This article is mentioned in:

Introduction

Aging and other factors, including increased oxidative stress and chronic inflammation, can diminish the function of the retinal pigment epithelium (RPE), which is located between the photoreceptors and vascular choroid in the eye, leading to eventual degeneration (1). Degenerated RPE cells may cause irreversible injury to photoreceptors and a disturbance in central vision, resulting in age-related macular degeneration (AMD) (2).

AMD is a progressive disease associated with retinal degeneration that can cause irreversible vision loss in elderly populations. A gradual increase in the prevalence of AMD has been observed among older individuals (3). AMD is a complex and multifaceted disease that is caused by a combination of genetic and environmental factors (4). Drusen accumulation in the macula between the RPE and Bruch's membrane has been observed in the early stages of AMD (4). Subsequently, AMD progresses to interconvertible ‘dry’ and ‘wet’ forms in the later stages (5). Dry AMD is characterized by the extensive loss of RPE cells, the overlying photoreceptors and the underlying choriocapillaris (6). Therefore, RPE cell degeneration by factors including increased oxidative stress and chronic inflammation may promote the irreversible injury of photoreceptors and loss of central vision, thus inducing AMD (7). However, at present, no effective treatment is available for AMD (8). Therefore, developing a new AMD model to investigate strategies for the prevention or delayed progression of the disease is important.

Oxidative stress is an important factor in the pathogenesis of AMD (9) that may initiate RPE cell damage, resulting in chronic inflammation (10). RPE cells can digest photoreceptor outer rod segments (POSs); however, the digestive rate of POSs decreases during oxidative stress, resulting in accumulated POS, which promotes lipofuscin formation, one of the major pathological alterations observed in AMD (11). Therefore, reducing oxidative stress may serve as a potential therapeutic strategy to prevent AMD.

Hydrogen peroxide (H2O2) is widely utilized to induce oxidative stress in human RPE cells (12), displaying dose- and time-dependent effects (12,13). Chemical oxidants, including tert-butyl hydroperoxide (t-BHP) (14) and bisphenol A (15), are also used to induce oxidative stress in vitro. Additionally, acrolein-induced reactive oxygen species (ROS) overproduction has been employed for the induction of oxidative stress (16). RPE cells are sensitive to oxidative stress as they contain photosensitizers that produce ROS in response to visible light (10). The antioxidative effects of RPE cells decrease with age, and aged RPE cells display an oxidative burden, which is a sign of AMD progression (12).

Potassium bromate (KBrO3) is an oxidant that is used as a food additive during the bread-baking process because it decomposes into a stable compound, potassium bromide (17). KBrO3-induced genotoxicity in HepG2 cells indicates that KBrO3 generates a variety of ROS, resulting in oxidative DNA damage (18). Additionally, KBrO3 induces an effect in cultured cardiac cells that was similar to the effect mediated by H2O2 (19). Superoxide anion radical generation by KBrO3 has also been observed (20). The aforementioned studies indicated that KBrO3 and H2O2 display the potential to induce AMD, thus it was hypothesized that KBrO3 may be used to induce AMD in cells.

In the present study, cultured ARPE-19 cells were employed to identify the effective dose of KBrO3 and explore the possibility of developing a new model of AMD.

Materials and methods

Cell culture

The human RPE cell line (ARPE-19) was obtained from The Culture Collection and Research Center, Food Industry Research and Development Institute [Bioresource Collection & Research Center (BCRC) no. 60383]. ARPE-19 cells were cultured in DMEM/F12 medium (Cytiva) with 10% fetal calf serum (Cytiva) and antibiotics (cat no. 15240062; Thermo Fisher Scientific, Inc.; solution contains 10,000 U/ml of penicillin, 10,000 µg/ml of streptomycin, and 25 µg/ml of Gibco Amphotericin B) at 37°C with 5% CO2 (12). 293 cells (BCRC no. 60019) were cultured in minimum essential medium (MEM) (Cytiva). After 24 h, the cell culture medium was refreshed. Following culture for 24 h, cells were used for subsequent experiments.

Cell viability assay

ARPE-19 and 293 cell viability was assessed by performing the MTT assay (19). The purple formazan was dissolved by DMSO. Briefly, ARPE-19 and 293 cells were seeded at 5,000 cells per well in 24-well plates and incubated with KBrO3 (25, 50 and 75 µM; Sigma-Aldrich, Merck KGaA) at 37°C for 24 h. 4-Hydroxynonenal (4-HNE; Cayman Chemical Company) and H2O2 (Sigma-Aldrich; Merck KGaA) was used as a reference to assess the effects of KBrO3. Subsequently, the culture medium was refreshed and 50 µl MTT (0.5 mg/ml) was added to each well for 4 h. The absorbance was measured at a wavelength of 595 nm using a Synergy HT Multi-Mode Microplate Reader (BioTek Instruments, Inc.).

Additionally, the cytotoxicity effect of KBrO3 in ARPE-19 and 293 cells was assessed using the lactate dehydrogenase (LDH) release assay kit (BioVision, Inc.) as previously described (12). The number of dead cells in the medium of cultured cells was measured to evaluate LDH activity. The absorbance was measured at a wavelength of 490 nm using a microplate spectrophotometer.

Determination of intracellular superoxide levels

According to the previously described protocol (19), ARPE-19 cells were seeded (7.5×103 cells/ml) into 24-well plates and incubated for 24 h. Following starvation in serum-free medium at 37°C for 4 h, cells were incubated with increasing amounts of KBrO3 (25, 50 and 75 µM) at 37°C for 72 h. Superoxide levels were measured by incubating cells with dihydroergotamine (DHE; Thermo Fisher Scientific, Inc.), which reacts with intracellular superoxide ions, for 30 min at 37°C. The entire field of vision was observed using an IX71 fluorescence microscope (Olympus Corporation; magnification, ×200) and a DP2-BSW imaging system (Olympus Corporation). The results were quantified using ImageJ software v2.0.0 (National Institutes of Health) (21).

Assay of intracellular ROS levels

ROS levels in ARPE-19 cells were assessed using the 2′,7′-dichlorodihydro-fluorescein (DCFH-DA) fluorescent probe (Sigma-Aldrich; Merck KGaA). ARPE-19 cells at 1×106 cells/ml were incubated with 25 µM DCFH-DA for 1 h at 37°C. Cells were pretreated with tiron (1 and 5 µM, Sigma-Aldrich; Merck KGaA) and phloroglucinol (1 and 5 µM, Sigma-Aldrich; Merck KGaA) for 30 min prior to KBrO3 at 37°C. The resulting fluorescence was measured at an excitation wavelength of 485 nm and an emission wavelength of 535 nm using a fluorescence plate reader as previously described (22).

Assay of antioxidant enzymes

Cells were harvested and sonicated at high frequency on ice. The lysates were incubated on ice for 30 min, to ensure protein solubilization. Cell debris was centrifuged at 10,000 × g for 20 min at 4°C. Total protein concentrations were measured by bicinchoninic acid protein assay (Thermo Fisher Scientific, Inc.).

Antioxidant enzyme activities, including superoxide dismutase (SOD) and glutathione peroxidase (GSHPx), were measured in total cell protein extracts using colorimetric assay kits (cat no. K555; BioVision, Inc.) according to the manufacturer's instructions. The absorbance was measured at a wavelength of 560 nm using a microplate spectrophotometer.

Western blotting

Total protein was extracted from ARPE-19 and 293 cell lysates using ice-cold RIPA buffer (Thermo Fisher Scientific, Inc.) (19). Total protein concentrations were measured by bicinchoninic acid protein assay (Thermo Fisher Scientific, Inc.). Proteins (10 µg) were separated by 10% or 15% SDS-PAGE and electro-transferred onto polyvinylidene fluoride membrane (EMD Millipore). After blocking with 5% (v/v) bovine serum albumin (MDBio, Inc.) for 1 h at room temperature, membranes were incubated with primary antibodies, Bcl2 (cat. no. 2870, 1:1,000; Cell Signaling Technology, Inc.), Bax (cat. no. 2772, 1:1,000; Cell Signaling Technology, Inc.), phosphorylated (p-)STAT3 (cat. no. ab76315, 1:1,000; Abcam), STAT3 (cat. no. ab68153, 1:1,000; Abcam) or β-actin (cat. no. sc-1615, 1:5,000; Santa Cruz Biotechnology, Inc.) at 4°C overnight. The membranes were then incubated with the corresponding secondary antibodies, goat anti-rabbit IgG (cat. no. 12-348, 1:5,000; Sigma-Aldrich; Merck KGaA) or goat anti-mouse IgG (cat. no. 12-349, 1:5,000; Sigma-Aldrich; Merck KGaA) at room temperature for 1 h. Protein bands were visualized using a chemiluminescence kit (PerkinElmer, Inc.). The optical densities of the bands for Bax (18 kDa), Bcl2 (100 kDa) and β-actin (43 kDa) were quantified using software (Gel-Pro Analyzer v4.0 software; Media Cybernetics Inc.) (23).

Reverse transcription-quantitative PCR (qPCR)

VEGF mRNA expression levels in ARPE-19 cells were measured. Total RNA was extracted from ARPE-19 and 293 cells using TRIzol® reagent (Invitrogen; Thermo Fisher Scientific, Inc.). Reverse-transcriptase reactions contained 5 µg of total RNA extract, 4 µl 5X PCR buffer, 1 µl 0.1 M DTT, 1 µl SuperScript™ III RT (cat. no. 18080093; Invitrogen; Thermo Fisher Scientific, Inc.), which were performed at room temperature for 10 min, then at 42°C for 30 min and at 95°C for 5 min. Subsequently, qPCR was performed using a LightCycler (Roche Diagnostics GmbH). The qPCR reactions contained LightCycler DNA Master Hybridization Probes reaction mix (Roche Diagnostics GmbH) and the dual-labeled fluorescent probes. PCR cycles were: Pretreatment at 95°C for 10 sec, 96°C for 10 sec, 60°C for 30 sec, 72°C for 1 sec, 40°C for 30 sec (45 cycles) and a final extension at 72°C for 10 min. Analysis of relative gene expression data was performed using the 2−ΔΔCq method (24). The following primers were used for qPCR: VEGF forward, 5′-AGGAGGGCAGAATCATCACG-3′ and reverse, 5′-CAAGGCCCACAGGGATTTTCT-3′; caspase-3 forward, 5′-CCGACTTCCTGTATGCTTACTCTA-3′ and reverse, 5′-CATGACCCGTCCCTTGAA-3′; and β-actin forward, 5′-CTAAGGCCAACCGTGAAAAG-3′ and reverse, 5′-GCCTGGATGGCTACGTACA-3′. VEGF mRNA, caspase-3 mRNA expression levels were normalized to the internal reference gene β-actin.

Statistical analysis

Data are presented as the mean ± SEM (n=6). All statistical analyses were carried out by SPSS v21 (IBM Corp.). Comparisons among multiple groups were analyzed using one-way ANOVA followed by Tukey's post hoc test. P<0.05 was considered to indicate a statistically significant difference.

Results

Effects of KBrO3 on ARPE-19 cell viability

Compared with the control group, incubation with KBrO3 for 24 h significantly inhibited ARPE-19 cell viability in a concentration-dependent manner KBrO3 decreased cell viability, and increased cell apoptosis and cell cytotoxicity at all three tested concentrations (Fig. 1A). The Bcl2/Bax ratio is a useful index for evaluating KBrO3-induced apoptosis (25). KBrO3 significantly decreased the Bcl2/Bax ratio in a concentration-dependent manner compared with the control group (Fig. 1B). Caspase-3 activity also serves as a biomarker of apoptosis (26). KBrO3 significantly decreased Caspase-3 expression levels in a concentration-dependent manner compared with the control group (Fig. 1C). The results were further supported by the results of the LDH assay that demonstrated that KBrO3 significantly increased cell cytotoxicity in a concentration-dependent manner compared with the control group (Fig. 1D). In contrast to the other 2 pro-oxidants (4-NHE, H2O2), KBrO3 showed lower cell toxicity at 25 µM. Moreover, 75 µM KBrO3 and 30 µM 4-HNE, which has been widely used to induce oxidative stress in RPE cells, displayed similar effects on ARPE-19 cells (23).

Figure 1. KBrO3 induces ARPE-19 cell injury in vitro. ARPE-19 cells were incubated with KBrO3, 4-HNE or H2O2 for 24 h. (A) ARPE-19 cell viability was assessed by performing the MTT assay. ARPE-19 cell apoptosis was assessed by measuring (B) the Bcl2/Bax ratio and (C) caspase-3 mRNA expression levels. (D) ARPE-19 cell viability was also assessed by performing the lactate dehydrogenase assay. Data are presented as the mean ± SEM (n=6). *P<0.05 vs. control; #P<0.05 vs. KBrO3 25 µM, potassium bromate; 4-HNE, 4-hydroxynonenal.

Effects of KBrO3 on oxidative stress biomarkers in ARPE-19 cells

In ARPE-19 cells, KBrO3 significantly inhibited the activities of antioxidant enzymes, including SOD or GSHPx, in a concentration-dependent manner compared with the control group (Fig. 2A and B). Therefore, the results indicated that KBrO3 induced oxidative stress in ARPE-19 cells.

Figure 2. KBrO3-mediated effects on oxidative stress biomarkers in ARPE-19 cells. ARPE-19 cells were incubated with KBrO3 for 24 h. Effect of KBrO3 on (A) SOD and (B) GSHPx activities in ARPE-19 cells. Effects of the antioxidants (C) tiron and (D) phloroglucinol on KBrO3-induced oxidative stress. Data are presented as the mean ± SEM (n=6). *P<0.05 and **P<0.01 vs. control; #P<0.05 and ##P<0.01 vs. KBrO3. KBrO3, potassium bromate; SOD, superoxide dismutase; GSHPx, glutathione peroxidase; ROS, reactive oxygen species.

To determine intracellular superoxide levels in ARPE-19 cells, DCFH-DA was used to detect intracellular ROS levels and DHE was used to detect intracellular superoxide levels. Compared with the control group, KBrO3 treatment significantly elevated ROS and superoxide levels (Fig. 2C and D). KBrO3-mediated effects were significantly inhibited by the antioxidants tiron and phloroglucinol in a concentration-dependent manner (Fig. 2C and D).

Inhibitory effects of antioxidants on KBrO3-mediated effects in ARPE-19 cells

Intracellular ROS levels were measured using the DCFH-DA fluorescence probe. Compared with the control group, KBrO3 significantly increased intracellular ROS levels in a concentration-dependent manner (Fig. 3A). Additionally, the effects of KBrO3 on intracellular ROS levels were significantly inhibited by the antioxidants tiron and phloroglucinol (Fig. 3B and C).

Figure 3. KBrO3 induces ARPE-19 cell apoptosis. (A) Effect of KBrO3 on intracellular ROS levels in ARPE-19 cells. Effects of the antioxidants (B) tiron and (C) phloroglucinol on KBrO3-induced intracellular ROS levels in ARPE-19 cells. Effects of the antioxidants (D) tiron and (E) phloroglucinol on KBrO3-induced ARPE-19 cell apoptosis were determined by measuring the Bcl2/Bax ratio. Data are presented as the mean ± SEM (n=6). *P<0.05 and **P<0.01 vs. control; #P<0.05 and ##P<0.01 vs. KBrO3. KBrO3, potassium bromate; ROS, reactive oxygen species.

Moreover, compared with the control group, KBrO3 notably increased Bax protein expression levels in ARPE-19 cells. The antioxidant phloroglucinol markedly inhibited KBrO3-induced Bax protein expression levels in a concentration-dependent manner (Fig. 3E). Additionally, phloroglucinol reversed KBrO3-mediated downregulation of Bcl2 protein expression levels in ARPE-19 cells (Fig. 3E). Therefore, phloroglucinol significantly attenuated KBrO3-mediated downregulation of the Bcl2/Bax ratio in ARPE-19 cells (Fig. 3E). The antioxidant tiron displayed similar effects on the Bcl2/Bax ratio (Fig. 3D). The results indicated that antioxidants inhibited KBrO3-mediated effects on apoptotic factors in ARPE-19 cells.

Role of 4-HNE in KBrO3-mediated effects

KBrO3-induced apoptosis might be mediated via endogenous 4-HNE in ARPE-19 cells, as 4-HNE is produced in RPE cells during high oxygen consumption (27). Recently, 4-HNE was reported to serve as an agonist of the nicotinic receptor GPR109A (28). Therefore, 293 cells lacking the GPR109A gene (29) were used to investigate the effects of KBrO3.

Compared with the control group, the three tested concentrations of KBrO3 significantly inhibited 293 cell viability in a concentration-dependent manner (Fig. 4A), although the effective dose was higher and the incubation time was two-fold higher in 293 cells compared with ARPE-19 cells. However, compared with the control group, 4-HNE failed to significantly inhibit 293 cell viability at a concentration sufficient to inhibit ARPE-19 cell viability. Additionally, compared with the control group, KBrO3 significantly induced 293 cell apoptosis, but 4-HNE did not significantly alter 293 cell apoptosis as demonstrated by significantly increased caspase-3 expression levels in KBrO3-treated cells compared with control cells, but no significant alteration in caspase-3 expression levels in 4-HNE-treated cells compared with control cells (Fig. 4B). KBrO3 showed higher apoptosis induction in all fractionated doses in contrast to 4-HNE. Similarly, KBrO3 significantly increased cytotoxicity in 293 cells compared with the control group, but 4-HNE did not significantly increase cytotoxicity in 293 cells (Fig. 4C). Moreover, the antioxidant tiron significantly increased the Bcl2/Bax ratio in KBrO3-treated 293 cells in a concentration-dependent manner (Fig. 4D). Therefore, the results suggested that 4-HNE did not serve a role in KBrO3-induced apoptosis.

Figure 4. Effect of 4-HNE on KBrO3-induced injury in 293 cells. (A) Cell viability was assessed by performing the MTT assay. (B) Cell apoptosis was assessed by measuring caspase-3 mRNA expression levels. (C) Cell viability was also assessed by performing the lactate dehydrogenase assay. (D) Effect of the antioxidants tiron on KBrO3-induced 293 cell apoptosis. Data are presented as the mean ± SEM (n=6). *P<0.05 and **P<0.01 vs. control; #P<0.05 vs. KBrO3. 4-HNE, 4-hydroxynonenal; KBrO3, potassium bromate.

Effects of KBrO3 on VEGF expression in ARPE-19 cells

Choroidal neovascularization (CNV) occurs during AMD, and VEGF is a critical signal in AMD-associated CNV (30). ROS inhibition may reduce VEGF mRNA expression levels (31). Therefore, alterations in VEGF mRNA expression levels in ARPE-19 cells treated with KBrO3 for 48 h were investigated.

Compared with the control group, KBrO3 significantly increased VEGF mRNA expression levels in a concentration-dependent manner at concentrations sufficient to induce ROS production (Fig. 5A). The upstream STAT3 may enhance VEGF gene expression (32). Compared with the control group, KBrO3 treatment also significantly increased STAT3 phosphorylation (Fig. 5B), which was inhibited by tiron. Cells were pretreated with stattic (0.5 and 1 µM; Sigma-Aldrich; Merck KGaA) for 30 min prior to addition of KBrO3. Stattic significantly attenuated KBrO3-mediated upregulation of VEGF mRNA expression levels (Fig. 5C). However, stattic failed to inhibit KBrO3-induced ROS production (Fig. 5D). Therefore, the results suggested that KBrO3 activated STAT3 via ROS to promote VEGF expression in ARPE-19 cells.

Figure 5. Effect of KBrO3 on VEGF expression in ARPE-19 cells. (A) Effect of KBrO3 on VEGF mRNA expression levels in ARPE-19 cells. (B) Effect of KBrO3 on STAT3 activation in ARPE-19 cells. The antioxidant tiron reversed KBrO3-mediated effects on STAT3 activation in ARPE-19 cells. Effect of static on KBrO3-mediated (C) VEGF mRNA expression levels and (D) ROS production in ARPE-19 cells. Data are presented as the mean ± SEM (n=6). *P<0.05 vs. control; #P<0.05 vs. KBrO3. KBrO3, potassium bromate; ROS, reactiveoxygen species; p, phosphorylated.

Discussion

The present study identified a novel AMD-like model using KBrO3 to induce oxidative stress in ARPE-19 cells. In the present study, KBrO3-induced cell injury was reversed by antioxidants. Moreover, the results suggested that KBrO3 induced apoptosis without mediating endogenous 4-HNE in ARPE-19 cells.

KBrO3 is an established oxidant (18), although to the best of our knowledge, its effects on ARPE-19 cells have not been previously reported. Oxidants, including H2O2 and t-BHP, have been reported to produce ROS in cells (12,13,16). In the present study, KBrO3 significantly induced ROS production and apoptosis in ARPE-19 cells in a concentration-depended manner. RPE cells may digest photoreceptor POSs (10); however, oxidative stress reduces the digestive rate of POS (10), and higher POS results in lipofuscin formation, one of the major alterations associated with AMD (11). Therefore, the present study indicated that KBrO3 may be useful to establish a cell model of AMD in ARPE-19 cells.

Similar to other oxidants, in the present study, KBrO3 significantly increased cytotoxicity in ARPE-19 cells compared with the control group. Moreover, compared with the control group, KBrO3 inhibited ARPE-19 cell viability in a dose-dependent manner at the tested concentrations (25, 50 and 75 µM). At 75 µM, KBrO3 induced ARPE-19 cell apoptosis to a similar level to ARPE-19 cell apoptosis induced by 30 µM 4-HNE. Also, 4-HNE produced in the RPE during high oxygen consumption is associated with AMD (33). Therefore, 4-HNE was used as a reference in the present study. Treatment with at higher concentrations (30 µM) causes cytotoxicity and ROS formation effect in cells compared with that in KBrO3 (25 µM) (34). In addition, t-BHP has been reported to induce cell apoptosis at doses >250 µM (14). In RPE cells, exposure to 1 mM H2O2 for 24 h markedly inhibited viability compared with the control group (12,13). Therefore, the results of the present study suggested that KBrO3 might be more potent than the aforementioned oxidants in ARPE-19 cells.

It has been reported that KBrO3-induced apoptosis is mediated via endogenous 4-HNE in ARPE-19 cells (33). 4-HNE is effective as an agonist for the nicotinic receptor GPR109A (28), thus 293 cells, which lack GPR109A expression (29), were used in the present study. In 293 cells, compared with the control group, KBrO3 significantly induced apoptosis and cytotoxicity, but 4-HNE did not display these effects. Therefore, the results suggested that endogenous 4-HNE was not involved in KBrO3-induced apoptosis.

The present study demonstrated that KBrO3 induced oxidative stress, which promoted cytotoxicity and apoptosis in ARPE-19 cells, but KBrO3-mediated effects were reversed by treatment with the antioxidants tiron and phloroglucinol. Tiron is a mitochondria-targeted antioxidant (35) and phloroglucinol protects RPE (36). The mediation of ROS in KBrO3-treated ARPE cells was also increased in a concentration-dependent manner. Moreover, KBrO3 treatment significantly inhibited antioxidant enzyme activities, including SOD, GSHPx in ARPE-19 cells, suggesting that KBrO3-induced oxidative stress in ARPE-19 cells. This was consistent with previous studies using hepatic (18) and cardiac (19) cells.

CNV occurs during the ‘wet’ phase of AMD, which occurs during the late stages of the disease (37). Hyperglycemia-induced oxidative stress may induce activation of STAT3-regulated VEGF expression in RPE cells (38). In the present study, the results suggested that KBrO3 activated STAT3 to promote VEGF expression in ARPE-19 cells. The results indicated that oxidative stress mediated this effect as KBrO3-induced effects were reversed by antioxidants. Moreover, the STAT3 inhibitor stattic significantly attenuated KBrO3-induced upregulation of VEGF mRNA expression levels at a dose that failed to inhibit KBrO3-induced ROS production. Stattic is an established STAT3 inhibitor that is pharmacologically applied in vitro and in vivo (39). Collectively, the results indicated that KBrO3-induced oxidative stress may activate STAT3 to promote VEGF expression in ARPE-19 cells.

KBrO3 is easily prepared in solution for both in vitro and in vivo investigations (40). As an oxidative stress inducer in cells, KBrO3 displays improved stability compared with H2O2 (41). However, the application of KBrO3 in animals may result in damage to the kidney when injury is initiated (42), whereas the ROS scavenger rutin may inhibit KBrO3 (43). Therefore, systemic application of KBrO3 to induce AMD may be difficult to achieve in animals. However, local application of KBrO3 in the eye may serve as an alternative method to prepare AMD model animals, but further investigation is required.

In conclusion, the present study identified a novel method to establish an AMD cell model using KBrO3 as an oxidant in ARPE-19 cells. KBrO3 treatment may serve as a simple and useful method to induce AMD cell models in the future.

Acknowledgements

The authors would like to thank laboratory assistants of Chi-Mei Medical Center, Ms. Y.L. Yen and Ms. Y.P. Lin, for their kind help with the experiments.

Funding

No funding was received.

Availability of data and materials

The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Authors' contributions

SCK and JTC designed the study and analyzed the data. YL assisted with performing the experiments and prepared the manuscript. JTC supervised the project. CCH analyzed the data and prepared the manuscript. KCC analyzed the data and revised the manuscript. HHL contributed to the study design and performed the VEGF assay. All authors read and approved the final manuscript.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References