Contractility as a Prerequisite for TGF-β–Induced Myofibroblast Transdifferentiation in Human Tenon Fibroblasts

Abstract

purpose. To assess the significance of Rho-kinase–dependent contractility in TGF-β–induced myofibroblast transdifferentiation of human tenon fibroblasts to characterize possible pharmacological targets for the inhibition of postoperative scarring after glaucoma surgery.

methods. Human tenon fibroblasts (HTFs) were grown in culture and stimulated with TGF-β1. The effect of TGF-β on Rho-GTPase activity was assessed by GST-rhotekin binding domain pulldown assay and detected by Western blot analysis. Contractility was evaluated in a silicone substrate wrinkling assay and in fibroblast-populated collagen gels. The actin cytoskeleton and focal adhesions were visualized by immunofluorescence microscopy. α-SMA transcripts were measured by real-time RT-PCR. TGF-β–induced Smad- and p38-activation and expression of α-SMA were detected by Western blot analysis. Nuclear translocation of Smad2/3 was determined by confocal immunofluorescence microscopy. The influence of Rho-dependent kinase (ROCK) and myosin light chain kinase (MLCK) were studied by using specific kinase inhibitors (Y-27632, HA-1077, H-1152, and ML-7).

results. Within 10 minutes of stimulation, TGF-β induced Rho activation that was associated with an increase in cell tension and followed by actin stress fiber enhancement. ROCK inhibitors released cell tension and averted TGF-β–induced cytoskeletal changes, p38 activation and subsequent α-SMA expression, whereas Smad2-phosphorylation and nuclear translocation were preserved. MLCK inhibition also blocked α-SMA expression. In fibroblast-populated collagen lattices, ROCK inhibitors prevented TGF-β–induced stress fiber assembly and contraction.

conclusions. TGF-β induces a rapid contractile response in HTFs that precedes myofibroblast transdifferentiation. ROCK inhibitors release this contraction and block subsequent TGF-β–induced myofibroblast transdifferentiation and may therefore serve to modulate postoperative scarring after glaucoma filtering surgery.

Postoperative scarring is the most frequent cause of failure after glaucoma filtering surgery. Currently, antimetabolites are used to prevent subconjunctival scar formation in patients at risk, but severe side effects limit their application. ¹ The cytokine TGF-β is a pivotal mediator of wound healing and is recognized as a major driving force of postoperative scarring after glaucoma surgery. ^{2 3} A better understanding of the molecular mechanisms underlying conjunctival scarring is desirable, as it may provide additional approaches to safe modulation of wound healing. 

In a process that is crucial to wound healing and scar formation, TGF-β induces the transition of fibroblasts to myofibroblasts characterized by α-smooth muscle actin (α-SMA) expression. ⁴ Myofibroblasts deposit extracellular matrix proteins ⁵ and exert contractile forces to drive the remodeling phase of wound healing. ⁶ Although myofibroblasts are only transiently present in normal wounds, their persistence is associated with scar formation. ⁷ TGF-β–scavenging antibodies or a lack of TGF-β release from macrophages reduces myofibroblast transdifferentiation and prevents scarring in wound-healing animal models. ^{8 9 10} TGF-β binds to its heterodimeric receptor to activate intracellular signaling cascades, such as the classic Smad and mitogen-activated protein kinase (MAPK) pathways. We have shown recently that activation of the MAPK p38 is mandatory for TGF-β–induced myofibroblast transdifferentiation. ¹¹  

Increasing evidence indicates that biomechanical factors are also major determinants of wound healing and myofibroblast transdifferentiation. ⁶ In a rat wound-healing model, myofibroblast transdifferentiation was markedly increased when mechanical strain was exerted on surgical wounds. ¹² Studies on fibroblast-populated collagen gels in vitro have shown that the mechanical stiffness of the gels determines the intensity of TGF-β–induced myofibroblast transdifferentiation and subsequent gel contraction. ¹³ Furthermore, myofibroblast persistence is prolonged in response to mechanical strain. ^{12 14} On the cellular level, mechanical strain on the extracellular matrix is transmitted through integrins to intracellular focal adhesion complexes and the actin cytoskeleton, to alter cell signaling. ¹⁵ This process, termed mechanotransduction, integrates mechanical cues and growth factor–induced intracellular signals to modulate cell functions. Thus, cellular mechanotransduction has a strong influence on wound healing and scarring, and a release of mechanical tension may reduce myofibroblast transdifferentiation. 

The GTPase Rho is a master regulator of the actin cytoskeleton and cell contractility. ¹⁶ It governs the formation of stress fibers and focal adhesions ^{17 18} and is a participant in mechanotransduction. ^{15 19} Rho activation increases cytoskeletal tension by activating Rho-associated serine-threonine kinase (ROCK), to increase myosin light-chain phosphorylation and engage actomyosin contraction. ²⁰ In reverse, ROCK inhibitors decrease cellular tension, which leads to disassembly of actin stress fibers and focal adhesions. ^{21 22} ROCK inhibitors are used systemically to treat cerebral vasospasms. ^{23 24} They also act on the trabecular meshwork and Schlemm’s canal endothelium to decrease intraocular pressure ^{22 25 26} and have been applied topically to the eye without serious side effects. ^{21 22 27 28}  

Herein, we report that pharmacologic inhibition of ROCK prevents TGF-β–induced early cell contraction, cytoskeletal changes, p38 activation, and subsequent myofibroblast transdifferentiation and collagen gel contraction in human tenon fibroblasts. ROCK inhibitors may therefore provide a novel approach to the modulation of wound healing after glaucoma surgery. 

Materials and Methods

Reagents

A total RNA purification kit (RNeasy) and RNase-free DNase were purchased from Qiagen (Hilden, Germany), and Taq Polymerase and M-MLV reverse transcriptase (Impro II) from Promega (Promega, Mannheim, Germany). Antibodies raised against the following proteins were used: α-SMA, vinculin, and tubulin (Sigma-Aldrich, St. Louis, MO.), active p38 (Promega), total p38 (NEB, Frankfurt, Germany), active Smad2/3 (Zymed/Zytomed, Berlin, Germany), total Smad2/3 (Chemicon, Temecula, CA), Alexa-488–conjugated goat anti-mouse (Invitrogen-Molecular Probes, Eugene, OR), and HRP-conjugated secondary (Jackson/Dianova, Hamburg, Germany) antibodies. Recombinant TGF-β1 was obtained from Tebu-Bio (Offenbach, Germany) and used at 2 ng/mL (final concentration) in all experiments. The ROCK inhibitors Y-27632, HA-1077, and H-1152 were purchased from Merck Biosciences (Bad Soden, Germany). 

Cell Culture

Small tenon biopsy specimens were obtained during standard intraocular surgery after selected patients received comprehensive information and provided written consent. The protocol complied with the tenets of the Declaration of Helsinki, and institutional ethics committee approval was granted. Primary HTFs were obtained in an expansion culture of the human tenon explants and propagated in Dulbecco’s modified Eagle’s medium (DMEM; PAA Laboratories GmbH; Pasching, Austria) supplemented with 10% heat-inactivated fetal calf serum (FCS; Biochrom, Berlin, Germany), 100 U/mL penicillin, and 100 μg/mL streptomycin (both from PAA Laboratories). Cells were maintained in logarithmic growth phase. In all experiments, cells from passages 3 to 6 were used. All experiments were performed at least three times with similar results. 

Rho-GTPase Assay

HTFs were plated to subconfluence at 1.2 × 10⁶ cells in 150-mm dishes, serum-starved for 24 hours, and stimulated with vehicle, TGF-β (2 ng/mL), or lysophosphatidic acid (LPA; 10 μM), as indicated. Cells were washed with ice-cold phosphate-buffered saline and processed with a Rho-GTPase assay kit (Upstate Biotechnology-Biomol, Hamburg, Germany), according to the manufacturer’s recommendations. Briefly, cells were lysed on ice in 500 μL lysis buffer containing protease inhibitors (Roche Diagnostics, Mannheim, Germany), lysates were cleared by centrifugation, and equal amounts of supernatant were incubated with the GST-Rho–binding domain of rhotekin (RBD) beads for 40 minutes, washed three times with lysis buffer, and eluted with SDS sample buffer. Rho protein was detected by Western blot analysis with a monoclonal antibody against Rho-A (Upstate Biotechnology-Biomol). 

ROCK Inhibitors

Stock solutions (10 mM) of the ROCK inhibitors Y-27632 and H-1152 were prepared in water and handled under safelight conditions. The inhibitors were diluted in nonsupplemented DMEM, added to the cell culture 30 minutes before the stimulation, and were maintained with the stimuli. 

Substrate Wrinkling Assay

To visualize changes in cell tension, we applied the technique described by Harris et al. ²⁹ Briefly, glass coverslips were coated with polydimethylsiloxan-co-diphenylsiloxan (Sigma-Aldrich). Exposure to a Bunsen burner flame for 2 to 3 seconds was sufficient to cross-link a superficial layer of silicone. Cells were then plated on the coated coverslips in Leibovitz medium supplemented with 10% FCS, allowed to spread for 6 to 12 hours, stimulated with vehicle, TGF-β (2 ng/mL), or Y27632 (10 μM) and viewed with a live cell videomicroscope (Carl Zeiss MicroImaging, Göttingen, Germany). 

RNA Isolation and Real-Time RT-PCR

HTFs were plated in 60-mm dishes at 2 × 10⁵ cells/dish, serum-starved for 12 hours and stimulated with TGF-β in the presence or absence of ROCK inhibitors for 24 hours. Total RNA was harvested using RNeasy spin columns (Qiagen, Hilden, Germany), according to manufacturers’ recommendations. To eliminate contamination with genomic DNA, DNase digestion was performed for 15 minutes. First-strand cDNA was synthesized with reverse transcriptase (Superscript II; Invitrogen, Karlsruhe, Germany) at 42°C with 500 ng total RNA extract. 

Real-time RT-PCR was conducted in a 96-well microtiter plate with a total reaction volume of 25 μL containing 2 μL of a 1:4 dilution of first-strand reaction product, 0.4 μM specific upstream and downstream primers, 240 μM dNTP-mix, 1× reaction buffer, Taq polymerase and a nucleic acid stain (SYBR-green; Sigma-Aldrich) as a fluorescent marker. Amplification of cDNA fragments and analysis was performed on a thermocycler (iCycler; Bio-Rad, Munich, Germany). Cycling conditions were initial denaturation at 95°C for 3 minutes, followed by 40 cycles consisting of a 15-second denaturation interval and an 1-minute interval for annealing and primer extension at 60°C. Amplification of the housekeeping gene hypoxanthine-guanine-phosphoribosyl-transferase 1 (HPRT1) mRNA, which served as a normalization standard, was performed with HPRT1 forward (GACCAGTCAACAGGGGACAT) and HPRT1 reverse (ACACTTCGTGGGGTCCTTTT) primers. Side-strand–specific primers for α-SMA were as follows: α-SMA-forward (CTGTTCCAGCCATCCTTCAT) and α-SMA-reverse (CCGTGATCTCCTTCTGCATT). α-SMA mRNA levels were measured as C _T threshold levels and normalized with the individual HPRT1 control C _Ts. 

Western Blot Analysis

Cells were rinsed with ice-cold PBS, and total cell protein extracts were prepared with a TNE lysis buffer (20 mM Tris, 150 mM NaCl, 1 mM EDTA, and 1% Triton X-100) containing phosphatase inhibitors (1 mM sodium vanadate and 50 mM NaF) and protease inhibitors (0.1% phenylmethylsulfonyl fluoride [PMSF] and Complete protease inhibitor; Roche). Protein concentrations were measured with a BCA assay (KMF, Lohmar, Germany). Protein extracts (10 μg) were boiled in Laemmli sample buffer and subjected to SDS polyacrylamide gel electrophoresis. The proteins were transferred onto a polyvinylidene difluoride (PVDF) membrane (GE Healthcare, Braunschweig, Germany) by using a gel blot apparatus (Bio-Rad). The membranes were blocked in 3% BSA in TBST (10 mM Tris-HCl [pH 7.4], 150 mM NaCl, and 0.1% Tween-20) for 1 hour and were incubated with primary antibody overnight at 4°C and with a peroxidase-conjugated secondary antibody for 45 minutes at room temperature. After each incubation step, the membranes were washed in TBST for 20 minutes. Peroxidase was visualized by chemiluminescence (ECL) and exposed to autoradiograph film (Hyperfilm; GE Healthcare) for the appropriate times. Densitometry analysis was performed using ImageJ 1.32j software (available by ftp at zippy.nimh.nih.gov/ or at http://rsb.info.nih.gov/nih-image; developed by Wayne Rasband, National Institutes of Health, Bethesda, MD). 

Immunofluorescence Confocal Microscopy

Glass coverslips were coated with laminin for fibronectin for 2 hours at 37°C in a cell incubator and washed with PBS. To assess short-term cytoskeletal changes, nonspecific binding sites on laminin-coated coverslips were blocked with 1% heat-denatured BSA for 1 hour at 37°C and washed again. HTFs were serum starved in 0.2% FCS in DMEM for 24 hours, trypsinized using soybean trypsin inhibitor, plated on laminin-coated coverslips, and allowed to spread in starvation medium overnight. To study α-SMA expression and localization, we plated HTFs on FN-coated coverslips and deprived them of serum for 12 hours. For stimulation, cells were preincubated with ROCK-inhibitors or vehicle for 30 minutes and subsequently stimulated with TGF-β (2 ng/mL). Subsequently, cells were fixed in 2% paraformaldehyde, permeabilized in 0.2% Triton X-100, blocked in 2% normal goat serum (Jackson ImmunoResearch, Hamburg, Germany) and labeled with primary antibody against vinculin or α-SMA at a dilution of 1:400 in blocking buffer. Alexa-488 conjugated secondary antibody against mouse IgG was used at a 1:500 dilution in blocking buffer. Rhodamine-TRITC was used to counterstain the F-actin cytoskeleton. Fixed collagen gels were permeabilized with 1% Triton X-100, blocked in 2% normal goat serum, 2% FCS, and 0.2% Triton X-100 and stained with rhodamine-phalloidin. Cells were viewed with a laser scanning confocal microscope (TCS SP-2; Leica Microsystems, Bensheim, Germany). 

Fixed Collagen Gel Cultures

Experiments were essentially conducted as described by Grinnell et al. ³⁰ Briefly, 24-well plates were precoated with 0.2% BSA (KMF, Lohmar, Germany) for 1 hour. HTFs resuspended in DMEM supplemented with 0.2% FCS were added to a neutralized collagen solution of 16 parts type I collagen (Vitrogen-100; Cohesion-Invitrogen, Palo Alto, CA), two parts 10× DMEM (Sigma-Aldrich), and one part NaHCO₃ (Sigma-Aldrich) in a 1:1 ratio, yielding a final concentration of 150,000 cells/mL and 1.2 mg/mL collagen. The collagen cell suspension was added to each well and incubated at 37°C for 1 hour for polymerization. The gels were preincubated with ROCK inhibitor Y27632 (10 μM) or vehicle for 1 hour. After 48 hours of stimulation with TGF-β1 (2 ng/mL) for 48 hours in the presence or absence of Y27632, the gels were detached, and contraction was digitally photodocumented at various time points. Contraction quantification was performed with ImageJ 1.32j software. 

Results

TGF-β–Induced Activation of Rho

To characterize contractility-related signaling events at the outset of TGF-β–induced myofibroblast transdifferentiation, we assessed the influence of TGF-β on Rho, a crucial regulator of the cytoskeleton. ¹⁶ After serum starvation for 24 hours, HTFs were treated with vehicle, TGF-β, or LPA for different times, lysates were prepared, and active Rho was precipitated with GST-RBD and analyzed by Western blot. TGF-β stimulation for 10 minutes strongly increased Rho activity, which had declined little after 60 minutes (Fig. 1) . LPA served as the positive control and induced a robust Rho activation after 5 minutes. 

ROCK-Dependent Cell Contraction

Active Rho can induce cell contraction. To address this possibility, we examined alterations in cell tension in a silicone substrate wrinkling assay. HTFs were plated on a thin layer of polymerized silicone and stimulated by vehicle or TGF-β while cell shape and tension-induced wrinkles in the substrate were recorded by live cell videomicroscopy. Migrating cells exert a well orchestrated tension field on the substrate, as seen in Figures 2A 2B 2C 2D . Within 30 minutes of TGF-β stimulation, cell tension increased in a centripetal contraction pattern, as indicated by the deeper wrinkles, smaller cell size, and increased wrinkle curvature (Figs. 2E 2F , see Fig. 2Hfor illustration). Rho activates ROCKs, which stimulate myosin light chain phosphorylation to promote actin and myosin contraction. ³¹ To assess the relevance of this pathway in TGF-β–induced early cell contraction, we tested the ROCK inhibitor Y-27632 which rapidly reversed the TGF-β–induced changes in this model (Fig. 2G) . 

These data indicate that TGF-β rapidly activated ROCK-dependent cell contraction. 

ROCK-Dependent Early Cytoskeletal Alterations

Next, we studied the effect of TGF-β on actin stress fibers and focal adhesions, since Rho-stimulated contractility is known to mediate cytoskeletal rearrangements. ¹⁸ When plated on stiff collagen- or fibronectin-coated substrates in vitro, HTFs develop strong focal adhesions and stress fibers. To visualize early TGF-β–induced cytoskeletal changes, HTFs were plated on laminin, which evokes delicate cortical stress fibers in control conditions (Figs. 3A 3B 3C) . All cells were serum-starved for 12 hours, trypsinized using soybean trypsin inhibitor (Sigma-Aldrich), plated, and allowed to spread overnight. They were then stimulated with vehicle or TGF-β for 6 hours, fixed, and stained with phalloidin-TRITC and an anti-vinculin antibody. Treatment with TGF-β induced pronounced actin stress fibers and elongated focal adhesions as detected by phalloidin and anti-vinculin staining, respectively (Figs. 3D 3E 3F) . This effect of TGF-β was blocked by Y-27632 (Figs. 3G 3H 3I) . Thus, it appears that ROCK-mediated contractility is required for TGF-β–induced early cytoskeletal changes. 

TGF-β–Induced α-SMA Expression

To explore a possible functional link of early TGF-β–induced contractility and subsequent myofibroblast transdifferentiation, we measured TGF-β–induced α-SMA mRNA (Fig. 4A)and protein (Fig. 4B)levels in the presence or absence of ROCK inhibitors. We used Y-27632 (10 μM), the less specific inhibitor HA-1077 (10 μM), and the ROCK-specific substance H-1152 (10 μM). In the absence of inhibitors, TGF-β induced a robust increase in α-SMA mRNA and protein expression after 12 and 48 hours of stimulation, respectively (Figs. 4A 4B) . This time course indicates that the early changes we observed in contractility and cytoskeletal arrangement precede α-SMA expression. ROCK inhibitors reduced the TGF-β–induced increase of α-SMA mRNA and protein levels. This effect was observed with all inhibitors tested (Fig. 4)and occurred in a dose-dependent manner (data not shown), suggesting that ROCK-mediated contractility is a prerequisite for α-SMA expression. 

To clarify further the role of cell tension in TGF-β–induced myofibroblast transdifferentiation, we tested a ROCK-independent inhibitor of cell contractility, the myosin light chain kinase (MLCK) inhibitor ML-7. α-SMA protein expression was also blocked by ML-7, thus strongly suggesting a general requirement of contractility in this process. 

TGF-β–Induced Recruitment of α-SMA to Stress Fibers

Besides the total level of α-SMA expression, the subcellular localization of α-SMA is of functional importance in myofibroblasts. To enhance cell contractility, α-SMA is incorporated into actin stress fibers. To characterize the effect of ROCK inhibitors on α-SMA recruitment to stress fibers, we visualized α-SMA localization by confocal immunofluorescence microscopy. Untreated fibroblasts plated on fibronectin for 48 hours showed elaborate actin stress fibers and weak diffuse cytoplasmic staining for α-SMA (Figs. 5A 5B 5C) . TGF-β treatment for 48 hours led to pronounced actin stress fibers (Fig. 5E)and the recruitment of α-SMA to actin stress fibers (Figs. 5D 5E 5F) . Incubation with the ROCK inhibitor H-1152 for 48 hours diminished actin stress fibers in control conditions (Figs. 5G 5H 5I) . In the presence of TGF-β, the ROCK inhibitor prevented α-SMA recruitment to stress fibers (Figs. 5J 5K 5L) . 

TGF-β–Induced p38 Activation

To elucidate underlying signaling pathways, we determined TGF-β–induced p38 and Smad2 phosphorylation in the presence or absence of ROCK inhibitors by Western blot analysis with phospho-specific antibodies. TGF-β induced a delayed and sustained activation of a small isoform of p38 (Fig. 6A , arrow), which was blocked by ROCK inhibitors. In contrast, Smad2 activation was rapid, sustained for 24 hours (Fig. 6A) , and was unaffected by ROCK inhibitors (Fig. 6A) . 

Activated Smad2 translocates to the nucleus to influence gene expression. Although Smad phosphorylation appeared unaffected by ROCK inhibitors, a failure in nuclear translocation could also alter Smad signaling, even though Smad2 is phosphorylated. To address this possibility, we determined Smad2/3 localization by immunofluorescence microscopy. In untreated cells, Smad2/3 was found in the nucleus and in the perinuclear region, with a slight, diffuse staining throughout the cell body (Fig. 6B) . On stimulation with TGF-β, the nuclear signal increased strongly, whereas the perinuclear and cytosolic regions were depleted (Fig. 6D ; imaging parameters were identical for all images). ROCK inhibitors did not alter these staining patterns, before or after TGF-β stimulation (Figs. 6C 6E) . Hence, inhibition of ROCK appeared to alter TGF-β–induced p38 phosphorylation without disrupting Smad2 phosphorylation or translocation. 

TGF-β–Induced Contraction and Stress Fiber Formation in Fibroblast-Populated Collagen Gels

The previous experiments indicated the possibility that ROCK inhibitors block myofibroblast transdifferentiation. To investigate this possibility in a functional assay, we studied cell contractility and stress fiber assembly in tethered fibroblast-populated collagen gels. After they were seeded in a neutralized collagen solution, HTFs were incubated in the presence or absence of TGF-β (2 ng/mL), Y-27632 (10 μM), or both for 48 hours. When released after 2 days of incubation (Figs. 7A 7B) , untreated gels contracted to 34% of their initial size within 3 hours. TGF-β induced a strong increase in contractility (gel size, 15%). The ROCK inhibitor Y-27632 blocked the effect of TGF-β on contractility (39%) as well as intrinsic contractility (56% in the absence of TGF-β). Actin stress fibers (Fig. 7D)were apparent in TGF-β–treated cells but were absent in control cells or cells treated with Y-27632 (Figs. 7C 7E) . Thus, ROCK inhibitors prevent the TGF-β–induced emergence of a myofibroblast phenotype on a functional and structural level. 

Discussion

Mechanical tension and rigid extracellular matrices promote myofibroblast transdifferentiation and scar formation. ^{6 12 13} We now provide further evidence of a role of cell tension and contractility in TGF-β–mediated myofibroblast transdifferentiation. 

We first studied Rho activity to assess whether TGF-β is linked to cell tension at the initial stage of myofibroblast transdifferentiation. The small GTP-binding protein Rho is a molecular switch that regulates cell tension and induces actin stress fibers and focal adhesions. ^{16 17} Rho activation rapidly leads to cytoskeletal contraction and force generation by actomyosin engagement, ¹⁸ which is achieved through the activation of ROCK, leading to actin polymerization ³² and phosphorylation of myosin light chain. ³¹ In HTFs, TGF-β treatment strongly increased Rho activity within 10 minutes (Fig. 1) , which is well in line with the observation that growth factors can rapidly activate Rho ¹⁷ and with comparable results that have been obtained in human lens epithelial cells ³³ and hyalocytes. ³⁴  

To explore whether TGF-β–induced Rho activation increases cell contractility, we visualized cell tension directly in a substrate wrinkling assay. Cell adhesions transmit cytoskeletal tension to induce wrinkles in elastic substrates that are easily observed by phase or differential interference contrast microscopy. ²⁹ Migrating HTF exerted orchestrated directional tensile forces on the substrate, to induce fine parallel wrinkles (Figs. 2A 2B 2C 2D) . In the presence of TGF-β, cell tension increased within minutes, and a more global centripetal contraction of the cells ensued, as indicated by deepened, less parallel, curved wrinkles and a decrease in cell surface area (Figs. 2E 2F) . Inhibition of ROCK by Y-27632 reversed this effect of TGF-β and released cell tension (Fig. 2G) , supporting the idea that ROCK inhibitors decrease cell contractility. 

Rho activation is typically followed by cytoskeletal changes. ¹⁷ Accordingly, an increase in stress fibers and focal adhesion size was observed within hours of TGF-β treatment (Fig. 3) . These changes were prevented by the ROCK inhibitor Y-27632. Similar results have recently been reported for TGF-β stimulation of Swiss 3T3 cells. ³²  

Our data on Rho activation, cell tension, and cytoskeletal changes indicate that TGF-β elicits a rapid contractile response in HTFs. This response precedes myofibroblast-driven contracture, which is usually detected only 24 to 48 hours after stimulation in vitro. ^{13 35 36}  

To seek the possible link of early contraction and subsequent myofibroblast transdifferentiation, we examined whether the prevention of early contractile and cytoskeletal responses to TGF-β by ROCK inhibitors would affect subsequent α-SMA expression and distribution. α-SMA is a hallmark of myofibroblasts and is responsible for the enhanced contractile forces exerted by this cell type. ³⁷ All three ROCK inhibitors tested, Y-27632, hydroxy-fasudil (HA-1077), and H-1152 blocked the increase in α-SMA expression elicited by TGF-β after 48 hours (Fig. 4B) . In keeping with our data, ROCK inhibition has recently been shown to downregulate TGF-β–induced α-SMA expression in hyalocytes ³⁴ and human corneal stromal cells. ³⁸ α-SMA expression was also prevented by the ROCK-independent myosin light chain kinase inhibitor ML-7 (Fig. 4B) , suggesting a more general role of contractility as a prerequisite for transdifferentiation. 

ROCK inhibitors also prevent the recruitment of α-SMA to stress fibers (Fig. 5)which is required for increased myofibroblast contractility. After 48 hours of TGF-β treatment in the presence of H-1152, some cells showed a slightly increased staining for α-SMA (Fig. 5J , arrows). However, the staining appeared as a diffuse perinuclear and cytoplasmic signal, whereas stress fibers were negative for α-SMA (Fig. 5J 5K 5L , arrows). A recent study in corneal fibroblasts ³⁹ had similarly reported a ROCK-dependence of stress fiber formation. 

Further work is needed to elucidate the functional implications of an immediate increase in cell tension after TGF-β stimulation. However, our findings and those in other reports suggest that an early contraction may serve to initiate the subsequent transdifferentiation process. A mere application of mechanical strain on the cytoskeleton through collagen-coated magnetized beads was sufficient to induce α-SMA expression in osteoclasts. ⁴⁰ Similarly, tissue strain increased α-SMA expression in a rat wound-healing model. ¹² In reverse, ROCK inhibitors, which release cell tension (Fig. 2) , prevented subsequent α-SMA expression (Fig. 4) . One might speculate that an early cell contraction allows cells to probe tissue elasticity ⁴¹ and adjust subsequent TGF-β–induced myofibroblast transdifferentiation to the level of mechanical strain in the tissue. 

To gain further insight into TGF-β–induced cellular signals that may be modulated by tension, we studied the Smad2 and p38 signaling pathways in the absence and the presence of ROCK inhibitors. As we observed earlier, ¹¹ TGF-β induced a pronounced transient Smad2 activation, whereas the MAPK p38 was persistently activated after a delay of 12 to 24 hours. (Fig. 6A) . ROCK inhibitors did not alter the Smad2 phosphorylation pattern, but inhibited TGF-β–induced phosphorylation of p38. This was true of all three inhibitors tested (data for HA-1077 and Y-27632 not shown). We also addressed Smad nuclear translocation and found that cytoskeletal alteration by ROCK inhibitors had no influence on Smad localization (Fig. 6B 6C 6D 6E) . These data add further evidence to the p38-dependence of TGF-β–induced myofibroblast transdifferentiation and may provide a mechanism for the effect of ROCK inhibitors on α-SMA expression, suggesting that TGF-β–induced contractility and p38 activation are coupled. Indeed, tensile forces have been shown to activate p38, ⁴² and TGF-β–induced α-SMA expression requires p38 activity in HTFs ¹¹ and in epithelial–mesenchymal transdifferentiation. ⁴³ It therefore appears that early cytoskeletal contraction could serve to initiate p38 signaling and α-SMA expression. ^{6 12 44}  

In a functional assay using tethered fibroblast-populated collagen gels, ROCK inhibitors decreased intrinsic contractility and prevented a TGF-β–induced increase in contractility (Figs. 7A 7B) . This effect correlated structurally with a failure of the cells to form TGF-β–induced actin stress fibers in the presence of ROCK inhibitors (Fig. 7E) , thus indicating that ROCK inhibitors prevent myofibroblast transdifferentiation on a functional level in vitro. 

Currently, various data suggest ROCK inhibitors as candidates for wound-healing modulation. Contractile forces exerted by established myofibroblasts are Rho-dependent ⁴⁵ and will conceivably be reduced by ROCK inhibitors, as has been shown in hyalocyte-populated collagen gels. ³⁴ Cytoskeletal relaxation may thus disrupt the feed-forward amplification loop of increasing cytoskeletal tension and α-SMA expression that has been described. ⁴⁴ Furthermore, ROCK inhibitors can inhibit cell migration, invasion ⁴⁶ and cytokinesis, ^{38 47} all of which have a role in wound healing and scar formation. 

Here we provide evidence of the role of Rho-mediated signaling and contractility at the outset of TGF-β–induced myofibroblast transdifferentiation in HTFs. Rho was activated within minutes after TGF-β stimulation, followed by an increase in cell tension and cytoskeletal reorganization. This process preceded the well-described myofibroblast contraction typically observed 24 to 48 hours after TGF-β stimulation. We showed that ROCK inhibitors released cell tension and inhibited TGF-β–induced cytoskeletal rearrangements, p38 activation, and α-SMA expression and its recruitment to stress fibers, as well as contraction of fibroblast-populated collagen gels. Inhibition of myosin phosphorylation by an MLCK inhibitor also prevented TGF-β–induced α-SMA expression. 

In summary, our data suggest that contractility is essential in the initiation of TGF-β–induced myofibroblast transdifferentiation. ROCK inhibitors may therefore serve as complementary pharmacological agents to modulate the conjunctival healing process after filtering glaucoma surgery. 

 Supported by the Interdisziplinäres Zentrum für Klinische Forschung (IZKF) Würzburg.

 Submitted for publication February 3, 2006; revised June 21 and July 25, 2006; accepted September 22, 2006.

 Disclosure: T. Meyer-ter-Vehn, None; S. Sieprath, None; B. Katzenberger, None; S. Gebhardt, None; F. Grehn, None; G. Schlunck, 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.

 

Figure 1.

 

TGF-β activated Rho in HTFs. The HTFs were serum-starved for 24 hours and stimulated with vehicle, TGF-β, or LPA for the times indicated. Active GTP-loaded Rho was captured in a GST-rhotekin binding domain pulldown assay and detected by Western blot analysis. TGF-β activated Rho within 10 minutes, and the activation was sustained for at least 60 minutes. LPA served as a positive control. Unprecipitated whole cell lysates were run in parallel, to assess total Rho.

Figure 2.

 

TGF-β rapidly increased cell tension. HTFs were plated on elastic silicone and observed by phase contrast (A) and differential interference contrast (B–G) microscopy in the presence of serum. Vehicle, TGF-β, and Y-27632 were sequentially added, as indicated. Cell adhesions transmit cell tension to the elastic substrate and cause it to wrinkle (as depicted in H). In the presence of serum, the cells evolved a migrational phenotype (D, arrows: main migration direction). Thirty minutes after addition of TGF-β, a global contraction became apparent as the cell covered a smaller area and the wrinkles deepened. At 60 minutes, deep wrinkles obtained a curved shape and perpendicular direction (E, F, arrows: assumed major forces). Twenty minutes after addition of Y-27632, the wrinkles were dramatically reduced. Scale bar, 20 μm.

Figure 3.

 

Effect of Y-27632 on TGF-β–induced cytoskeletal alterations. HTFs were serum starved, plated on laminin-coated (2 μg/mL) coverslips using soybean trypsin inhibitor, allowed to spread overnight, and stimulated with vehicle (A–C), TGF-β (D–F), or TGF-β and Y-27632 (G–I) for 6 hours. Vinculin and F-actin were stained with a monoclonal antibody and phalloidin-TRITC, respectively, and visualized by confocal microscopy. Composite images (C, F, I) depict vinculin (green), F-actin (red), and the DAPI-stained nucleus (blue). TGF-β–treated cells showed prominent stress fibers (E, F), and elongated focal adhesions (D, F). In contrast, Y-27632 reduced stress fibers (E, I) and focal adhesions (G, I). Scale bar, 40 μm.

Figure 4.

 

Contractility inhibitors decreased TGF-β–induced α-SMA protein expression. HTFs were serum starved overnight and stimulated with TGF-β in the absence or presence of Y-27632, HA-1077, H-1152, or ML-7, as indicated. The expression of α-SMA mRNA (A) and protein (B) were analyzed by real-time RT-PCR and Western blot analysis, respectively. After 48 hours, TGF-β had induced α-SMA protein expression, which was blocked by inhibitors (arrows, for comparison). Blots were reprobed for tubulin as a loading control.

Figure 5.

 

ROCK inhibitors prevented recruitment of α-SMA to actin stress fibers. HTFs were treated with TGF-β and H-1152 or vehicle, as indicated, and α-SMA and F-actin localization was detected by confocal immunofluorescence microscopy. Untreated cells expressed only small amounts of α-SMA which were predominantly localized to the perinuclear region (A, G). H-1152 provoked disassembly of stress fibers (H). TGF-β induced α-SMA expression and recruitment to actin stress fibers in most cells (D, F). In the presence of H-1152 less α-SMA was expressed, and it failed to be recruited into stress fibers (J, L, arrows: cells with mildly increased cytosolic α-SMA signal). Identical acquisition settings were used for all images. Scale bar, 40 μm.

Figure 6.

 

ROCK inhibitors prevented TGF-β–induced p38 activation. (A) HTFs were treated with TGF-β in the presence or absence of H-1152 for the indicated times. TGF-β induced phosphorylation of Smad2 which was unaffected by H-1152 and declined after 24 hours. The MAPK p38 was strongly phosphorylated after 24 hours of TGF-β treatment (arrows). H-1152 abolished the TGF-β–induced p38 phosphorylation. (B–E) Nuclear translocation of Smad2/3 after TGF-β stimulation (D, E) was unaltered by H-1152 (E). Arrows: nuclear signals.

Figure 7.

 

ROCK inhibitors blocked myofibroblast functions in fibroblast-populated collagen gels. HTFs were seeded in neutralized collagen solution and incubated with vehicle or TGF-β (2 ng/mL) in the presence or absence of Y-27632 (10 μM) for 48 hours, as indicated. Gels were detached, and subsequent gel contraction was digitally photodocumented (A) and measured as the reduction in gel surface area 3 hours after detachment (B). Cell morphology in contracted collagen gels (C–E). After 6 hours of contraction, gels treated with vehicle (C), TGF-β (D), or TGF-β and Y27632 (E) were fixed, and F-actin was stained with rhodamine-TRITC and viewed with a laser scanning microscope. Shown are single confocal slices depicting the most prominent cytosolic actin fiber stain. (D, arrows) stress fibers. Scale bar, 20 μm.

The authors thank Michael Sendtner and his coworkers at the Institute for Clinical Neurobiology, Würzburg, for generously sharing advice and equipment and particularly thank Michael Glinka for expert help with the live cell microscopy setup.