Activation of the Rho/Rho Kinase Signaling Pathway Is Involved in Cell Death of Corneal Endothelium

Abstract

Purpose: Rho kinase (ROCK) pathways control fundamental cell functions, making ROCK an important therapeutic target in several pathophysiologic conditions. The purpose of this study was to investigate whether inhibition of ROCK can suppress apoptosis of the corneal endothelium and to determine the role of ROCK signaling in regulating apoptosis.

Methods: The effects of inhibitors of ROCK or myosin light chain (MLC) were evaluated in cultured monkey corneal endothelial cells (MCECs) irradiated with ultraviolet (UV) (100 J/m²) to induce apoptosis. Annexin V and TUNEL staining and Western blot for apoptosis-related proteins and focal adhesion complexes were then performed. RhoA activation was further evaluated by pull-down assays. ROCK inhibitor and caspase inhibitor effects on apoptosis were also evaluated in MCECs treated with ethylene glycol tetraacetic acid (EGTA) to induce MLC phosphorylation.

Results: ROCK or MLC inhibition suppressed the caspase-3 cleavage and Annexin V and TUNEL expression typically seen during UV-mediated apoptosis of MCECs. The apoptotic stimulus activated RhoA and then induced phosphorylation of MLC via ROCK activation. EGTA-mediated phosphorylation of MLC was sufficient to induce the loss of cell contact with the substrate and subsequent apoptosis. Western blot showed that ROCK inhibition upregulated the expression of the focal adhesion complex in adhered cells, following UV stress.

Conclusions: Apoptotic stimuli activated Rho/ROCK/MLC phosphorylation in the corneal endothelium, and subsequent actomyosin contraction induced apoptosis by loss of cell adhesion. ROCK inhibition suppressed MLC phosphorylation and subsequent cell death, and it counteracted the loss of cell adhesion by activating the focal adhesion complex.

The corneal endothelium controls corneal hydration, thereby maintaining corneal transparency. The cell density of corneal endothelial cells (CECs) in healthy individuals ranges from 2000 to 3000 cells/mm², but can drop below a critical cell density (<1000 cells/mm²) for various pathologic reasons, resulting in corneal haziness. 

Fuchs endothelial corneal dystrophy (FECD) is the leading cause of corneal transplantation, with a prevalence as high as 4% of the population older than 40 years in the United States.¹ The corneal endothelium of FCED patients shows progressive damage, and several possible explanations have been proposed for the pathogenesis of FECD, including oxidative stress, mitochondrial abnormalities, microRNA, extracellular matrix, and unfolded protein response. However, histologic assessments and several basic research studies have demonstrated that apoptosis plays an important role in the cell loss associated with FECD.^2–7 The only treatment for loss of corneal transparency due to FECD is corneal transplantation,⁸ but research continues for the identification of drug therapies to regulate corneal endothelium cell loss.^9,10 

Rho kinase (ROCK) is a serine/threonine kinase that undergoes activation by interaction with Rho GTPases. The Rho/ROCK signaling pathways control a wide range of fundamental cell functions, including cell adhesion, motility, proliferation, differentiation, and apoptosis. The wide spectrum of biological events influenced by ROCK has led to the recognition of ROCK as an important therapeutic target in a variety of pathophysiologic conditions.^11–13 Indeed, inhibition of ROCK has been intensively researched for treating various diseases ranging from vascular disease, cancer, and neuronal degenerative disease to asthma and glaucoma.^11–13 One ROCK inhibitor, fasudil, has been approved for the treatment of cerebral vasospasm in 1995, and another, ripasudil, has been approved in eye drop form in Japan in 2014 for the treatment of glaucoma and ocular hypertension.^12,14,15 Recently, we have proposed ROCK inhibitors as eye drops for promoting CEC proliferation^16–18 and as adjunct drugs for cell-based therapy to enhance engraftment when treating corneal endothelial dysfunction.^19,20 

We previously have reported that the ROCK inhibitor Y-27632 suppresses the apoptosis of cultured monkey CECs (MCECs) during cell passaging, but the underlying mechanisms remain unclear.²¹ The possible involvement of an anti-apoptotic effect motivated us to investigate the feasibility of using a ROCK inhibitor as a drug to regulate pathologic conditions, especially those associated with FECD. In the execution phase of apoptosis, ROCK is pivotal in regulating morphologic events such as membrane blebbing, nuclear disintegration, and formation of apoptotic bodies in various cell types.^11,22,23 However, the importance of ROCK in the early stages of apoptosis is highly dependent on the cell type and stimulus.¹¹ The aim of the current study was to investigate whether inhibition of ROCK suppresses apoptosis of the corneal endothelium and to determine the role of ROCK signaling in the regulation of apoptosis. 

Methods

Cell Culture

Animals were housed and treated in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Six corneas from three cynomolgus monkeys (3–5 years of age; estimated equivalent human age: 5–20 years) housed at NISSEI BILIS Co., Ltd. (Osaka, Japan) were used for the experiments. The MCECs were cultivated as described previously.¹⁹ Briefly, the Descemet's membrane, including the MCECs, was stripped from the cornea and digested with Dulbecco's modified Eagle's medium (DMEM; Life Technologies Corp., Carlsbad, CA, USA) supplemented with 1 mg/mL collagenase A (Roche Applied Science, Penzberg, Germany) at 37°C for 12 hours. The isolated MCECs were recovered in culture medium, seeded on culture plates coated with FNC Coating Mix (Athena Environmental Sciences, Inc., Baltimore, MD, USA), and cultured in DMEM supplemented with 10% fetal bovine serum, 50 U/mL penicillin, 50 μg/mL streptomycin, and 2 ng/mL fibroblast growth factor 2 (Life Technologies Corp.). Monkey corneal endothelial cells at passages 4 through 8 were used for these experiments. In some experiments, cells were treated with culture medium supplemented with ethylene glycol tetraacetic acid (EGTA) (3 mM; Wako Pure Chemical Industries, Ltd., Osaka, Japan), Z-VAD-FMK (10 μM, Wako Pure Chemical Industries, Ltd.), Y-27632 (10 μM, Wako Pure Chemical Industries, Ltd.), or blebbistatin (10 μM, Wako Pure Chemical Industries, Ltd.). Cells were examined with a phase contrast microscope (DMI4000 B; Leica Microsystems, Wetzlar, Germany) or a time-lapse phase contrast microscope (BZ-9000; KEYENCE, Osaka, Japan). 

For UV stimulation, MCECs were washed gently with phosphate-buffered saline (PBS) and exposed to UV (100 J/m²) by using a UV CrossLinker CX-2000 (UVP, Upland, CA, USA), and were further cultured with fresh culture medium. For hydrogen peroxide stimulation, MCECs were grown in culture medium supplemented with hydrogen peroxide (1000 μM) and cultured for a further 24 hours. 

Ultraviolet or Hydrogen Peroxide Treatment of Corneal Specimens

Rabbit corneal specimens were placed corneal endothelial side up and exposed to UV (100 J/m²) by using the UV CrossLinker CX-2000, followed by incubation in DMEM for 24 hours. Rabbit corneal specimens were stimulated with hydrogen peroxide by culturing in DMEM supplemented with hydrogen peroxide (1000 μM) for 24 hours. 

Immunoblotting

The cultured MCECs were washed with ice-cold PBS and lysed with ice-cold RIPA buffer containing phosphatase inhibitor cocktail 2 (Sigma-Aldrich Corp., St. Louis, MO, USA) and protease inhibitor cocktail (Roche Applied Science). Following centrifugation, the supernatant containing the total proteins was fractionated by SDS-PAGE. The separated proteins were transferred to polyvinylidene difluoride (PVDF) membranes, blocked with 3% nonfat dry milk, and incubated overnight at 4°C with the following primary antibodies: caspase-3 (1:1000; Cell Signaling Technology, Inc., Danvers, MA, USA), ROCK1 (1:1000; Santa Cruz Biotechnology, Santa Cruz, CA, USA), phosphorylated myosin light chain (MLC, 1:1000; Merck Millipore, Billerica, MA, USA), phosphorylated focal adhesion kinase (FAK, 1:1000; Cell Signaling Technology), FAK (1:1000; Cell Signaling Technology), phosphorylated paxillin (1:1000; Cell Signaling Technology), poly (ADP-ribose) polymerase (PARP) (1:1000; Cell Signaling Technology), and glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (1:3000; Medical & Biological Laboratories Co., Ltd., Aichi, Japan). The blots were probed with horseradish peroxidase–conjugated secondary antibodies (1:5000; GE Healthcare, Piscataway, NJ, USA), developed with luminal for enhanced chemiluminescence using the ECL Advanced Western Blotting Detection Kit (Nacalai Tesque, Kyoto, Japan), and documented with an LAS4000S (Fuji Film, Tokyo, Japan) cooled charge-coupled device camera gel documentation system. Molecular weight markers (Bio-Rad, Hercules, CA, USA) were run alongside all samples. 

Immunohistochemistry

Cultured MCECs or corneal specimen samples were fixed for 20 minutes with 4% paraformaldehyde, and excess paraformaldehyde was removed by washing with Dulbecco's PBS. The samples were permeabilized with 0.5% Triton X-100 (Nacalai Tesque) and then incubated with 1% bovine serum albumin to block nonspecific binding. Specimens were incubated with primary antibodies against phosphorylated MLC (1:200; Merck Millipore). Alexa Fluor 488–conjugated goat anti-mouse (Life Technologies Corp.) antibodies were used as secondary antibodies at a 1:1000 dilution. Actin staining was performed by incubation with a 1:400 dilution of Alexa Fluor 546–conjugated phalloidin (Life Technologies Corp.). Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI; Vector Laboratories, Burlingame, CA, USA). 

The mitochondrial membrane potential was evaluated by 5,5′,6,6′-tetrachloro-1,1′,3,3′-tetraethylbenzimidazolylcarbocyanine iodide (JC-1) staining according to the manufacturer's protocol. Briefly, MCECs were exposed to UV (100 J/m²) and further cultured for 10 hours with culture medium supplemented with 10 μM Y-27632 or 10 μM Z-VAD-FMK. The MCECs were then incubated with MitoScreen (10 μM; Merck Millipore) for 15 minutes and fixed with 4% formaldehyde for 10 minutes. 

Cell apoptosis was evaluated by incubating the samples in medium supplemented with Annexin V (Medical & Biological Laboratories Co., Ltd.) for 15 minutes, followed by fixation with 4% paraformaldehyde for 10 minutes. Excess paraformaldehyde was removed and the samples were washed with PBS. TUNEL staining was also performed for the analysis of apoptosis according to the kit manufacturer's protocol. Briefly, MCECs were fixed with 4% formaldehyde for 10 minutes, permeabilized with 0.5% Triton X-100 solution for 5 minutes, and equilibrated by covering them with equilibration buffer for 10 minutes. The samples were then incubated with terminal deoxynucleotidyl transferase incubation buffer, which reacts with the 3′-OH ends of fragmented DNA. The slides were examined with a fluorescence microscope (DM 2500; Leica Microsystems). 

Rho Pull-Down Assay

The Rho activation assay (Merck Millipore) was performed according to the manufacturer's protocol. Briefly, MCECs were cultured until confluency, irradiated with UV (100 mJ/m²), and further incubated for 6 hours. The MCECs were washed with ice-cold Tris-buffered saline, lysed with ice-cold Mg²⁺ lysis/wash buffer (Merck Millipore) containing phosphatase inhibitor cocktail 2 (Sigma-Aldrich Corp.), and then agitated. The samples were reacted with Rho Assay Reagent (Merck Millipore) to bind GTP-Rho. The supernatant, representing total proteins, was analyzed by immunoblotting, as described as above. A primary antibody for anti-Rho, clone 55 (Merck Millipore), was used at 3:1000 dilution. 

Flow Cytometry

Cell apoptosis was evaluated by using Annexin V Assay Kits (Medical & Biological Laboratories Co., Ltd.) according to the manufacturer's protocol. Briefly, cells were incubated with DMEM supplemented with Annexin V for 15 minutes and then harvested by digestion with Accumax (Innovative Cell Technologies, San Diego, CA, USA). Recovered cells were then analyzed by flow cytometry using a CellQuest Pro software (BD Biosciences, Franklin Lakes, NJ, USA). 

Statistical Analysis

The statistical significance (P value) of differences between mean values of the two-sample comparison was determined with the Student's t-test. The comparison of multiple sample sets was analyzed by using Dunnett's multiple-comparison test. The data represent the mean ± SEM. 

Results

Suppression of Plasma Membrane Blebbing by ROCK Inhibition

Phase contrast images showed that UV stimulation induced a loss of cell adhesion of MCECs to the culture plate and that the released cells showed plasma membrane blebbing, which is a phenotypic feature associated with the execution phase of apoptosis. However, the selective ROCK inhibitor Y-27632 suppressed this loss of cell adhesion as well as membrane blebbing of MCECs after stimulation with UV (Fig. 1A). Time-lapse phase contrast microscopy showed that the numbers of MCECs exhibiting membrane blebbing increased in a time-dependent manner after UV irradiation, but membrane blebbing was almost totally suppressed by Y-27632 treatment (Fig. 1B). Two isoforms, ROCK1 and ROCK2, were isolated as RhoA-interacting proteins. Membrane blebbing is induced by ROCK1 activation, which is caused by ROCK1 cleavage triggered by caspase-3 activation in other cell types. Western blot showed that slight cleavage of caspase-3 occurred after 3 hours and this cleavage became more apparent after 6 hours of UV stimulation, with a coincident cleavage of ROCK1 in the MCECs (Fig. 1C). Time-lapse phase contrast imaging showed that the blebbing became evident after 6 hours of UV stimulation and paralleled the cleavage of caspase-3 and ROCK1. 

Figure 1

 

Effect of ROCK inhibition on membrane blebbing in MCECs. (A) Cultured MCECs were washed gently with PBS and irradiated with UV (100 J/m²). Phase contrast images showed that UV stimulation induced a loss of cell adhesion from the culture plate and the surrounding cells, associated with plasma membrane blebbing. However, the selective ROCK inhibitor Y-27632 suppressed this loss of cell adhesion, as well as the membrane blebbing. Representative phase contrast images of MCECs stimulated by UV after 6 hours are shown. Scale bar: 50 μm. (B) Blebbing cells were evaluated by time-lapse phase contrast microscopy. *P < 0.01. (C) Western blot was performed after UV stimulation to evaluate cleavage of caspase-3 and ROCK1. A slight amount of cleavage of caspase-3 was observed after 3 hours, and cleavage became more apparent after 6 hours of UV stimulation, accompanied by a coincident cleavage of ROCK1 in the MCECs.

Effect of Inhibition of ROCK/MLC Pathway on Apoptosis of MCECs

We then evaluated the effect of ROCK signaling inhibition on UV-mediated apoptosis in MCECs. Phase contrast images showed that UV-induced cell detachment was suppressed by the pan-caspase inhibitor Z-VAD-FMK after 12 hours of UV irradiation (Fig. 2A). Western blot showed the presence of cleavage products of caspase-3 (19-kDa partially cleaved and 17-kDa cleaved products) in UV-stimulated cells. PARP helps cells to maintain viability, so cleavage of PARP, which is caused by cleaved caspase-3, is used as a marker of cells undergoing apoptosis. The UV treatment of the MCECs induced an identical PARP cleavage to that seen for caspase-3. By contrast, Z-VAD-FMK suppressed the formation of the 17-kDa cleavage products of caspase-3 and the 89-kDa cleavage products of PARP (Fig. 2B), indicating that the damage to MCECs caused by UV is mainly due to apoptosis associated with caspase-3 activation. 

Figure 2

 

Effect of inhibition of ROCK/MLC pathway on apoptosis of MCECs. (A) Cultured MCECs were washed gently with PBS and exposed to UV (100 J/m²), followed by further incubation for 12 hours in culture medium supplemented with the pan-caspase inhibitor Z-VAD-FMK. Representative phase contrast images show that UV-induced cell detachment was suppressed by caspase inhibition. Scale bar: 100 μm. (B) Activation of caspase-3 and PARP was evaluated by Western blot after 12 hours of UV stimulation. Ultraviolet exposure induced partially cleaved 19-kDa and fully cleaved 17-kDa products of caspase-3. The UV treatment also induced PARP cleavage in MCECs. By contrast, Z-VAD-FMK suppressed the appearance of the 17-kDa cleavage product of caspase-3 and the 89-kDa cleavage product of PARP, showing that damage to the MCECs caused by UV is mainly due to apoptosis. (C) Cleavage of caspase-3 was evaluated by Western blot after UV irradiation of MCECs. Y-27632 and blebbistatin (an inhibitor of MLC) treatment suppressed cleavage of caspase-3 in the UV-treated cells. (D, E) Monkey corneal endothelial cells were exposed to UV (100 J/m²) and then cultured in Dulbecco's modified Eagle's medium supplemented with Y-27632/blebbistatin for 24 hours. The presence of apoptotic cells was evaluated by staining the MCECs with Annexin V. Evaluation of the numbers of Annexin V–positive apoptotic cells produced in response to UV irradiation revealed a significant suppression by Y-27632 or blebbistatin treatment. Values are the averages of four independent images. Experiments were performed in triplicate. Nuclei were stained with DAPI. Scale bar: 50 μm. *P < 0.01. (F, G) Apoptotic cells were evaluated by TUNEL staining. The numbers of TUNEL-positive apoptotic cells were evaluated. Values are the averages of four independent images. Experiments were performed in triplicate. Nuclei were stained with DAPI. Scale bar: 50 μm. *P < 0.01. (H) Monkey corneal endothelial cells were stimulated with hydrogen peroxide by culturing with Dulbecco's modified Eagle's medium supplemented with hydrogen peroxide (1000 μM) for 24 hours. Phosphorylation of MLC was evaluated by immunostaining. Actin was stained by phalloidin and nuclei were stained with DAPI. Scale bar: 50 μm. (I, J) After stimulation by hydrogen peroxide (1000 μM) for 24 hours, MCECs were evaluated by phase contrast microscopy. Annexin V–positive apoptotic cells were evaluated by flow cytometry. Scale bar: 50 μm. *P < 0.01.

We also evaluated the effect of inhibition of MLC activity on apoptosis of MCECs, as MLC is downstream of ROCK. Blebbistatin (an inhibitor of MLC) suppressed the cleavage of caspase-3 in a similar fashion to that seen with the ROCK inhibitor (Fig. 2C). Annexin V staining showed that early apoptosis induced by UV was significantly suppressed by Y-27632 or blebbistatin after 24 hours of UV irradiation (Figs. 2D, 2E). TUNEL staining also showed that late apoptosis was significantly decreased in the MCECs treated with Y-27632 or blebbistatin (Figs. 2F, 2G). These data indicated that inhibition of the ROCK/MLC pathway counteracts UV-mediated apoptosis of MCECs. Apoptosis was induced by UV or by hydrogen peroxide, and effect of Y-27632 on corneal endothelial apoptosis was evaluated. Immunostaining showed that hydrogen peroxide induced phosphorylation of MLC (which is downstream from ROCK1), which is associated with cell contraction and membrane blebbing (Fig. 2H). Phase contrast images showed that hydrogen peroxide caused cell detachment after 24 hours, associated with membrane shrinkage and blebbing, but Y-27632 suppressed this cell detachment (Fig. 2I). Flow cytometry demonstrated that hydrogen peroxide induced Annexin V–positive apoptotic cells at 7.8%, but Y-27632 treatment significantly suppressed the apoptotic cell percentage to 5.0% (P < 0.01) (Fig. 2J). 

Effect of Inhibition of ROCK on Apoptosis of Ex Vivo Corneal Endothelium

We next conducted ex vivo experiments with rabbit corneal specimens. Annexin V–positive apoptotic CECs were observed at a level of 10.3% in rabbit corneas after 24 hours of UV irradiation, but treatment with Y-27632 suppressed this percentage to 2.9% (Fig. 3A). A similar apoptotic response was seen with hydrogen peroxide treatment, where the appearance of Annexin V–positive cells was significantly suppressed by Y-27632 treatment (10.8% [control] versus 3.0% [Y-27632]) after 24 hours (Fig. 3B). In addition, we confirmed that Y-27632 caused a similar suppression of UV-induced corneal endothelial apoptosis in human corneal specimens as was observed in the rabbit specimens (Supplementary Figs. S1A, S1B). 

Figure 3

 

Effect of inhibition of ROCK on apoptosis of ex vivo rabbit corneal endothelium. (A, B) Rabbit corneal specimens were placed corneal endothelial side up and exposed to UV (100 J/m²), followed by incubation in Dulbecco's modified Eagle's medium for 24 hours. Apoptotic cells were detected by staining the corneal specimens with Annexin V and counting the numbers of Annexin V–positive apoptotic cells. Values are the averages of data from four independent corneas. Experiments were performed at least in triplicate. Scale bar: 50 μm. *P < 0.01. (C, D) Hydrogen peroxide stimulation of rabbit corneal specimens incubated in Dulbecco's modified Eagle's medium supplemented with hydrogen peroxide (1000 μM) for 24 hours. Specimens were stained with Annexin V and the numbers of Annexin V–positive cells were evaluated. Values are averages of data from four independent corneas. Experiments were performed at least in triplicate. Scale bar: 50 μm. *P < 0.01.

Phosphorylation of MLC by RhoA Activation During the Initiation Phase of Apoptosis in MCECs

RhoA is recognized as a central modulator for ROCK activation; therefore, we examined the activation of Rho in UV-stimulated MCECs. Exposure of MCECs to UV irradiation resulted in substantial recognition of GTP-bound RhoA in pull-down assays, suggesting that cellular stress induced the activation of RhoA (Fig. 4A). Consistent with RhoA activation, immunofluorescent images showed UV-induced phosphorylation of MLC after 3 hours of UV irradiation. However, Y-27632 counteracted the MLC phosphorylation in a similar fashion to blebbistatin, indicating that MLC phosphorylation is a result of ROCK activation (Fig. 4B). Monkey corneal endothelial cells were phosphorylated beginning approximately 30 minutes after UV irradiation, and 33.3% of the MCECs were phosphorylated at 6 hours (Fig. 4C). This MLC phosphorylation was observed earlier than ROCK1 cleavage and the subsequent membrane blebbing is shown in Figure 1, suggesting that ROCK was activated by RhoA activation during the early phase of apoptosis. ROCK1, in turn, was activated by ROCK1 cleavage by casplase-3 activation during the execution phase of apoptosis. Immunostaining and Western blot performed 3 hours after UV irradiation demonstrated that ROCK inhibitor treatment suppressed the MLC phosphorylation during the initiation phase (Figs. 4B, 4D). 

Figure 4

 

RhoA activation and subsequent phosphorylation of MLC during the initiation phase of apoptosis. (A) Monkey corneal endothelial cells were cultured until confluency, irradiated with UV (100 mJ/m²), and incubated. Activation of Rho was then evaluated by a pull-down assay. (B, C) Myosin light chain phosphorylation was assessed by immunostaining. Ultraviolet induced the phosphorylation of MLC after 3 hours of UV irradiation, while Y-27632 and blebbistatin counteracted this MLC phosphorylation. The percentages of MCECs showing MLC phosphorylation were evaluated. Nuclei were stained with DAPI. Scale bar: 50 μm. (D) Monkey corneal endothelial cells were stimulated by exposure to UV and then incubated with or without Y-27632/blebbistatin. Phosphorylation of MLC was evaluated by Western blot 3 hours after UV treatment. Experiments were performed in triplicate.

Effect of MLC Phosphorylation on Apoptosis of MCECs

We recently have reported that cell dissociation results in phosphorylation of MLC in MCECs.²⁰ In the present study, we used EGTA to induce MLC phosphorylation to assess its effects on apoptosis. As found previously, EGTA treatment resulted in MLC phosphorylation and subsequent actin contraction, followed by cell detachment from the substrate and surrounding cells (Figs. 5A, 5B). By contrast, both Y-27632 and blebbistatin counteracted this induction of MLC phosphorylation and actin contraction, and the cells maintained a monolayer confluent sheet rather than assuming the round morphology associated with actomyosin contraction. Phase contrast images showed that Y-27632 and blebbistatin suppressed cell detachment (Figs. 5A, 5B). However, MCECs exhibited MLC phosphorylation associated with actin contraction even when treated with Z-VAD-FMK, showing that cell detachment was not a consequence of apoptosis, whereas actin contraction was due to MLC phosphorylation. Western blot also showed that the MLC phosphorylation due to EGTA treatment induced the cleavage of caspase-3 and PARP, whereas Y-27632 and blebbistatin treatment suppressed these cleavages (Fig. 5C). 

Figure 5

 

RhoA activation and subsequent phosphorylation of MLC during the initiation phase of apoptosis. (A) Confluent MCECs were incubated in culture medium supplemented with EGTA (3 mM) to induce MLC phosphorylation, which was assessed by immunostaining. Representative images after 6 hours of incubation with EGTA supplemented with Y-27632, blebbistatin, or Z-VAD-FMK. Scale bar: 50 μm. (B) Representative phase contrast images of MCECs after 6 hours of incubation with EGTA. Scale bar: 100 μm. (C) Western blot to evaluate cleavage of caspase-3 and PARP performed after 24 hours of EGTA treatment. Ethylene glycol tetraacetic acid mediated MLC phosphorylation, which induced cleavage of caspase-3 and PARP, but Y-27632 and blebbistatin suppressed these cleavages. Experiments were performed in triplicate. (D) Ultraviolet-mediated apoptosis was induced by UV irradiation, and activation of focal adhesion complexes were evaluated by Western blot. Expression of focal adhesion complexes in adhered cells under stress of UV irradiation was promoted by Y-27632. Experiments were performed in triplicate.

We also evaluated the effect of ROCK inhibition on the focal adhesion complex, as it may also be involved in modulation of cell detachment. Focal adhesion kinase and paxillin are important components of focal adhesion and transmit downstream signaling of integrins. Western blot showed that Y-27632 treatment promoted the phosphorylation of FAK and paxillin in adhered cells exposed to UV irradiation stress (Fig. 5D). Taken together, the data support the possibility that perception of apoptotic stimuli by the corneal endothelium induced the activation of the RhoA/ROCK/MLC pathway, promoted actomyosin contraction, and finally caused apoptosis or anoikis. Treatment with a ROCK inhibitor suppressed apoptosis by counteracting the RhoA/ROCK/MLC signaling as well as by activation of the focal adhesion complex (Fig. 6). 

Figure 6

 

The molecular pathway for a role for ROCK signaling in apoptosis and mechanisms to explain how ROCK inhibitor suppresses apoptosis of the corneal endothelium. Apoptotic stimuli perceived by the corneal endothelium induce the activation of the RhoA/ROCK/MLC pathway. Actomyosin contraction, induced by MLC phosphorylation, causes a loss of adhesion, followed by apoptosis or anoikis. ROCK inhibitor treatment suppresses apoptosis by counteracting RhoA/ROCK/MLC signaling, as well as by activating expression of the focal adhesion complex.

Discussion

The execution phase of apoptosis is characterized by morphologic events, and ROCK activation is recognized as its major regulator.^11,22,23 Loss of contact with the extracellular matrix and surrounding cells leads to a transient membrane blebbing by apoptotic cells, which is regulated by actomyosin contraction via MLC phosphorylation. In 2001, caspase-3–mediated cleavage and activation of ROCK1 was demonstrated as pivotal in inducing MLC phosphorylation and subsequent membrane blebbing in NIH3T3 cells treated with TNF-α and in Jurkat cells treated with agonistic anti-Fas antibodies.^22,23 Subsequent research has confirmed the same mechanism in various cell types and with different kinds of apoptotic stimuli.^24–27 

Our current data agree with previous findings with other cell types, as the corneal endothelium shared the same system for regulating membrane blebbing involving caspase-3 and ROCK1 cleavage during the execution phase. However, the contribution of ROCK signaling to the initiation of the apoptotic phase varies widely, depending on the cell type and apoptotic stimulus. The first reports showing a role for caspase-3–mediated cleavage of ROCK1 in membrane blebbing have indicated that inhibition of ROCK does not suppress caspase-3 activation and subsequent apoptosis in TNF-α–treated NIH3T3 and anti-Fas antibody–treated Jurkat cells.^22,23 Likewise, in lung epithelial cells, ROCK activation is required for membrane blebbing but is not required for other apoptosis-related events.²⁴ These reports suggest that activation of ROCK signaling is indispensable for the morphologic events occurring during the execution phase but not for earlier apoptosis processes. 

However, several lines of evidence indicate that ROCK activation contributes to the initiation phase as well as the execution phase. For example, MLC phosphorylation plays a central role in the TNF-α–mediated apoptosis of vascular endothelial cells.²⁸ ROCK-dependent rearrangement of actin cytoskeleton is also a critical event in early apoptosis, possibly due to assembly of the death-inducing signaling complex, and inhibition of ROCK attenuates TNF-α–mediated apoptosis.²⁹ Similarly, ROCK signaling is activated by inhibition of ERK-MAPK signaling in vascular endothelial cells and causes cell death, while inhibition of ROCK rescues the cells from the effects of ERK-MAPK inhibition.³⁰ Apoptosis of erythroblast cells mediated by phorbol-12-myristate-13-acetate is also triggered by upregulation of the Rho/ROCK pathway and is mediated by caspase-8 and caspase-10 associated with myosin contraction.³¹ Human embryonic stem cells show poor survival after cell dissociation, but this is attenuated by a ROCK inhibitor.³² Other research has demonstrated that dissociation-induced apoptosis is caused by a ROCK-dependent hyperactivation of actomyosin, and that Rho-GEF (guanine nucleotide exchange factor) is an indispensable regulator of Rho/ROCK/actomyosin activation.³³ 

Our data agree with these previous reports, as we demonstrated that activation of the Rho/ROCK pathway contributes to the initiation phase of apoptosis in the corneal endothelium. We showed that UV stimulation activated Rho and was followed by induction of actin cytoskeletal contraction by MLC phosphorylation via ROCK activation. We also showed that EGTA-mediated MLC phosphorylation was sufficient to cause apoptosis associated with caspase-3 activation, and that MLC phosphorylation was sufficient for inducing the loss of cell adhesion even in the absence of apoptosis. These data suggest that the loss of contact with the substrate or with the surrounding cells due to MLC phosphorylation regulates apoptosis of the corneal endothelium. 

The Rho subfamily members function as molecular switches that cycle between GDP-bound inactive and GTP-bound active forms. This transition is controlled by the cooperation of positive regulators (GEFs) and negative regulators (GTPase activating proteins [GAPs]).³⁴ Further research is required on the types of proapoptotic signals that activate Rho and the involvement of GEF and GAPs in the activation. Another important question awaiting further study is whether the activation of Rho/ROCK/MLC contributes to cell loss in the diseased condition of the corneal endothelium, especially the FECD condition, in order to determine the usefulness of ROCK signaling as a target for treating disease in the clinical setting. 

In summary, we showed that cellular stresses to the corneal endothelium activated Rho/ROCK/MLC phosphorylation. Subsequently, phosphorylated MLC induced the loss of cell adhesion, which then promoted cell death due to apoptosis or anoikis. We demonstrated that ROCK inhibition suppressed MLC phosphorylation and subsequent cell death and counteracted the loss of cell adhesion by activating the focal adhesion complex. These findings suggest that regulation of apoptosis in a pathologic status such as FECD might be achieved by targeting the Rho/ROCK signaling pathway. 

Acknowledgments

The authors thank Yuji Sakamoto for technical assistance for handling the monkey corneas. 

Supported by the Program for the Strategic Research Foundation at Private Universities from MEXT (NK and NO). 

Disclosure: N. Okumura, None; K. Fujii, None; T. Kagami, None; N. Makiko, None; M. Kitahara, None; S. Kinoshita, P; N. Koizumi, P