Emodin suppresses alkali burn-induced corneal inflammation and neovascularization by the vascular endothelial growth factor receptor 2 signaling pathway

Abstract

OBJECTIVE:

To investigate the effects of emodin on alkali burn-induced corneal inflammation and neovascularization.

METHODS:

The ability of emodin to target vascular endothelial growth factor receptor 2 (VEGFR2) was predicted by molecular docking. The effects of emodin on the invasion, migration, and proliferation of human umbilical vein endothelial cells (HUVEC) were determined by cell counting kit-8, Transwell, and tube formation assays. Analysis of apoptosis was performed by flow cytometry. CD31 levels were examined by immunofluorescence. The abundance and phosphorylation state of VEGFR2, protein kinase B (Akt), signal transducer and activator of transcription 3 (STAT3), and P38 were examined by immunoblot analysis. Corneal alkali burn was performed on 40 mice. Animals were divided randomly into two groups, and the alkali-burned eyes were then treated with drops of either 10 μM emodin or phosphate buffered saline (PBS) four times a day. Slit-lamp microscopy was used to evaluate inflammation and corneal neovascularization (CNV) in all eyes on Days 0, 7, 10, and 14. The mice were killed humanely 14 d after the alkali burn, and their corneas were removed and preserved at －80 ℃ until histological study or protein extraction.

RESULTS:

Molecular docking confirmed that emodin was able to target VEGFR2. The findings revealed that emodin decreased the invasion, migration, angiogenesis, and proliferation of HUVEC in a dose-dependent manner. In mice, emodin suppressed corneal inflammatory cell infiltration and inhibited the development of corneal neovascularization induced by alkali burn. Compared to those of the PBS-treated group, lower VEGFR2 expression and CD31 levels were found in the emodin-treated group. Emodin dramatically decreased the expression of VEGFR2, p-VEGFR2, p-Akt, p-STAT3, and p-P38 in VEGF-treated HUVEC.

CONCLUSION:

This study provides a new avenue for evaluating the molecular mechanisms underlying corneal inflammation and neovascularization. Emodin might be a promising new therapeutic option for corneal alkali burns.

1. INTRODUCTION

Alkali-induced ocular damage accounts for approximately 60% of chemical injuries of the eye, while chemical eye injuries cause 11.5%-22.1% of all ocular traumas.¹ Corneal alkali burns can cause corneal neovascularization (CNV), resulting in loss of corneal transparency or permanent vision loss. Although the mechanism of CNV is not entirely known, an imbalance between antiangiogenic and angiogenic factors is assumed to be the major etiology of corneal opacity and neovascularization following alkali damage.^2,3 Most current therapy techniques for alkali-induced corneal damage, inflammatory neovascularization, and inflammation are not very successful in restoring normal corneal functionality and transparency.^1,2 As a result, novel therapeutic techniques are required to enhance the treatment of corneal alkali damage.

The key regulator of angiogenesis in health and illness is vascular endothelial growth factor (VEGF), which is one of the endogenous molecules that induces angiogenesis.⁴ VEGF’s impact on vascular endothelial cells is primarily regulated by two receptors, vascular endothelial growth factor receptor (VEGFR) 1 and 2. The angiogenic response to VEGF in vivo is mostly mediated by VEGFR2 activation, which initiates signaling cascades with multiple intermediates. Finally, endothelial cells produce different responses: proliferation, cell migration, vascular permeability, and invasion of surrounding tissues.⁵ Therefore, the VEGFR2 signaling pathway has been proposed as a target to inhibit blood vessel formation.

Emodin (1,3,8-trihydroxy-6-methyl-anthraquinone), a common chemical component found in many Chinese herbs, has been used for more than 2000 years in China.⁶ Emodin has several pharmacological actions, including anti-inflammatory, antiviral, antibacterial, and hepatoprotective properties.^7,⇓,⇓-10 Moreover, the anticancer property of emodin is considered the most important activity, and some studies have demonstrated its potential antitumor mechanism. Emodin can prevent human breast cancer cells from migrating, invading, and metastasizing in vitro and in vivo in MDA-MB-231 cells.¹¹ Additionally, emodin can inhibit the adhesion, growth, and migration of cells in human colorectal cancer by effectively inhibiting the VEGFR2 signaling pathway.¹² However, existing research on emodin has primarily focused on its antitumor activity, and its effect on eye diseases, especially on corneal neovascularization, is still poorly understood. These findings led us to hypothesize that emodin may inhibit corneal inflammation and neovascularization by inhibiting the VEGFR2 signaling pathway.

We used an in vitro model to assess emodin’s effects in the first phase of validating this hypothesis.¹³ The alkali burn model, a well-established corneal neova-scularization model, was then employed to assess the therapeutic effects of emodin on corneal inflammation and angiogenesis. Our findings offer additional insight into the critical role of emodin in angiogenic regulation and corneal inflammation.

2. MATERIALS AND METHODS

2.1. Cell culture

Endothelial cell medium (ECM, ScienCell, Carlsbad, CA, USA) with 5% heat-inactivated fetal bovine serum (ScienCell, Carlsbad, CA, USA), 1% endothelial cell growth supplement (ScienCell, Carlsbad, CA, USA) and 1% penicillin/streptomycin (ScienCell, Carlsbad, CA, USA) was used to culture human umbilical vein endothelial cells (HUVEC) at 37 ℃ and 5% CO₂.

2.2. Cell Counting Kit-8 assay

We digested logarithmic growth phase HUVEC with 0.25% trypsin-ethylenediaminetetraacetic acid (trypsin-EDTA, Beyotime, Shanghai, China), prepared a cell suspension and inoculated cells into a 96-well plate (1 × 10⁴ cells/well). Cells were treated in a CO₂ incubator at 37 ℃ with emodin (0, 2. 5, 5, 10, 15, and 20 µM). Emodin was dissolved in 80% alcohol and 20% Tween 80 solution. After 24 h, we added 10 µL of cell counting kit-8 (CCK-8) solution (Vazyme, Nanjing, China) per well, and the cells were incubated at 37 ℃ for 1-2 h. A microplate reader (Varioskan LUX, Thermo Scientific, Waltham, MA, USA) was used to measure each well’s absorbance at 450 nm.

2.3. Transwell migration and invasion assays

During our cell migration assay, 1 × 10⁵ cells were loaded into the upper chamber inserts (Corning Inc., Corning, NY, USA) with 200 µL of serum-free ECM. The lower chambers contained 600 µL of ECM. After 24 h, we fixed the cells with paraformaldehyde, performed crystal violet staining and counted them in 3 different fields via a light microscope. Cell invasion assays were performed using chambers precoated with Matrigel (BD, Franklin Lakes, NJ, USA). All of the assays and trials were carried out three times.

2.4. Tube formation assay

The 24-well plates were covered with 200 µL of Matrigel basement membrane matrix (BD, Franklin Lakes, NJ, USA) for this test, and HUVEC (2 × 10⁴ cells/well) were cultured for an hour in growth factor-free medium at 37 ℃. Afterward, the cells were treated with phosphate buffered saline (PBS), 10 µM emodin, or 20 µM emodin at 37 ℃ for 6 h. After 6 h of incubation, tube-like structures were photographed by an inverted microscope. At least three fields were then selected from each well. The average length was determined by calculating the tube length in three random fields from each well.

2.5. Analysis of apoptosis by flow cytometry

The 6-well plates were seeded with HUVEC (1 × 10⁶ cells/well) for 24 h and subsequently treated with 10 and 20 µM emodin for 24 h. HUVEC were collected and centrifuged for 5 min at 1000 × g at 4 ℃ before being washed with cold PBS twice. Next, we resuspended the cells at a density of 5 × 10⁵ cells/mL in 100 μL of binding buffer. Then, 5 μL of propidium iodide (PI) solution and 5 μL of Annexin V/fluorescein isothiocyanate (FITC) were added. Finally, we added 400 µL of 1 × binding buffer to each tube before detecting the apoptosis rate with a flow cytometer (BD, Franklin Lakes, NJ, USA) and FACSDiva software (BD, Franklin Lakes, NJ, USA).

2.6. Western blot assay

The cells and corneal tissue were homogenized by ultrasonication after being lysed in ice-cold radioimmunoprecipitation assay (RIPA) buffer containing protease inhibitors. The bicinchoninic acid (BCA) technique was used to assess protein content. The same protein quantities were exposed to sodium dodecyl sulfate-polyacrylamide gel electrophoresis and transferred to polyvinylidene difluoride (PVDF) membranes. Next, we used 2% bovine serum albumin (BSA) to block the membranes for 1 h and blotted with primary antibodies against p-VEGF2 (Cat# AF3279) (Affinity, Changzhou, China), VEGFR2 (Cat# AF6281) (Affinity, Changzhou, China), p-signal transducer and activator of transcription 3 (p-STAT3; Cat# AF3293) (Affinity, Changzhou, China), STAT3 (Cat# AF6294) (Affinity, Changzhou, China), p-P38 (Cat# 28796-1-AP) (Proteintech, Wuhan, China), P38 (Cat# 14064-1-AP) (Proteintech, Wuhan, China), p-serine/threonine kinase (p-Akt; Cat# 66444-1-Ig) (Proteintech, Wuhan, China), Akt (Cat# 10176-2-AP) (Proteintech, Wuhan, China), and β-actin (Cat# 20536-1-AP) (Proteintech, Wuhan, China) at 4 ℃ overnight. After the tris-buffered saline Tween-20 (TBST) wash, the blots were incubated for 1 h with a secondary antibody at room temperature with horseradish peroxidase. In a dark room, after washing with TBST, enhanced chemiluminescence (ECL) reagent was applied to the membrane to develop the film. ImageJ (National Institutes of Health, Bethesda, MD, USA) was utilized to determine the gray values of each band, with β-actin acting as the reference control.

2.7. Alkali-burned mouse cornea model and treatment

C57BL/6 mice of SPF grade (40 males), two-month-old, weighing (20 ± 2) g, were purchased from Shanghai SLAC Laboratory Animal Co, Ltd., Shanghai, China (License No. SCXK [HU] 2017-0012). All animal research was approved by the Experimental Animal Ethics Review Committee of Fuzhou University in Fuzhou, China, and the Regional Association for Research in Vision and Ophthalmology. The mice were anesthetized with an intraperitoneal injection of 40 mg/kg pentobarbital and examined for neovascularization, inflammatory reactions, and other abnormalities. Before the ocular surface was rinsed with 20 mL of 0.9% saline solution, a filter paper disc (2.5 mm in diameter) saturated in 1 M NaOH was placed on the central cornea for 30 s. The alkali-burned eyes were then treated with drops of either 10 μM emodin or PBS four times a day (n = 20). Slit-lamp microscopy was used to evaluate inflammation and CNV in all eyes on Days 0, 7, 10, and 14. The mice were killed humanely 14 d after the alkali burn, and their corneas were removed and preserved at －80 ℃ until histological study or protein extraction.

2.8. Hematoxylin-eosin staining

After mice were sacrificed, the cornea samples were separated and fixed in eye fixative (Servicebio, Wuhan, China) for 24 h, subsequently embedded in paraffin, and sliced 5 μm. Histologic sections of tissues were stained with hematoxylin and eosin, observed under an upright fluorescence microscope (Nikon Eclipse E100, Tokyo, Japan), and photographed. Five fields were taken from each slide at 20 ×.

2.9. Immunofluorescence analyses

Four percent paraformaldehyde was used to fix the eyes after removal for one day before being implanted and sectioned. For staining, 5 μm thick paraffin slices were cut and stained. For reduction of the inherent peroxidase activity, slices were submerged briefly in xylene, after which they underwent hydration with graded ethanol solutions and incubation for 5 min in 3% hydrogen peroxide. Then, they were exposed to 3% (w/v) BSA for 30 min during PBS blocking. The sections were then treated overnight at 4 ℃with anti-CD31 primary antibodies (Servicebio, Wuhan, China), followed by 2 h at room temperature with secondary antibodies (Servicebio, Wuhan, China). The nuclei were stained with a 4',6-diamidino-2-phenylindole mounting solution after multiple rinses (DAPI). Each slide was examined, and images were acquired using an upright fluorescence microscope (Nikon Eclipse C1, Tokyo, Japan) and Nikon DS-U3 cameras (Nikon DS-U3; Tokyo, Japan). Every experiment was carried out in triplicate.

2.10. Molecular docking

Docking analysis was performed with Auto Dock Tool software (version 1.5.7, http://mgltools.scripps.edu) to study the interaction between emodin and VEGFR2 proteins. The crystal structure of VEGFR2 with a resolution of 1.64Å (3VNT) was obtained from the Protein Data Bank (PDB) database. The water molecules in the target protein were removed using Auto Dock Tool software. It was saved in PDBQT format, and the processed VEGFR2 was hydrogenated and charged in Auto Duck Tools software. It was then molecularly docked with the emodin. The affinity was analyzed by looking at the binding energy, and the docking results were analyzed using PyMol software (version 1.7.6.0, http://www.pymol.org/pymol).

2.11. Statistical analyses

Every experiment was carried out in triplicate, and statistical analyses were carried out with the JASP software program (Version 0. 14. 1, JASP Team, Amsterdam, Netherlands). For comparison of more than two groups, Student’s t test was employed to assess notable differences, and a least significant difference test via one-way analysis of variance was utilized. We reported data as the mean ± standard deviation and a value of P < 0. 05 was considered statistically significant.

3. RESULTS

3.1. Molecular docking analysis of emodin with VEGFR2

In recent years, molecular docking methods have become an important technology in the field of computer-aided drug research. The semiflexible docking method allows the conformation of small molecules to change to a certain extent during the docking process. With semiflexible docking with AutoDock Tool software, the docking binding energy of emodin and VEGFR2 was －6.75 kcal/mol. When the docking binding energy is less than －1.2 kcal/mol, the docking result is considered good. Therefore, the docking site is reliable. Five hydrogen bonds were generated at the docking site, and the positions are at the 863th, 864th, 917th and 918th amino acids. The docking results were analyzed by PyMol software. The amino acids that formed hydrogen bonds with emodin were Arg-863, Thr-864, Glu-917 and Phe-918, and the hydrogen bond distances formed were between 1.9 and 3.1 (supplementary Figure 1).

3.2. Emodin decreased the viability of HUVEC by regulating HUVEC proliferation and apoptosis

We initially assessed cell viability in the presence of various emodin concentrations using CCK-8 assays, which revealed that the inhibitory effect of emodin on HUVEC viability started at a concentration of 10 μM in a dose-dependent manner (Figure 1A). Furthermore, emodin induced apoptosis in HUVEC (Figure 1C). The percentages of early apoptotic cells following treatment with 10 µM emodin and 20 µM emodin were 9.19% and 14.7%, respectively (Figure 1B).

Figure 1. Apoptosis and cell viability with exposure to different doses of emodin were identified using flow cytometry and the CCK-8 assay.

A: HUVEC were incubated with different concentrations (0-20 µM, 24 h) of emodin, CCK-8 assay was applied to determine cells viability; B: the rate of cell apoptosis in each group, the ratio of the second quadrant plus the fourth quadrant represents the ratio of cell apoptosis; C: apoptosis was assessed by flow cytometry assay. C1: 0 μM emodin group; C2: 10 μM emodin group; C3: 20 μM emodin group. CCK8: cell-counting-kit-8; HUVEC: human umbilical vein endothelial cell. The mean ± standard deviation is indicated by each value (n = 3). ^aP < 0.05, compared with 0 μM emodin treatment group.

3.3. Emodin inhibited cell migration, invasion, and tube formation in HUVEC

After confirming the influences of emodin on cell proliferation and apoptosis, we investigated its effects on HUVEC migration and invasion using Transwell experiments with varying doses of emodin. In the Transwell chamber, the number of cells migrating to the bottom side of the membrane in the emodin-treated group was much lower than that in the normal control group (Figure 2A, 2D). The data obtained in the invasion experiment also showed that emodin inhibited the invasive behavior of HUVEC (Figure 2B, 2E). The findings of the subsequent tube-formation assay indicated that emodin significantly inhibited the formation of capillary-like structures (Figure 2C, 2F). Based on results 3.1 and 3.2, 10 µM emodin was chosen for the following in vivo experiment.

Figure 2. Emodin inhibited migration, invasion, and tube formation of HUVEC .

A: HUVEC were treated with various concentrations of (0, 10 and 20 μM) emodin 24 h, and the migration of HUVEC was determined by transwell migration assay; B: HUVEC were treated with various concentrations of emodin (0, 10 and 20 μM), and the migration of HUVEC was determined transwell invasion assay; C: HUVEC were treated with various concentrations of emodin (0, 10 and 20 μM), and the migration of HUVEC was determined by matrigel tube-formation assay; D: results of quantitative analysis of HUVEC transwell migration assay. E: results of quantitative analysis of HUVEC transwell invasion assay. F: results of quantitative analysis of HUVEC matrigel tube-formation assay. A1, B1, C1: cells treated with 0 μM emodin 24 h; A2, B2, C2: cells treated with 10 μM emodin 24 h; A3, B3, C3: cells treated with 20 μM emodin 24 h. HUVEC: human umbilical vein endothelial cells. The mean ± standard deviation is indicated by each value (n = 3). ^aP < 0.01, ^bP < 0.05, compared with 0 μM emodin group.

3.4. Emodin inhibited inflammation and neovascularization in alkali-injured mouse corneas

Next, we evaluated the effect of emodin on CNV by comparing the total neovascularization area between the emodin-treated and control (PBS) groups 14 d after corneal cauterization. Images on Day 14 post-injury revealed visibly less neovascularization in the group treated with emodin than in the group treated with PBS (Figure 3A). At 14 d after alkali burn, inflammatory cell infiltration was greatly enhanced in the mouse cornea and was significantly reduced by emodin (Figure 3B). The CNV area was normalized to the entire corneal area. Compared to that of the group treated with PBS, the proportion of CNV in the group treated with emodin was considerably lower (Figure 3C). The level of inflammatory cells in the PBS group was higher than that in the emodin group at Day 14 (Figure 3D).

Figure 3. In mouse corneas, emodin inhibits alkali burn-induced neovascularization and inflammatory cell infiltration.

A: slit-lamp microscopy was used to detect CNV on Days 7 and 14; A1: pictures of corneas of mouse 7 d after treatment of corneal alkali burns with PBS eye drops. A2: pictures of corneas of mouse 14 d after treatment of corneal alkali burns with PBS eye drops. A3: pictures of corneas of mouse 7 d after treatment of corneal alkali burns with emodin eye drops. A4: pictures of corneas of mouse 14 d after treatment of corneal alkali burns with emodin eye drops. B: corneal inflammatory cell infiltration in various groups of mouse (hematoxylin-eosin staining, × 100); B1: normal mouse corneas without any treatment; B2: treatment of mouse corneal alkali burns with PBS drops for 14 d; B3: treatment of mouse corneal alkali burns with emodin drops for 14 d; C: statistical analysis of the CNV-covered area at various time points; D: quantification of inflammatory cell infiltration on 14 d. CNV: corneal neovascularization. PBS: phosphate buffered saline. The mean ± standard deviation is indicated by each value (n = 3). ^aP < 0.05, ^bP < 0.01, compared with PBS group.

3.5. Emodin inhibited neovascularization through the VEGFR2 signaling pathway

To further evaluate the antineovascularization effects in the cornea, we performed immunofluorescence analysis for CD31. The number and area of CD31-positive blood vessels decreased in the group treated with emodin compared to the group treated with PBS, as per our prediction (Figure 4A, 4B). We also used Western blotting to examine VEGFR2 expression in the mouse cornea. The levels of VEGFR2 in the control and PBS-treated groups were considerably greater than those in the group treated with emodin (Figure 4E, 4F). Although our findings suggest that emodin modulated VEGFR2 expression in the corneal alkali burn model, the mechanism of emodin’s action has remained unknown.

Figure 4. Effect of emodin on the expression of CD31 and VEGFR2 signaling pathway after corneal alkali burns.

A: expression of CD31 in mouse cornea by immunofluorescence detection; A1: normal mouse corneas without any treatment; A2: treatment of mouse corneal alkali burns with PBS drops for 14 d; A3: treatment of mouse corneal alkali burns with emodin drops for 14 d; B: quantitative analysis of CD31 expression. Control: blank control group mice, without any treatment. PBS: negative control group mice treated with phosphate buffered saline eye drops. Emodin: experimental group mice, treated with emodin eye drops; C: quantitative analysis of VEGFR2, pVEGFR2, pSTAT3, tSTAT3, pPI3K, tPI3K, pAkt, and tAkt; D: representative Western blot of VEGFR2, pVEGFR2, pSTAT3, tSTAT3, pPI3K, tPI3K, pAkt, and tAkt after treatment of VEGF-stimulated HUVEC with PBS and 20 µM emodin for 24 h; E: VEGFR2 expression bands in alkali-burned mouse corneas from the emodin treatment and control groups on 14 d; F: quantitative analysis of VEGFR2 in alkali-burned mouse corneas from the emodin treatment and control groups on 14 d. CD31: platelet endothelial cell adhesion molecule-1; VEGF: vascular endothelial growth factor; VEGFR2: vascular endothelial growth factor receptor 2; p-VEGFR2: phospho-vascular endothelial growth factor receptor 2. STAT3: signal transducer and activator of transcription 3; p-STAT3: phospho-signal transducer and activator of transcription 3; PI3K: phosphoinosmde-3-kinase; p-PI3K: phospho-phosphoinosmde-3-kinase; Akt: total protein kinase B; p-Akt: phospho-protein kinase B. The mean ± standard deviation is indicated by each value (n = 3). ^aP < 0.05, ^bP < 0.01, compared with PBS group.

We identified the expression levels of VEGFR2-related signaling pathways in HUVEC to investigate the underlying molecular mechanism of these alterations. The levels of VEGFR2 and p-VEGFR2 in the group treated with emodin were lower than those in the group treated with PBS, according to the Western blot assay (Figure 4D). To better understand the signaling cascade initiated by VEGFR2, we also measured the expression of p-STAT3, p-Akt, and p-P38 (Figure 4D). Following emodin treatment, the levels of p-STAT3, p-Akt, and p-P38 were drastically reduced in the VEGF-cultured HUVEC (Figure 4C).

4. DISCUSSION

As the key and difficult point in the study of ophthalmology treatment, ocular alkali burn is also one of the most difficult emergencies in ophthalmology. In general, when alkali burns occur, the blood vessels around the corneal rim dilate, which will increase vascular permeability, and the aggregation and migration of inflammatory cells occur, together with the activation of endothelial cells, to form new blood vessels.¹⁴ In this study, the corneal alkali burn model was adopted as a representative model of ocular neovascularization and inflammatory disease, where corneal ulcers, severe keratitis, corneal neovascularization, and corneal scarring occurred in mice.

Emodin is a rhubarb-derived anthraquinone derivative that has been proven in recent years to have a variety of pharmacological characteristics, with its potential anticancer benefits being established by various researchers. These studies have revealed that it promotes a variety of antitumor processes, including apoptosis induction, antimetastatic activity, and cell cycle arrest.^15,⇓-17 Acute corneal inflammation was observed in lipopoly-saccharides-induced rats, while the inflammatory response and structure of the cornea were improved in rats pretreated with emodin.¹⁸ Nonetheless, the mechanism behind the antiangiogenic properties of emodin remains to be fully explored; hence, this study investigated the effects of emodin on HUVEC activity, invasion, tube formation, and migration.

The proliferation of HUVEC in vitro was obviously inhibited by emodin, which also significantly increased the early apoptosis of HUVEC. Moreover, emodin played an essential role in migration and invasion as well as in inhibiting tube formation. Experiments in vitro verified our hypothesis and confirmed that emodin had an inhibitory effect on neovascularization, followed by the testing of our hypothesis in vivo. To study the effects of emodin on keratitis and neovascularization as well as explore its underlying molecular mechanisms, we used a mouse corneal alkali burn model for in vivo experiments. The results showed that the degree of corneal neovascularization of the alkali-burned mice significantly decreased in the emodin-treated group compared with the PBS-treated group. On Day 14, hematoxylin-eosin staining showed that the infiltration of corneal inflammatory cells in the PBS control group was more obvious than that in the emodin treatment group, while the corneal inflammation index and neovascularization area in the emodin-treated group were lower than those in the PBS-treated group. CD31 is a vascular marker that plays an important role in promoting leukocyte and EC migration during inflammation and angiogenesis.¹⁹ Immunofluorescence analysis further demonstrated that the expression of CD31 in the cornea of the emodin-treated group was lower than that of the PBS-treated group. Thus, evidence from in vitro and in vivo studies suggests that emodin inhibits corneal neovascularization.

Recent studies have confirmed that endothelial cell proliferation, differentiation, migration, and lumen formation can lead to the formation of new blood vessels. VEGFR2 activates a variety of downstream signal transduction molecules, including P38 mitogen-activated protein kinase (MAPK), Akt, and STAT3, thereby promoting vascular endothelial growth and tumor progression. Among them, MAPK is an important transmitter of signals from the cell surface to the interior of the nucleus, while p38 inhibitors have been found to show good anti-inflammatory effects in animal models of inflammation-related diseases.²⁰ In addition, STAT3 is a transcription factor activated by a variety of cytokines and growth factors. Upon activation, it translocates into the nucleus and regulates apoptosis, proliferation, migration, and the expression of related genes in the nucleus.²¹ In retinal microvascular endothelial cells, small interfering RNA-mediated knockdown inhibited activation of VEGFR2 and phosphorylation of STAT3 protein, the latter of which inhibited microvascular endothelial cell proliferation and pathological neovascularization.²² Moreover, the PI3K/Akt signaling pathway is involved in a variety of important cellular processes, including cell growth, metabolism, survival, and angiogenesis.²³ Therefore, we hypothesized that emodin suppresses angiogenesis by inhibiting downstream signal transduction mediated by VEGFR2. Western blot analysis showed that compared to that of the control group and PBS-treated group, the level of VEGFR2 decreased markedly in the emodin-treated mouse corneal group. In addition, we examined the levels of VEGFR2, STAT3, Akt, and P38 in HUVEC treated with VEGF and emodin. The results revealed that the levels of p-Akt, VEGFR2, p-STAT3, p-VEGFR2, and p-P38 were considerably lower in the emodin treatment group than in the control group, which suggested that emodin reduces neovascularization by inhibiting VEGFR2 and its downstream STAT3/Akt/P38 signaling pathway.

In conclusion, our findings suggest that emodin, which targets VEGFR2 protein, can inhibit corneal inflammation and neovascularization after alkaline burn. In addition, under regulation of the VEGFR2-driven STAT3/Akt/P38 signaling pathway, the anti-inflammatory and antiangiogenic effects of emodin are mediated. Given the complex mechanism of corneal neovascularization, however, quantitative real-time polymerase chain reaction, clustered regularly interspaced short palindromic repeats, and assessment of the effects of anti-corneal neovascularization drugs will be included in future research. Therefore, emodin-based strategies, having good potential in clinical applications, can provide guidance for further CNV preclinical studies.

5. SUPPORTING INFORMATION

Supporting data to this article can be found online at http://journaltcm.cn.