/mL aprotinin [Sigma-Aldrich]), 1 mM sodium orthovanadate, and 100 μg/mL PMSF), in the protocol recommended by Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Aliquots of cell extracts containing 10 or 20 μg of proteins were subjected to SDS-PAGE. For the analyses of secreted keratan sulfate, the volume of culture supernatants in each was adjusted with PBS to a final volume of 5 mL/10⁶ cells or 5 mL/100 μg proteins in the cell extracts. Forty microliters of culture supernatant was used per lane for Western blot analyses. Protein bands were electrophoretically transferred by SDS-PAGE to a membrane (Immobilon-P; Millipore Corp., Bedford, MA), stained with 0.1% Coomassie blue 250 in 40% methanol–1% acetic acid for 1 minute, destained one to three times with 50% methanol-1% acetic acid, and scanned for densitometric comparisons to check equal loading of total proteins. The gels were also stained similarly to ascertain complete transfer of protein bands. The blots were then destained with methanol and subjected to Western blot analysis. The immunoreactive bands were detected with chemiluminescence (Immobilon Western Chemiluminescent HRP Substrate; Millipore Corp.), according to the manufacturer’s protocols. 

Quantitative RT-PCR

Cells were collected by scraping into cold RLT, and RNA was then isolated (RNeasy Mini kit; Qiagen, Valencia, CA). Quantitative RT-PCR for the mRNAs of α3(IV) collagen, ALDH1, lumican, and keratocan were performed by using SYBR Green RT-PCR reagents (Applied Biosystems, Inc. [ABI], Foster City, CA) according to the manufacturer’s instructions. The reactions were performed with a detection system (Model 7700; ABI) for 45 cycles of 15 seconds at 95°C and 60 seconds at 60°C after initial incubations for 10 minutes at 95°C. 

The reaction mixture ³⁷ contained 1× power SYBR Green PCR master mix, 1.0 μL of cDNA, and forward and reverse primers (Table 1)at optimized concentrations. Gene-specific C _t values were standardized based on 18S C _t values obtained for each cDNA. Data were recorded as the mean ± SD of the analyses of the RNA samples (run in duplicate or triplicate) from three separate experiments. After amplification, a dissociation curve for each reaction was generated to confirm the absence of nonspecific amplification. Gel electrophoresis confirmed the amplified product to be of the expected size. Primers used for these experiments are given in Table 1 . Since the sequence of rabbit keratocan had not been reported, its partial sequence was determined by amplifying a short region of rabbit cDNA encoding keratocan by PCR (GenBank DQ239829; http://www.ncbi.nlm.nih.gov/Genbank; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD). Primers for generating the PCR product were designed from the conserved sequences of this gene in other animal species. 

Statistical Analysis

All data are presented as the mean ± SD. Statistical analysis of data from three or more separate experiments was performed with repeated-measures ANOVA. The differences were considered significant at P ≤ 0.05. 

Results

Rho and Rho/ROCK Signaling in the Regulation of TGF-β1–Induced α-SMA Expression

As expected from previously reported findings ^{26 29} keratocytes cultured in SF medium or medium with 0.05% FBS exhibited dendritic morphology and expressed negligible levels of α-SMA. TGF-β1 induced more than 90% of the keratocytes (cultured 0.05% FBS medium) to attain the myofibroblast phenotype, as evident from the expression of α-SMA and development of an extensive network of stress fibers containing α-SMA (Fig. 1) . To evaluate whether these phenotypic changes are regulated by the activation of Rho signaling pathway(s) during the activation of keratocytes with TGF-β1, we analyzed the effects of inhibition of Rho activity or Rho/ROCK activity with C3 and Y27632, respectively. ROCK inhibition resulted in the assembly of significantly fewer stress fibers and reduced levels of α-SMA as evident from the results of immunostaining (Fig. 1) . RhoA inhibition with C3 resulted in the inhibition of morphologic changes, development of stress fibers and expression of α-SMA associated with transdifferentiation of keratocytes into myofibroblasts (Fig. 1) . Reduced synthesis of α-SMA resulting from Rho/ROCK and Rho inhibition was confirmed by Western blot analysis of the cell extracts (Fig. 2A) . The levels of α-SMA mRNA were also lower by 65% ± 10% in ROCK-inhibited cells and by 87% ± 7% in C3-inhibited cells (Fig. 2B) . 

Rho and Rho/ROCK Signaling in the Regulation of FGF-2– and TGF-β1–Induced Loss of α3(IV) Collagen Expression

Activation of rabbit keratocytes by FGF-2 and TGF-β1 to differentiate into fibroblasts and myofibroblasts, respectively, leads to a substantial loss in the expression of α3(IV) collagen. ²⁶ The expression of α3(IV) collagen in keratocytes cultured in SF medium was evident from immunostaining, and no significant change in immunostaining was evident in the keratocytes treated with the ROCK inhibitor, Y27632 (Fig. 3) . However, the inhibition of Rho with C3 significantly reduced the staining intensities of α3(IV) collagen in the keratocytes (Fig. 3) . The inhibition of α3(IV) collagen expression induced by FGF-2 was not prevented by treating the cells with Y27632 or C3 during the activation (Fig. 3) . Similarly, TGF-β1–induced loss in α3(IV) collagen was not prevented by either ROCK or Rho inhibition (data not shown). 

Quantitative mRNA analyses indicated that C3 treatment resulted in a significant decrease (84% ± 8%) in the level of mRNA in nonactivated keratocytes. TGF-β1– or FGF-2–induced loss of expression of α3(IV) collagen was not affected by Rho or ROCK inhibition (Fig. 4) . 

Rho and Rho/ROCK Signaling in the Regulation of FGF-2– and TGF-β1–Induced Loss of KSPG Expression

KSPGs are secreted by keratocytes in culture medium and can be detected by Western blot analyses. ²³ To evaluate cell-associated and intracellular KS, immunostaining of keratocytes with anti-KS was performed (Fig. 5A) . KS was detectable on the cell surface (as evident from a similar staining pattern with nonpermeabilized cells, data not shown) and intracellular perinuclear regions of the cells. Brighter staining was seen in the keratocytes treated with Y27632 and C3. However, an aggregated pattern of staining in the matrix and on the cell surface was detected in C3-treated keratocytes. Cell-associated staining was significantly decreased in TGF-β1– and FGF-2–activated keratocytes (Figs. 5B 5C , respectively). Keratocytes activated in the presence of Y27632 and C3 immunostained with significantly more intensity than the controls. Treatment with C3 during activation led to deposition of KSPG in aggregates, as seen in C3-treated nonactivated keratocytes. 

Western blot analyses of the culture supernatants with anti-KS antibody indicated that significantly elevated concentrations of KSPGs were secreted by the C3-treated keratocytes than by the control cells. In FGF-2– and TGF-β1–activated cells, the concentrations of KS PGs secreted over a period of 48 hours were substantially lower than those secreted by nonactivated keratocytes. However, considerably higher levels of KSPGs were evident in the culture supernatants of keratocytes activated in the presence of Y27632- and C3-treated cells than in the corresponding nontreated cells (Fig. 6) . Quantitative analyses of the levels of intracellular mRNAs of lumican and keratocan demonstrated that FGF-2 and TGF-β1 activation resulted in a significant decrease in keratocan mRNA levels (70% ± 17% and 79% ± 4% lower, respectively). Lumican levels decreased to a lesser extent (by 55% ± 11% and 60% ± 10%, respectively). Cells activated with either FGF-2 or TGF-β1 in the presence of Y27632 had higher lumican levels (117% ± 26% and 70% ± 16% higher) than did the corresponding controls without inhibitor, but keratocan levels were not significantly different. When Rho activity was inhibited by C3 during activation by FGF-2, keratocan and lumican mRNA levels were 327% ± 36% and 365% ± 100% higher, respectively, than in the control cells without any inhibitor. However, when Rho activity was inhibited during the activation of keratocytes with TGF-β1, lumican mRNA levels were 175% ± 25% higher, but keratocan levels were not significantly different from those in the keratocytes activated without any inhibitor. 

Rho and Rho/ROCK Signaling in the Regulation of FGF-2– and TGF-β1–Induced Tenascin-C Expression

Tenascin-C expression, as evaluated by immunostaining, indicated that keratocytes cultured in SF medium expressed negligible levels of tenascin-C (Fig. 7A) . Treatments with Y27632 or C3 did not have any effect on its expression. Although tenascin-C expression was induced by both TGF-β1 and FGF-2, higher levels of expression were induced by FGF-2 (Figs. 7B 7C) . Although TGF-β1–induced expression of tenascin-C was reduced by both ROCK and Rho inhibition, b-FGF–induced expression of tenascin-C was reduced only by Rho inhibition and not by ROCK inhibition. These observations were confirmed by Western blot analyses of the proteins in the cell extracts (Fig. 8A) . An anti–tenascin-C reactive band, in the range of 200 kDa, detected in the FGF-2–activated cells was of significantly stronger intensity than that in TGF-β1–activated cells and the latter was further reduced in Y27632- and C3-treated cells. Reduced tenascin-C levels were also evident in the keratocytes activated with FGF-2 in the presence of C3 but not in the presence of Y27632. Tenascin-C mRNA levels reflected changes in the protein levels, as judged from immunocytochemical and Western blot analyses (Fig. 8B) . 

Discussion

Activation of quiescent keratocytes in response to an injury is accompanied by the assembly coupled with dynamic rearrangement of actomyosin filaments necessary for cellular migration and proliferation. In addition, activated keratocytes express proteins that are not components of normal stroma and downregulate the expression of some of the normal stromal components. It is now well known that Rho-signaling pathways, including the Rho/ROCK pathway, regulate the assembly of actomyosin filaments. It has also becoming evident that Rho has other downstream effects—for example, the regulation of transcription. ^{38 39} In the present study, we evaluated whether some of the phenotypic changes associated with activation of keratocytes are regulated by the activation of the Rho signaling pathway(s) using an in vitro tissue culture model of keratocyte activation. Keratocytes were activated either with TGF-β1 or with FGF-2/HS to attained phenotypic characteristics of myofibroblasts and fibroblasts, respectively. Activation with TGF-β1 and FGF-2 induced tenascin-C expression and downregulated the expression of KSPGs (keratocan and lumican) and α3(IV) collagen. 

Inhibition of Rho small GTPases with a Clostridium botulinum C3 exoenzyme which catalyzes ADP-ribosylation as well as inhibition of ROCKI and -II with a ROCK inhibitor, Y27632, suppressed TGF-β1–mediated expression of α-SMA in keratocytes. Rho-dependent regulation of α-SMA expression has been observed in other cell types—for example, Tenon fibroblasts, ^{40 41} pancreatic stellate cells, ⁴² human gingival fibroblast, ⁴³ prostatic fibroblasts, ⁴⁴ and lung fibroblasts, ⁴⁵ and during epithelial mesenchymal transition (EMT) in lens epithelial cells ⁴⁶ and renal epithelial cells. ⁴⁷ Activation of RhoA and Smad-3 has been found to be necessary for TGF-β1–mediated expression of α-SMA in lens epithelial cells. ⁴⁶ It is not known whether RhoA-regulated α-SMA expression induced by other stimuli involves Smad3. We had found in a prior study that the assembly of actomyosin filaments in transdifferentiated myofibroblasts was regulated by Rho via both ROCK-dependent and -independent pathways. ⁴⁸ The present study indicated that Rho/ROCK signaling regulates de novo expression of α-SMA as well as the assembly of actomyosin filaments containing α-SMA during the differentiation of keratocytes. Although some stress fibers assembled when ROCK activity was inhibited during keratocytes activation only cortical microfilaments were evident when Rho was inhibited. Therefore, a ROCK-independent Rho pathway can regulate assembly of some actomyosin filaments, not containing α-SMA during TGF-β1–induced activation of keratocytes. Differentiation of most keratocytes into myofibroblasts induced by TGF-β1 in the SF medium takes more than 7 days. However, myofibroblast transdifferentiation occurred in less than 48 hours when a small amount of serum (0.05%) was added to the medium. Therefore, a factor(s), possibly epidermal growth factor (EGF), present in the serum had a synergistic effect on Rho-mediated expression of α-SMA. EGF has been reported to act synergistically with TGF-β1 in transdifferentiation of keratocytes to myofibroblasts and the expression of α-SMA. ⁴⁹  

The undesirable phenotypic changes in the activated keratocytes not only include expression of proteins that are not present in normal stroma but also downregulation of the expression of normal proteins in the stroma. We had determined that the expression of α3(IV) collagen, a component of the normal corneal stroma, is diminished on activation of keratocytes by either FGF-2 or TGF-β1. Our studies indicated that Rho inhibition but not Rho/ROCK inhibition downregulated α3(IV) collagen expression in nonactivated keratocytes. Thus, a Rho downstream pathway, other than the Rho/ROCK pathway, regulates the expression of α3(IV) collagen in keratocytes. This signaling pathway may be downregulated on keratocyte activation. Neither Rho inhibition nor ROCK inhibition prevented downregulation of α3(V) collagen expression in activated keratocytes indicating that activation of Rho/ROCK or any other Rho pathway did not lead to inhibition of α3(IV) collagen expression. 

KSPGs were downregulated on keratocyte activation by either FGF-2 or TGF-β1. Immunocytochemical and Western blot analysis with anti-KS antibodies indicated that RhoA and ROCK inhibition during TGF-β1– or FGF-2–induced activation of keratocytes not only prevented downregulation of KSPGs but had a stimulatory effect on their synthesis. Stimulation of KSPG expression by Rho inhibition was more pronounced than that by ROCK inhibition. Thus, activation of Rho/ROCK and possibly a different Rho pathway is involved in downregulation of KS or KSPGs during keratocyte activation. The relative mRNA levels showed that Rho and ROCK inhibition prevented TGF-β1– and FGF-2–induced loss in the expression of lumican, and Rho inhibition prevented FGF-2–induced loss in keratocan expression. These results indicate that Rho/ROCK signaling downregulates expression of lumican induced by both FGF-2 and TGF-β1. The Rho pathway inhibits expression of keratocan induced by FGF-2 but not TGF-β1. Although significantly higher levels of KS were detected in Rho-inhibited keratocytes than in the control keratocytes, there was no corresponding increase noted in either lumican or keratocan mRNA levels, suggesting that Rho signaling may also regulate the sulfation of KS chains and/or glycation of KSPGs. 

Tenascin-C, another marker of fibrosis, ^{50 51} is not present in the normal central cornea but is expressed in healing stromal wounds ^{52 53} and other corneal disorders, such as pseudophakic bullous keratopathy (PBK). ⁵⁴ Tenascin-C forms a hexameric structure of six chains approximately 100 nm in length. Its presence in the stromal matrix is likely to disrupt the spacing of collagen fibrils and affect corneal transparency. Kenney et al. ⁵⁵ showed that insulin-like growth factor (IGF)-I and TGF-β1 upregulate the expression of tenascin-C in the normal corneal fibroblasts and fibroblasts isolated from PBK. Similarly, we determined that while negligible levels of tenascin-C protein and mRNA were detectable in quiescent keratocytes, its expression was induced by both TGF-β1 and FGF-2. However, FGF-2–induced transcription of tenascin-C mRNA was more than six times greater than that induced by TGF-β1. Based on the effects of inhibitors, Rho signaling was found to be involved in both TGF-β1– and FGF-2–mediated upregulation of tenascin-C, although different downstream signaling pathways are regulated by FGF-2 and TGF-β1. However, whereas the Rho/ROCK pathway regulated TGF-β1–mediated expression, a different Rho downstream pathway regulated the FGF-2–mediated expression of tenascin-C. In summary, Rho/ROCK and/or other Rho downstream signaling pathways are involved in differentially regulating phenotypic changes during the transdifferentiation of keratocytes into fibroblasts and myofibroblasts. In addition to actomyosin filament assembly, Rho-regulated changes include the expression of α-SMA and tenascin-C and the downregulation of the expression of KSPGs (lumican and keratocan) and α3(IV) collagen. Therefore, ROCK and Rho inhibitors may prove to be useful therapeutic drugs not only to prevent wound contraction but also to prevent other undesirable phenotypic changes associated with activated keratocyte during wound healing. 

 Supported by National Eye Institute Grants EY03263 (NS), Core Grant EY008098; Research to Prevent Blindness, and The Eye and Ear Foundation (Pittsburgh, PA).

 Submitted for publication December 8, 2008; revised January 26, and February 17, 2009; accepted June 5, 2009.

 Disclosure: J. Chen, None; E. Guerriero, None; Y. Sado, None; N. SundarRaj, 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.

 

Primers used in Quantitative RT-PCR

Primer Names	Oligonucleotide Sequence (5′→3′)
18S	Forward: CTCAACACGGGAAACCTCAC
	Reverse: ACCACCCACAGAATCGAGAA
α3(IV) collagen (GenBank LA7283)	Forward: GCTGTCAACACCAGCTCT GA
	Reverse: CGGTGCACCTGCTAATGTAA
Lumican (GenBank AF020292)	Forward: TGCAGCTTACCCACA ACA AG
	Reverse: TGAAGGTGAACGAAGGTCAA
Keratocan (GenBank DQ239829)	Forward: CTCACGTGGCTTTGATGTGT
	Reverse: GACCTTTGTGAGGCGATTGT
α-Smooth muscle actin (GenBank X60732)	Forward: TTCAATGTCCCAGCCATGTA
	Reverse: TGCCAGTTGTACGTCCAGAG
Tenascin-C (GenBank FJ480400)	Forward: CACGGTGGAGTATGCTCTGA
	Reverse: GGACCTTTCTCCGCAAAGAT

Figure 1.

 

Immunocytochemical analyses showing decreased α-SMA expression in Rho- and ROCK-inhibited TGF-β1–activated keratocytes. Keratocytes isolated from rabbit corneas, cultured in DMEM/F12 without FBS (SF medium) were activated to differentiate into myofibroblasts when the SF medium was supplemented with 0.05% FBS and 8 ng/mL of TGF-β1, without (control) or with the ROCK inhibitor Y27632 or the Rho inhibitor C3. After 48 hours of incubation, the cells were analyzed by double fluorescence staining with mouse anti–α-SMA antibody followed by Alexa Fluor 488 anti-mouse IgG antibodies (right) and Alexa Fluor 546 phalloidin (left). Bar, 50 μm.

Figure 2.

 

Western blot analyses (A) showing decreased α-SMA levels in Rho- and ROCK-inhibited TGF-β1–activated keratocytes. Keratocytes isolated from rabbit corneas, cultured in DMEM/F12 without FBS (SF medium), were activated to differentiate into myofibroblasts by supplementing the culture medium with 0.05% FBS and 8 ng/mL of TGF-β1 without (control) or with the ROCK inhibitor Y27632 or the Rho inhibitor C3. After 48 hours of incubation, the cells were extracted in RIPA buffer, and the extracts (10 μg protein equivalent) were subjected to Western blot analysis with mouse anti–α-SMA antibody. Quantitative RT-PCR analysis (B) showing significantly increased levels of α-SMA mRNAs in TGF-β1–activated cells (P < 0.05), and lower concentrations of mRNA transcripts of α-SMA in ROCK- or Rho-inhibited TGF-β1–activated keratocytes (*P < 0.005). For mRNA analysis, keratocytes in culture were treated as in (A). Levels of α-SMA mRNA were analyzed by real-time RT-PCR in which 18S rRNA levels were used to normalize values in each experiment. The data are presented as the mean ± SD of relative mRNA levels compared with those in control TGF-β1–treated cells from three different experiments.

Figure 3.

 

Immunocytochemical analyses of corneal stromal cells in culture showing inhibition of α3(IV) collagen expression after inhibition of Rho, and no apparent effect of Rho or ROCK inhibitors on α3(VI) collagen loss in FGF-2 (b-FGF)–activated keratocytes. Keratocytes isolated from rabbit corneas, cultured in DMEM/F12 FBS (SF medium), were activated to differentiate into fibroblasts by supplementing the medium with FGF-2 (40 ng/mL)+HS (5 μg /mL) without (control) or with the ROCK inhibitor Y27632 or the Rho inhibitor C3. After 48 hours of incubation, the cells were analyzed by double fluorescence staining with rat anti-α3(ΙV) collagen antibody followed by Alexa Fluor 488 rat IgG antibodies (right) and Alexa Fluor 546 phalloidin (left). Bar, 50 μm.

Figure 4.

 

Quantitative RT-PCR analysis showing significant reductions in α3(IV) collagen mRNA levels after activation of keratocytes with TGF-β1 and FGF-2 (b-FGF; P < 0.05). Rho inhibition resulted in a reduction of α3(IV) collagen mRNA levels (*P < 0.05). Rho or ROCK inhibition had no effect on the diminished α3(VI) collagen mRNA levels in FGF-2– or TGF-β1–activated keratocytes. For mRNA analysis, keratocytes isolated from rabbit corneas, cultured in DMEM/F12 FBS (SF medium), were activated to differentiate into fibroblast or myofibroblasts by supplementing the medium with FGF-2 (40 ng/mL)+HS (5 μg /mL) or with 8 ng/mL TGF-β1+0.05% FBS, respectively. Activation was performed without (control) or with the ROCK inhibitor Y27632 or the Rho inhibitor C3. Relative concentrations of mRNA transcripts of α3(IV) collagen were analyzed by real-time RT-PCR, using 18S rRNA levels to normalize values in each experiment. The data are presented as the mean ± SD of relative mRNA levels compared with those in keratocytes (as 100) from three different experiments.

Figure 5.

 

Immunocytochemicalanalyses showing that ROCK and Rho inhibition increased cell-associated (cell surface and intracellular) keratan sulfate (KS) in keratocytes and prevented FGF-2 (b-FGF)– and TGF-β1–induced decreases in cell-associated KS. Corneal stromal keratocytes isolated from rabbit corneas, cultured in DMEM/F12 without FBS (SF medium) (A) were activated to differentiate into myofibroblasts or fibroblasts by supplementing the medium with 0.05% FBS and 8 ng/mL of TGF-β1 (B) or 40 ng/mL FGF+5 μg /mL HS (C) without (control) or with the ROCK inhibitor Y27632 or the Rho inhibitor C3. After 48 hours of incubation the cells were analyzed by double fluorescence staining with mouse anti-keratan sulfate (KS) antibody followed by Alexa Fluor 488 anti-mouse IgG antibodies (right) and Alexa Fluor 546- phalloidin (left). Bar, 50 μm.

Figure 6.

 

Western blot analyses showing that ROCK and Rho inhibition increased KS secretion in nonactivated keratocytes and prevented TGF-β1– and FGF-2 (b-FGF)–induced decreases in KS secretion. Rabbit corneal keratocytes cultured in DMEM/F12 without FBS (SF medium) were activated to differentiate into myofibroblasts or fibroblasts by supplementing the medium with 0.05% FBS and 8 ng/mL of TGF-β1 or FGF-2 (40 ng/mL)+HS (5 μg /mL) without (control) or with the ROCK inhibitor Y27632 or the Rho inhibitor C3. (A) After 48 hours, culture supernatants were analyzed by Western blot analysis with a mouse anti-KS antibody. Quantitative RT-PCR analysis (B) showing decreased levels of lumican and keratocan mRNAs in TGF-β1– and FGF-2–activated keratocytes (P < 0.05), increased lumican mRNA in ROCK- and Rho-inhibited TGF-β1– and FGF-2–activated keratocytes than in the corresponding control cultures (*P < 0.05) and increased keratocan mRNA levels in Rho-inhibited FGF-2–activated keratocytes (*P < 0.05). For mRNA analysis keratocytes in culture were treated as in (A), and lumican and keratocan mRNA levels were analyzed by real-time RT-PCR using levels of 18S rRNA to normalize values in each experiment. The data are presented as the mean ± SD from three different experiments.

Figure 7.

 

Immunocytochemicalanalyses showing negligible tenascin-C expression in keratocytes (A), and the inhibition of TGF-β1–induced tenascin-C expression by Rho and ROCK inhibition (B), and inhibitor of FGF-2 (b-FGF)–induced expression of tenascin-C by Rho but not by ROCK inhibition (C). Keratocytes isolated from rabbit corneas were cultured in DMEM/F12 without FBS (SF medium) were activated to differentiate into myofibroblasts or fibroblasts by supplementing the medium with 0.05% FBS and 8 ng/mL of TGF-β1 (B) or FGF-2 (40 ng/mL)+HS (5μg /mL) (C) without (control) or with the ROCK inhibitor Y27632 or the Rho inhibitor (C3). After 48 hours of incubation, the cells were analyzed by double fluorescence staining using rat anti–tenascin-C antibody followed by Alexa Fluor 488-conjugated anti-rat IgG antibodies (right) and Alexa Fluor 546-conjugated phalloidin (left). Bar, 50 μm.

Figure 8.

 

Western blot analyses (A) showing inhibition of TGF-β1– and FGF-2 (b-FGF)–induced expression of tenascin-C by Rho and ROCK inhibition. Rabbit corneal keratocytes cultured in DMEM/F12 without FBS (SF medium) were activated to differentiate into myofibroblasts or fibroblasts by supplementing the medium with 0.05% FBS and 8 ng/mL of TGF-β1 or FGF-2 (40 ng/mL)+HS (5 μg /mL) without (control) or with the ROCK inhibitor Y27632 or the Rho inhibitor C3. After 48 hours of incubation, the cells were extracted in RIPA buffer, and the cell extracts (10 μg protein equivalent) were subjected to Western blot analyses with rat anti–tenascin-C antibody. Quantitative RT-PCR analysis (B) showed a TGF-β1–induced increase in tenascin-C mRNA in keratocytes (P < 0.05), which was reduced by inhibition of Rho or ROCK (*P < 0.05). The FGF-2–induced increase in tenascin-C mRNA was reduced by Rho inhibition (*P < 0.05) but not by ROCK inhibition. For mRNA analysis keratocytes in culture were treated as in (A), and lumican and keratocan mRNA levels were analyzed by real-time RT-PCR with levels of 18S rRNA used to normalize values in each experiment. The data are expressed as the mean ± SD from three different experiments.

The authors thank Kira Lathrop, Ryan Eberwine, Julie Wong Chong, and Kate Davoli for technical help and Yoshikazu Sado for the generous gift of anti-α3(IV) collagen antibodies.