Fasudil hydrochloride, a potent ROCK inhibitor, inhibits corneal neovascularization after alkali burns in mice.

PURPOSE: To investigate the effects and mechanisms of fasudil hydrochloride (fasudil) on and in alkali burn-induced corneal neovascularization (CNV) in mice.
METHODS: To observe the effect of fasudil, mice with alkali-burned corneas were treated with either fasudil eye drops or phosphate-buffered saline (PBS) four times per day for 14 consecutive days. After injury, CNV and corneal epithelial defects were measured. The production of reactive oxygen species (ROS) and heme oxygenase-1(HO-1) was measured. The infiltration of polymorphonuclear neutrophils (PMNs) and the mRNA expressions of CNV-related genes were analyzed on day 14.
RESULTS: The incidence of CNV was significantly lower after treatment with 100 μM and 300 μM fasudil than with PBS, especially with 100 μM fasudil. Meanwhile, the incidences of corneal epithelial defects was lower (n=15, all p<0.01). After treatment with 100 μM fasudil, the intensity of DHE fluorescence was reduced in the corneal epithelium and stroma than with PBS treatment (n=5, all p<0.01), and the number of filtrated PMNs decreased. There were significant differences between the expressions of VEGF, TNF-a, MMP-8, and MMP-9 in the 100 μM fasudil group and the PBS group (n=8, all p<0.05). The production of HO-1 protein in the 100 μM fasudil group was 1.52±0.34 times more than in the PBS group (n=5 sample, p<0.05).
CONCLUSIONS: 100 μM fasudil eye drops administered four times daily can significantly inhibit alkali burn-induced CNV and promote the healing of corneal epithelial defects in mice. These effects are attributed to a decrease in inflammatory cell infiltration, reduction of ROS, and upregulation of HO-1 protein after fasudil treatment.

Introduction

Commonly associated with inflammatory, infectious, and traumatic disorders of the ocular surface, corneal neovascularization (CNV) is a severe sight-threatening condition. When the effect of angiogenic factors overcome that of antiangiogenic factors in corneal burns, corneal neovascularization will be brought out [1-3]. According to current understanding, inflammatory cytokines and reactive oxygen species (ROS) are two of the major angiogenic factors in the development of CNV after injury.

The current agents for inhibiting CNV includes anti-angiogenic factors [4,5], anti-inflammatory agents [6,7], agents for anti-remodeling of the extracellular matrix [8,9], and agents for anti-oxidative stress [10]. As a potential regulator of cellular ROS metabolism, the small GTPase RhoA and its downstream effector ROCK in the RhoA/ROCK signal pathway has recently been shown to play a critical role in angiogenesis [11,12]. ROCK inhibitors could protect endothelial cells from inflammatory damage by suppressing nuclear factor kappa B signaling [13,14] and reducing ROS production [15]. As a potent inhibitor of Rho-kinase, fasudil has an inhibitory effect similar to ATP [16]. In this study, fasudil was demonstrated to inhibit alkali burn-induced CNV not only by decreasing inflammation but also by reducing ROS via the RhoA/ROCK pathway and by increasing the heme oxygenase-1 (HO-1) protein, a protective factor against ROS.

Methods

Animals

Female BALB/c mice aged 4 to 6 weeks and weighting between 16 and 20 g were purchased from the Guangdong Provincial Center for Animal Research in Guangzhou, China. The right eye of each mouse was selected for experimentation. All experiments on animals were conducted in accordance with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. The research protocol was approved by the Animal Care Committee of the Zhongshan Ophthalmic Center at Sun Yat-sen University in China.

Alkali burn- induced CNV

CNV was induced by alkali burns using a method outlined in previous reports [17]. In brief, after the mice were sedated with an intraperitoneal injection of general anesthesia consisting of 4.3% chloral hydrate [10 ml/kg] and a topical anesthesia consisting of a drop of 0.5% proparacaine hydrochloride (Alcaine eye drops, Alcon Inc., Fort Worth, TX), a 2 mm diameter filter paper soaked with 2 μl of 0.1 M NaOH solution was placed on the central cornea for 40 s, followed by immediate rinsing with 30 ml of 0.9% saline solution for 10 s. The entire corneal limbus and epithelium were then scraped off with a surgical blade under a microscope. Tobramycin ophthalmic ointment (Tobrex, Alcon Inc.) was administered after the operation.

Treatment with fasudil hydrochloride eye drops

The fasudil hydrochloride (Asahi Kasei Inc., Tokyo, Japan) was diluted in phosphate-buffered saline (PBS) to make different concentrations of fasudil eye drops. To observe the antiangiogenic effects of fasudil, 75 mice with alkali burns were randomly assigned to either the experimental groups to be treated topically with 30, 100, 300, and 1000 μM fasudil eye drops or to the control groups to be treated with PBS eye drops four times daily for 14 consecutive days (n=15 in each group). All the mice were killed on day 14 for immunohistopathological examination and reverse quantitative real-time polymerase chain reaction (PCR) analysis.

To detect the production of HO-1 in the murine corneas, another 50 mice with alkali burns were randomly treated with 100 μM fasudil eye drops or PBS eye drops four times daily for 4 consecutive days (n=25 in each group). To detect the production of ROS, 15 mice with alkali burns were randomly assigned to be treated with 100 μM fasudil eye drops, PBS eye drops, or nothing every 2 h for six consecutive hours (n=5 in each group), while another five normal mice without burns were assigned to a control group. The murine corneas were then procured for the analysis of ROS production.

Quantification of CNV and measurement of corneal epithelial defects

On days 4, 7, 10, and 14 after receiving alkali burns, the murine corneas were examined and photographed with a digital camera (Cannon, Tokyo, Japan) attached to a slit-lamp microscope (SL-120, Zeiss Inc., Jena, Germany). After the cycloplegia with a tropicamide eye drop (Wujing Pharmaceutical Company, Wuhan, China), CNV was observed under a white light. Corneal epithelial defects were then revealed using 0.5% fluorescein staining and were observed under a cobalt blue light. Three consecutive photos with satisfactory full-face imaging were used for image analysis. Image J software (ver. 1.62) was downloaded from the US National Institutes of Health (NIH) website and used for quantitative analysis, as documented in a previous report [18]. The consecutive photos of each mouse were used to measure the areas of CNV, epithelial defects, and the entire cornea, and then their percentages were calculated.

Measurement of ROS

Six hours after the alkali burn, the murine eyeballs were enucleated and immediately frozen in optimum cutting temperature (OCT) compound (Sakura Finetek, Torrance, CA). The cryopreserved blocks were cut until the first slide with corneal tissue was observed. The next five 30th slides of 10 μm thickness were then collected for immunostaining. when the first slide was observed, then the last section of every 30 consecutive cuts was procured to get five sections. Each section was 10 μm thickness.Unfixed cryosections were incubated with 5 μM dihydroethidium (DHE, ROS Fluorescent Probe, Eugene, OR) for 15 min at 37 °C, as previously reported [19]. The negative control was created by incubating sections with PBS instead of DHE. Sections were examined using a fluorescence microscope at 400× magnification (Olympus IX51, Tokyo, Japan) and photographed with an exposure time of 695 ms. The intensity of fluorescein staining was measured in the corneal epithelium and stroma using Image J 1.62 software.

Immunohistochemistry

Fourteen days after the alkali burns, seven mice in each of the 100 μM and PBS groups were randomly killed and their eyeballs were enucleated for immunohistochemistry. The eyeballs were fixed in a 10% neutral buffered formaldehyde solution and embedded in paraffin. The paraffin blocks containing the samples were cut until the first slide with corneal tissue was observed. Then the next three 20th slides of 5 μm thickness were collected for immunostaining, and another three 25th slides were collected for hematoxylin-eosin (H-E) staining. When the first slide was observed, then the 20th and 25th sections of every 30 consecutive cuts was procured to get three sections each for immunostaining and HE staining. For the immunostaining of polymorphonuclear neutrophils (PMNs), sections were deparaffinized and boiled in antigen retrieval solution (Dako, Glostrup, Denmark) for 15 min. Nonspecific staining was blocked with 1% bovine serum albumin (BSA; Sigma-Aldrich, St. Louis, MO) for 1 h at room temperature. Sections were then incubated with the primary antibody (monoclonal rat antibody against mouse neutrophil marker NIMP-R14; sc59338, Santa Cruz Biotechnology, Santa Cruz, CA) diluted with 1% BSA (1:300) overnight at 4 °C. After three washes with PBS for 15 min, sections were incubated with HRP-conjugated secondary anti-rat IgG antibody (ZSGB-BIO Institute of Biotechnology, Beijing, China) for 1 h at room temperature and then washed again with PBS. The negative control was created by incubating sections with PBS instead of the primary antibody. The sections were then counterstained with Hoechst 33342 for 2 min. The sections were then washed three times in PBS for 3 min before being mounted on glass slides. The corneal sections were procured for histopathological study, just as in the procedures for immunostaining, and the slides were deparaffinized and regularly stained with H-E. Corneal morphology and the infiltrated PMNs in the corneal stroma were visualized and photographed with a fluorescence microscope (Olympus IX51). The number of infiltrated inflammatory cells in the cornea, including PMNs, was counted in five randomly selected fields (at 400× magnification) of a slide.

Analysis of mRNA expression

Real-time quantitative reverse transcription PCR (real-time qRT-PCR) was performed to detect the mRNA expression levels of the VEGF, VEGFR1, VEGFR2, TNF-α, IL-1β, MMP-2, MMP-8, and MMP-9 genes in the murine corneas. On day 14 after the burns, 8 mice from each of the 100μM group and the PBS group were euthanized by cervical dislocation, and their right corneas were procured. The total RNAs of each cornea were immediately isolated using an RNeasy Micro Kit (Qiagen, Valencia, CA). After quantification of the RNA concentration, the total RNAs were treated with DNase I (Sigma-Aldrich) to remove any contaminated genomic DNA. A total of 0.5 μg RNA was reverse transcribed into cDNA in a 20 μl volume reaction system using a Maxima First Strand cDNA Synthesis Kit (Fermentas International Inc., Burlington, ON, Canada). Samples of synthesized cDNA were divided into aliquots and stored at -80 °C.

Real-time qRT-PCR was performed and the results analyzed using an ABI PRISM 7000 sequence detection system (Applied Biosystems Inc., Foster City, CA). PCR was performed in a 20 μl volume reaction system containing 10 μl 2SYBR Green Reaction Mix (Invitrogen, Carlsbad, CA), 0.4 mmol/l paired primers, and 1 μl cDNA. Each sample was simultaneously run in triplicate. The sequences of the PCR primer pairs are listed in Table 1. Thermal cycling consisted of denaturation for 3 min at 95 °C, followed by 40 cycles of 15 s at 95 °C and 30 s at 60 °C. The PCR amplification efficiency of the primer sets was determined to be 100% before qPCR. A comparative Ct (ΔΔCt) method was used to compare the mRNA expression levels of the genes of interest. The glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene was chosen to be an internal control gene.

Table 1

Primer sets for real-time PCR.

Gene	Forward (5′-3′)	Reverse (5′-3′)
GAPDH	GTTGTCTCCTGCGACTTCA	TGGTCCAGGGTTTCTTACT
VEGF	ACTATTCAGCGGACTCACC	AACCAACCTCCTCAAACC
VEGFR1	ACCCTGGTAAAGCAACTAA	GCACGGAGGTGTTGAAAG
VEGFR2	TGTGAACGCTTGCCTTAT	CAACATCTTGACGGCTACT
TNF-α	TCAGAATGAGGCTGGATAA	AAGAGGAGGCAACAAGGT
IL-1β	CTCCATGAGCTTTGTACAAGG	TGCTGATGTACCAGTTGGGG
MMP-2	CCCCGATGCTGATACTGA	CACTGTCCGCCAAATAAA
MMP-8	GATTATGGAAATGCCTCG	CTTCAGCCCTTGACAGC
MMP-9	CAGCCAACTATGACCAGGAT	CTGCCACCAGGAACAGG

Western blot

The protein levels of HO-1 in the murine corneas were measured by western blots. Briefly, the corneas procured from the mice in the 100 μM group and the PBS group were homogenized in a 1.5 ml microtube with a lysis buffer (Beyotime Institute of Biotechnology, Shanghai, China) on day 4 after the alkali injury. Twenty-five corneas in each group were randomly divided and pooled into five samples, each of which had five corneas. Supernatants from each sample homogenate were collected after centrifugation at 24949 ×g for 30 min at 4 °C. Protein concentrations were quantified using a BCA Protein Assay Kit (Beyotime Institute of Biotechnology). Next, 20 μg of protein from each sample was mixed with a 2× sample buffer and heated to 100 °C for 5 min. After electrophoresis, the proteins were electrotransferred onto 0.2 mm PVDF membranes (Bio-Rad, Hercules, CA). The membranes were blocked with 5% skimmed milk and incubated overnight at 4 °C with primary antibodies (1:500, sc-10789, Santa Cruz Biotechnology). The staining specificity was assessed by the omission of the primary antibody. After three washes with tris-buffered saline (TBS), the membranes were incubated with a horseradish peroxidase-conjugated anti-rabbit IgG secondary antibody solution (1:1000, sc-2004, Santa Cruz Biotechnology) and then diluted in TBS with 5% skim milk for 2 h at room temperature. The membranes were then washed in TBS and placed in the chemiluminescent working solution (ECL detection reagents, Thermo Fisher Scientific, Waltham, MA) for 5 min at room temperature. The membrane was imaged with a G:BOX BioImaging system (Syngene, Cambridge, England) at the same time, and the pictures were analyzed with Image J 1.62 software to measure the densitometric intensities of the respective bands. The experiment was repeated once.

Statistics

All data was expressed as mean ± standard deviation. The percentages of CNV and corneal epithelium defects were compared using one-way ANOVA between the four experimental groups and the control. A least-significant difference (LSD) analysis was performed to compare each experimental group and the control group. The expression levels of mRNAs, ROS production, and HO-1 proteins, as well as the number of infiltrated inflammatory cells, including PMNs, were compared using a Student t test for both the 100 μM fasudil group and the PBS group. A p value of less than 0.05 was considered to be statistically significant.

Results

Effects of fasudil on alkali burn-induced CNV

The area of CNV was found to increase with time in both the experiment groups and the control group after alkali burns (Figure 1A). The percentages of CNV areas in the 100 μM fasudil treated group were significantly lower than those in the PBS treated group at every check point time (all p<0.05, Figure 1B). The percentages of CNV areas in the 300 μM fasudil treated group were lower than in the PBS treated group on days 10 and 14 (all p<0.05, Figure 1B). The areas of CNV in the 100 μM fasudil treated group were smaller than in the 300 μM fasudil treated group. There were no statistical differences in the percentages of CNV areas between either the 30 μM or the 1000 μM fasudil treated groups and the control group at any checkpoint after injury.

Figure 1

Inhibitory effect of fasudil eye drops on corneal neovascularization (CNV). A: Representative images of the murine corneas after alkali burns with and without fasudil eye drops. B: The percentages of CNV in the fasudil and phosphate-buffered saline (PBS) groups at every checkpoint. The percentages of CNV in the 30, 100, 300, and 1000 μM fasudil groups were 40.29%±6.30%, 32.54%±3.76%, 38.49%±6.27%, and 41.67%±8.78% on day 4; 48.61%±9.65%, 40.40%±8.15%, 46.36%±10.81%, and 49.52%±11.65% on day 7; 78.28%±10.21%, 54.17%±10.71%, 63.23%±14.32%, and 78.00%±11.57% on day 10; 95.89%±3.23%, 67.45%±15.45%, 82.38%±12.43%, and 93.16%±3.70% on day 14. The percentages of CNV in the PBS group were 39.04%±8.07%, 52.24%±10.58%, 79.34%±12.96%, 96.07%±3.32% on day 4, 7, 10, and 14, respectively (n=15/group, * indicates p<0.05, ** indicates p<0.01).

Fasudil promotes corneal epithelial healing

The percentages of corneal epithelial defect area to the entire cornea in the fasudil treated groups were lower than those in the PBS treated group on days 4 and 7 after the alkali burns (all p<0.01), with the exception of the 30 μM fasudil treated group on day 4. By day 10, the murine corneas had re-epithelialized in all groups (Figure 2).

Figure 2

Fasudil promoted the healing of corneal epithelial defects after alkali burns. A: Representative images of the mice corneas with fluorescein staining after treatment with phosphate-buffered saline (PBS) or fasudil eye drops after alkali burns. B: The percentages of corneal epithelial defects were 22.21%±8.23%, 6.59%±3.27%, 11.50%±6.48%, and 13.90%±7.34% on day 4 and 4.20%±6.4%, 0%, 0%, and 0% on day 7 in the 30 μM, 100 μM, 300 μM, and 1000 μM fasudil groups. The corneal surfaces were re-epithelialized in all mice from day 10. So we don’t need to compare the difference after day 10.The percentages of corneal epithelial defects were 21.15%±9.21% and 8.60%±4.60% in the PBS group on days 4 and 7, respectively (n=15/group, * indicates p<0.05, ** indicates p<0.01).

The production of ROS in the cornea was enhanced after alkali burns and was then inhibited after treatment with 100 μM fasudil

To evaluate the effect of fasudil on ROS formation after the alkali burns, ROS production was examined using a DHE assay, in which DHE reacts with to show fluorescent imaging. The results show that the production of ROS in the corneal epithelium and stroma after an alkali injury without treatment was 2.29±0.21 and 2.0±0.20 times that of ROS production in the normal cornea (all p<0.01, n=5/group, Figure 3). After treatment with 100 μM fasudil eye drops, the intensity of ROS-associated DHE fluorescence was significantly reduced in the corneal epithelium and stroma (all p<0.01, n=5/group, Figure 3).There was no statistical difference in the production of ROS between the mice with and without PBS treatment after injury, either in the corneal epithelium or in the stroma.

Figure 3

The production of reactive oxygen species (ROS) in the burned murine corneas with and without fasudil treatment. A: To evaluate the effect of fasudil on the production of ROS in the cornea after alkali burns using the dihydroethidium (DHE) assay, the intensity of fluorescence was scaled by integral optical density. B: In the corneal epithelium, the intensity of DHE fluorescence in the burned corneas after treatment with or without phosphate-buffered saline (PBS) was significantly higher than that in the normal corneas (503.9±56.7 and 496.8±44.9 versus 216.9±28.6, respectively; n=5/group, **p<0.01;), while it was significantly reduced after treatment with 100 μM fasudil eye drops (281.2±33.0; **p<0.01). C: In the corneal stroma, the intensity of DHE fluorescence in the burned corneas after treatment with or without PBS was significantly higher than that in the normal corneas (300.7±65.8 and 266.3±27.6 versus 132.9±21.1; n=5 per group, **p<0.01), while it was significantly reduced after treatment with 100μM fasudil eye drops (162±34.1; **p<0.01).

100 μM fasudil reduced the infiltration of inflammatory cells in the alkali-burned murine corneas

The histopathological analysis reveals a significant decrease in the number of infiltrated macrophages, lymphocytes, and PMNs in murine corneas after treatment with 100 μM fasudil eye drops compared to corneas treated with PBS (18.4±5.31 versus 47.8±9.01 per field, p<0.01; Figure 4A-C).The number of monoclonal antibody stained PMNs in the corneal stroma was significantly lower in the 100 μM fasudil treated group than in the PBS treated group (5.0±4.6 versus 16.0±5.0 per field, p<0.01; Figure 4D-F).

Figure 4

Anti-inflammatory effects of fasudil on the alkali-burned corneas. A–C: Hematoxylin-eosin staining of the murine corneas (magnification ×400). The number of infiltrated polymorphonuclear neutrophils (PMNs), macrophages, and lymphocytes in the corneal stroma per field was lower with 100 μM fasudil treatment than with phosphate-buffered saline (PBS) treatment. D–F: Immunohistochemical staining of PMNs. The endothelium of corneal vessels is indicated with red arrows. The number of infiltrated PMNs (brown staining, black arrows) in the corneal stroma per field was lower after100 μM fasudil treatment than with PBS treatment (n=7/group, ** indicates p<0.01).

Effects of 100 μM fasudil on the mRNA expression of angiogenic genes in alkali-burned murine corneas

The mRNA expression levels of VEGF, VEGFR1/2, IL-1β, TNF-a, and MMP-2, MMP-8, and MMP-9 were detected and compared by quantitative RT-PCR. Gene expression was significantly lower in the 100 μM fasudil treated group; the expression levels were VEGF, 34.14%; TNF-α, 18.82%; MMP-8, 27.63%; and MMP-9, 12.57% compared to those in the PBS treated group. However, there were no statistical differences in the expression of VEGFR1/2, IL-1β, and MMP-2 between these two groups (Figure 5).

Figure 5

The mRNA expression of angiogenesis-related genes detected by real-time polymerase chain reaction (RT-PCR). All the data represented the relative fold change of mRNA expression of the genes of interest. There were statistically significant differences in the expression of VEGF, TNF-a, MMP-8, and MMP-9 between the 100 μM fasudil treated group and the phosphate-buffered saline (PBS) treated group (n=8/group, * indicates p<0.05, ** indicates p<0.01).

Effect of 100 μM fasudil on the expression of heme oxygenase-1 (HO-1) protein in the murine corneas after the alkali burn

To assess the effect of fasudil on HO-1 expression in the murine corneas, we examined the HO-1 protein on day 4 after the injury using western blotting. The results indicated that the optical intensity of HO-1 protein bands in the 100 μM fasudil treated group was 1.52±0.34 times that of HO-1 protein bands in the PBS treated group (p<0.05, Figure 6).

Figure 6

Effect of fasudil on the production of heme oxygenase-1 (HO-1) protein in the mice corneas after alkali burns. A: Representative image of western blotting bands for HO-1 protein in the murine corneas on day 4 after alkali burns. The intensity of fluorescence was scaled by integrated density. B: The protein expression of HO-1 after treatment with 100 μM fasudil eye drops was significantly enhanced compared to after phosphate-buffered saline (PBS) treatment (n=5 per group, * indicates p<0.05).

Discussion

The results of this study show that fasudil can inhibit alkali burn-induced CNV. The potency of fasudil was also demonstrated by its inhibition of hypoxia-induced angiogenesis in pulmonary hypertension and tumors [20,21] and diabetes-induced microvascular damage in the retina [22]. In this study, we found 100 μM fasudil hydrochloride eye drops to be most effective in inhibiting CNV. This effect of fasudil was not dose-dependent and inhibited cancer cell proliferation in a concentration-dependent manner [23].

Inflammation is heavily involved in CNV, especially in chemical burn-induced CNV. After alkali burns, the inflammatory cells are recruited into the injured cornea, and some inflammatory cytokines are released. Following this, CNV will be induced, and the healing of corneal epithelial defects is delayed [24,25]. Most anti-inflammatory agents have shown a capacity to inhibit CNV, such as anti-VEGF agents [4,5], VEGF receptor inhibitors [26], anti-TNF-α agents [27], and anti-MMP agents [8]. The immunohistopathological results of this study show that the number of infiltrated inflammatory cells, including PMNs, in the murine cornea after alkali burns significantly decreased after the application of fasudil. Meanwhile, the mRNA expression of inflammatory cytokines, such as VEGF, TNF-α, MMP-8, and MMP-9, were downregulated. In previous reports, ROCK inhibition by fasudil could protect the vascular endothelium by inhibiting neutrophil adhesion and reducing neutrophil-induced endothelial injury in diabetic rats [22] and by decreasing the level of involved inflammatory cytokines, such as VEGF [28], TNF-α [29], MMP-2, and MMP-9 [30]. Our study further demonstrates the effectiveness of fasudil in reducing the infiltration of inflammatory cells, including PMNs, to the injured sites after which the associated inflammatory cytokines decrease in number. However, this study demonstrates that fasudil could not significantly reduce the mRNA expression of VEGFR1 and VEGFR2. Kubota et al. also demonstrated that fasudil had no effect on VEGFR2 and its phosphorylation in bovine retinal microvascular endothelial cells (BRECs) [10]. Furthermore, fasudil did not downregulate the mRNA expression of IL-1β and MMP-2 in this study, as it did in LPS-induced inflammation in endotoxemic mice [31] and in A549 lung cancer cells [23]. The mechanisms that contributed to this difference are still unknown.

Beyond inhibiting PMNs via the ROCK pathway, fasudil has been shown to reduce ROS production and potency after alkali burns in this study. The results of an ROS-related DHE fluorescence assay show that fasudil significantly reduced the production of ROS caused by alkali burns in the corneal epithelium and stroma. ROS formation in the cornea was enhanced immediately after alkali injury [10]. ROS, such as the superoxide radical (), hydrogen peroxide (), and the hydroxyl radical (), have been shown to be a driving force in the occurrence and development of CNV [10]. ROS can upregulate the expression of inflammatory cytokines (such as VEGF, interleukins, and TNF-α) via the nuclear factor-κB pathway [32,33] and promote the formation of CNV. Kubota et al. demonstrated that the application of antioxidant agents such as N-acetyl-L-cysteine for six consecutive days beginning from day 3 before the alkali burn can significantly reduce ROS production [10]. As an antioxidant agent, fasudil can reduce ROS production by inhibiting ROCK activation in vitro [15].

Furthermore, fasudil increased the expression of HO-1 protein after alkali burns in this study. HO-1 has been recognized as an important therapeutic target for cardiovascular diseases with high oxidative-stress levels, such as hypertension, diabetes, obesity, and myocardial ischemia-reperfusion injury [34-36]. The protective effect of increased HO-1 expression against endogenous ROS includes a reduction in cellular heme levels, the induction of ferritin, and the increased formation of bilirubin [37]. The evidence gathered shows that upregulation of HO-1 can not only reduce ROS levels in cells [38] but also prevent inflammation by inhibiting TLR pathway activation in vitro [39]. Thus, it is reasonable to assume that fasudil may suppress inflammation through the upregulation of HO-1 protein and the reduction of ROS formation via the Rho/ROCK pathway.

As a Rho-associated protein kinase (ROCK) inhibitor, fasudil promoted the healing of corneal epithelial defects after alkali burns in this study. Yin et al. also confirmed that Y-27632, a selective Rho-kinase inhibitor like fasudil, could accelerate the healing of corneal epithelial defects in wounded human corneal epithelial cells (HCECs) via the Rho/Rock pathway in vitro [40]. ROCK has been suggested to be involved in corneal epithelial differentiation [41], cell cycle progression [42], cell-cell adhesion [43], and cell-matrix interaction [44]. The inhibition of ROCK could enhance wound closure in cultured HCECs [40]. The promoted recovery of corneal wound healing may be attributed in part to the reduction of inflammatory cell infiltration and subsequent inhibition of CNV formation [45]. Our results were consistent with a previous report in which a specific ROCK inhibitor, Y-27632, could promote epithelial migration and attenuate cell proliferation. ROCK may play a negative role in regulating corneal epithelial wound healing. When negative regulation is inhibited, corneal wound healing may be promoted. However, another report showed that the acceleration of corneal epithelial migration activated by LPA (1-acyl-2-hydroxy-sn-glycero-3-phosphate) could be prevented by exoenzyme C3, a Rho inhibitor [46]. LPA acts at a cell surface receptor and stimulates corneal epithelial migration in an exoenzyme C3-sensitive manner through Rho. These conflicting results may be due to the different effects of Rho and ROCK inhibitors. The downstream effectors of Rho include ROCK, protein kinase N, rhotekin, citron, and mDia [47-51]. Theoretically, the activation of ROCK by LPA may inhibit epithelial migration. However, the other concurrent activated downstream effectors like mDia may negate the effect of ROCK activation. A recent report suggested that mDia works concurrently with ROCKs. The Rho-induced actin reorganization and various actin fiber patterns rely on the balance between mDia and ROCK actions [50].

In summary, fasudil can inhibit the formation of CNV and accelerate the healing of corneal epithelial defects, as shown in this study. This may be attributed to the ability of fasudil to reduce the infiltration of inflammatory cells and the production of ROS as a potent ROCK inhibitor and its ability to increase HO-1 production to serve as an antioxidant. Further investigation is necessary to determine the pharmacodynamics, pharmacokinetics, safety, and tolerability of fasudil hydrochloride eye drops for ophthalmic application.