Quercetin and Resveratrol Decrease the Inflammatory and Oxidative Responses in Human Ocular Surface Epithelial Cells

Abstract

Purpose.: To determine the anti-inflammatory and antioxidant effects of quercetin (QCT) and/or resveratrol (RES) on human conjunctival (IOBA-NHC) and corneal (HCE) epithelial cell lines.

Methods.: IOBA-NHC and HCE cells were treated with QCT (0.5–25 μM), RES (0.5–50 μM) and a low-dose mixture of QCT (0.5 μM) and RES (5 μM) (QCT+RES) and stimulated with either TNF-α or ultraviolet (UV)-B radiation. Cytokine production (IL-6, IL-8, IP-10, and VEGF) was analyzed by an immune bead-based array, and intracellular reactive oxygen species (ROS) production was determined by a H₂DCF-DA dye assay.

Results.: Stimulation of IOBA-NHC and HCE cells with TNF-α induced an increase of IL-6, IL-8, and IP-10 secretion in both cell lines. Quercetin and RES decreased IL-6 and IP-10 secretion in a dose-dependent manner in both cell lines. Interleukin-8 secretion was also inhibited in a dose-dependent manner by QCT in HCE, but only at 20 and 25 μM QCT and 50 μM RES in IOBA-NHC and at 50 μM RES in HCE. QCT+RES decreased IL-6 and IL-8 secretion (P < 0.01 and P < 0.05, respectively) in IOBA-NHC cells. Ultraviolet-B induced a significant increase of ROS in both cell lines (P < 0.01 and P < 0.001 for IOBA-NHC and HCE cells, respectively), which was significantly decreased in a dose-dependent manner by QCT and RES in HCE cells. Reactive oxygen species production in IOBA-NHC cells was inhibited (P < 0.05) by 50 μM RES.

Conclusions.: Quercetin and RES have anti-inflammatory and antioxidant effects on IOBA-NHC and HCE cells. These in vitro data suggest that both polyphenols may have a therapeutic potential in the treatment of inflammatory ocular surface diseases.

Inflammatory ocular surface diseases, such as dry eye disease (DED), severe allergies, or immune-based cicatrizing conjunctivitis, are an increasing health care problem due to their high prevalence¹ and capacity to affect patients' quality of life, work-related issues, and health care resources.^2,3 Anti-inflammatory and immunosuppressive compounds are considered the pharmacologic agents for controlling the inflammatory cascade underlying these diseases. Particularly, corticosteroids and cyclosporine A are the elective therapies for more severe forms of DED. Nevertheless, prolonged topical corticosteroid treatment produces ocular long-term side effects (i.e., glaucoma and cataract),^4,5 and cyclosporine A is not available in all countries as its use has been approved in only some of them. In addition, inflammation is not the sole mechanism involved in these processes. Oxidative stress also plays a role in these pathologies as an increase of oxidative stress markers has been evidenced in the conjunctiva of DED,⁶ as well as in patients suffering from conjunctivochalasis⁷ or ocular allergy.⁸ For these reasons there is an increasing interest in finding new topical therapeutic alternatives for the treatment of chronic inflammatory ocular surface diseases. 

In the last decade, there has been growing evidence of the potential health benefits of dietary-derived plant polyphenols, such as flavonoids and stilbenes, due to their biological properties as antioxidant and anti-inflammatory compounds.⁹ 

Quercetin (3,3′,4′,5,7-pentahydroxyflavone; QCT) is a flavonoid polyphenolic compound found in fruits and vegetables such as apples, onions, and berries. Several in vitro and in vivo studies support its effect as an anticancer, antiviral, cardioprotective, and neuroprotective compound.^10,11 In particular, QCT is well known for its antioxidant capacity, being one of the most effective free radical scavengers,¹² and for its anti-inflammatory activity.¹³ 

Trans-resveratrol (trans-3,4′,5-trihydroxystilbene; RES) is a nonflavonoid polyphenolic compound found in several dietary sources, such as grapes, mulberries, peanuts, and red wine. The popularity of RES comes from the so-called French paradox, an inverse correlation between high-fat diet and low incidence of coronary heart disease due to red wine consumption.¹⁴ Like QCT, it has widely demonstrated biological effects as a neuroprotective, cardioprotective, and anticarcinogenic^15–18 as well as an anti-inflammatory and antioxidant compound.^19,20 

Currently, interest in possible health benefits of naturally occurring polyphenols has increased due to their potential effect as anti-inflammatory and antioxidant compounds on several diseases, such as cystitis and hypertension.^21,22 Some studies have addressed the effect of RES in certain ocular inflammatory pathologies, as in the endotoxin-induced uveitis mouse model.²³ However, to our knowledge, there are no studies addressing the effects of QCT and RES on ocular surface diseases. Therefore, the aim of the present study was to determine the anti-inflammatory and antioxidant effects of QCT, RES, and the combination of both compounds (QCT+RES) on ocular surface epithelial cells under two different stimuli: (1) tumor necrosis factor (TNF)-α–induced inflammation²⁴ and (2) ultraviolet (UV)-B light-induced oxidative stress.²⁵ Cytotoxicity of polyphenols and their effects on cytokine/chemokine secretion (interleukin [IL]-6, IL-8, interferon γ-induced protein [IP]-10, and vascular endothelial growth factor [VEGF]) and intracellular reactive oxygen species (ROS) production were measured in two different ocular surface epithelial cell lines. 

Materials and Methods

Materials

Reagents were purchased from the following suppliers. Dulbecco's modified Eagle's medium/Nutrient Mixture F-12 (DMEM/F-12), alamarBlue cell viability assay, and 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) were from Invitrogen (Inchinnan, UK); plastic culture dishes were from Nunc (Roskilde, Denmark); ethanol (EtOH) and D-glucose were from Panreac (Barcelona, Spain); the cytokine TNF-α was from PeproTech EC (London, UK); RES, QCT, DMEM (culture medium without sodium bicarbonate [NaHCO₃], sodium pyruvate, and phenol red), 2′,7′-dichlorodihydrofluorescein diacetate (H₂DCF-DA), L-glutamine, fetal bovine serum (FBS), cholera toxin, human epithelial growth factor (EGF), bovine insulin, penicillin, streptomycin, fungizone, hydrocortisone, and dimethyl sulfoxide (DMSO) were purchased from Sigma-Aldrich Corp. (St. Louis, MO, USA); and the bicinchoninic acid (BCA) assay was from Thermo Fisher Scientific (Rockford, IL, USA). 

Cell Lines and Culture

Two different ocular surface epithelial cell lines, the IOBA-NHC and the HCE, were used for these experiments. 

The IOBA-NHC (normal human conjunctiva) cell line is a nontransfected, spontaneously immortalized epithelial cell line derived from normal human conjunctiva.²⁶ IOBA-NHC cells were cultured in DMEM/F-12 L-glutamine supplemented with 10% FBS, 0.1 μg/mL cholera toxin, 2 ng/mL EGF, 1 μg/mL bovine insulin, 5000 U/mL penicillin, 5 mg/mL streptomycin, 2.5 μg/mL fungizone, and 0.5 μg/mL hydrocortisone. It was used from passage 62 to 72. 

The human corneal epithelium (HCE) cell line is an SV40-immortalized human corneal epithelial cell line²⁷ kindly gifted by Arto Urti (University of Helsinki, Finland). The HCE cells were cultured in DMEM/F-12 L-glutamine supplemented with 15% FBS, 0.5% DMSO, 0.1 μg/mL cholera toxin, 10 ng/mL EGF, 5 μg/mL insulin, 100 U/mL penicillin, and 0.1 mg/mL streptomycin. It was used from passage 45 to 55. 

Both cell lines were cultured at 37°C in a 5% CO₂ 95% air atmosphere. Media were changed every other day, and daily observations were made by phase-contrast microscopy. 

It has been shown that cell culture media play a key role when polyphenols are tested in vitro.²⁸ On the one hand, some culture medium components, such as sodium bicarbonate (NaHCO₃), can degrade and decrease polyphenol content.^29,30 On the other hand, cell culture media usually contain many potential antioxidant compounds, such as pyruvate and phenol red, which can interfere with antioxidant capacity of tested compounds.^31–34 In order to avoid these artifacts in cell culture medium, all experiments were carried out in a DMEM culture medium without NaHCO₃, sodium pyruvate, and phenol red, and supplemented with 3.15 g/L D-glucose, 2 mM L-glutamine, and 25 mM HEPES. 

Preparation of Polyphenol Solutions

Fresh stock solutions of QCT and RES in EtOH were prepared for each experiment, after which serial dilutions were carried out to achieve final concentrations, ranging from 0.5 to 25 μM QCT and 0.5 to 300 μM RES. A combination of 0.5 μM QCT and 5 μM RES (QCT+RES) was prepared by mixing QCT and RES solutions in order to reach the final concentrations for each experiment. All solutions were prepared in such a way that vehicle (EtOH) had a final concentration of 0.5% (nontoxic for either cell line, data not shown) in all samples when polyphenols were added to each well. 

Cytotoxicity Assay

The toxicity of QCT and RES on epithelial cells was assessed by the alamarBlue test. IOBA-NHC and HCE cells were seeded in 96-well plates and grown to 90% of preconfluence. They were then maintained for 24 hours in serum-free, nonsupplemented medium. At that time, media were removed and cells were treated with different concentrations of QCT (1, 5, 10, 15, 20, and 25 μM) or RES (1, 5, 10, 25, 50, 100, 150, and 300 μM) and incubated for 24 hours at 37°C. Control cells were treated with vehicle. Following incubation, supernatants were discarded and 10% alamarBlue, prepared in supplemented DMEM/F-12 cultured medium, was added. Cells were incubated for 4 hours at 37°C. Finally, medium from each sample was collected, and fluorescence was measured at 560 nm_ex/590 nm_em by UV/Vis spectrophotometry (SpectraMax M5; Molecular Devices, Sunnyvale, CA, USA). Benzalkonium chloride (0.005%) was used as positive control (data not shown). Three independent experiments were performed, and measurements were from eight replicates for each condition studied. 

Cell Cytokine Stimulation and Polyphenol Treatments

IOBA-NHC and HCE cells were seeded in 24-well plates and grown to 90% of preconfluence. They were then maintained in serum-free, nonsupplemented medium for 24 hours. At that time, media were removed and cells were pretreated with QCT (0.5, 1, 5, 10, 15, 20, and 25 μM), RES (0.5, 1, 5, 10, 25, and 50 μM), QCT+RES (0.5 μM QCT + 5 μM RES) or vehicle for 2 hours at 37°C. After that, pretreatment agents were removed; cells were stimulated with 25 ng/mL TNF-α in the presence of QCT, RES, QCT+RES, or vehicle (0.5% EtOH) and incubated for 24 hours. Unstimulated cells, treated with polyphenols but without TNF-α, were used as control. After the treatment period, the conditioned media were collected and centrifuged at 59g for 5 minutes. Supernatants and plates with adherent cells were stored at −80°C until use. Three independent experiments were performed in duplicate. 

Measurement of Cytokine/Chemokine Secretion

Cytokine/chemokine secretion was assessed by a multiplex bead-based array using Luminex x-MAP multiplexing bead technology as previously described.²⁴ Interleukin-6, IL-8, IP-10, and VEGF levels were determined in cell supernatants with a commercial Milliplex 4-plex human cytokine/chemokine immunobead-based assay (HCYTO; Millipore, Watford, UK), according to the manufacturer's instructions. Briefly, 25 μL cell supernatant from each sample was incubated in 96-well plates with antibody-immobilized beads overnight at 4°C. Then, beads were washed and incubated with biotinylated cytokine/chemokine antibody solution for 1 hour at room temperature, followed by incubation with streptavidin-phycoerythrin for 30 minutes at room temperature. Finally, beads were washed and read on a Luminex 100-IS instrument (Luminex Corporation, Austin, TX, USA). Standard curves of known concentrations of recombinant human cytokines/chemokines were used to convert fluorescent units to cytokine/chemokine concentration units (pg/mL). The minimum detectable level for each cytokine/chemokine, based on manufacturer specifications, was 0.3 pg/mL for IL-6, 0.2 pg/mL for IL-8, 1.2 pg/mL for IP-10, and 5.8 pg/mL for VEGF. When a cytokine level was not detectable, the minimum detectable level was used in the analysis. Data were analyzed with the BeadView Software (Upstate, UK) and normalized to total protein content for each sample. Total protein content was assessed using the BCA protein assay³⁵ in adherent cells according to manufacturer's instructions. 

Measurement of Reactive Oxygen Species Production Induced by UV-B Radiation

The generation of intracellular ROS by UV-B exposure of epithelial cells was assessed using the H₂DCF-DA dye. This compound is a nonfluorescent dye that passively diffuses into cells, where it is cleaved and deacetylated to H₂DCF by intracellular esterases. Nonfluorescent H₂DCF is rapidly oxidized to fluorescent DCF by intracellular ROS. 

IOBA-NHC and HCE cells were cultured in supplemented medium in 24-well plates until 90% of preconfluence. They were then maintained for 24 hours in serum-free, nonsupplemented medium. After that, media were removed and cells were pretreated with QCT (0.5, 1, 5, 10, and 25 μM), RES (0.5, 1, 5, 10, and 50 μM), QCT+RES (0.5 μM QCT + 5 μM RES), or vehicle for 1 hour at 37°C. At that point, supernatants were discarded, and cells were loaded with H₂DCF-DA, adding 500 μL 10 μM H₂DCF-DA solution, and incubated for 30 minutes. H₂DCF-DA solution was then aspirated; cells were treated with QCT, RES, QCT+RES, or vehicle (0.5% EtOH) at the same concentrations as used before and exposed to 8-W UV-B lamps (with an excitation peak of 302 nm) located 3 cm below cells. At that distance the UV-B radiation power density was 7.15 mW/cm², according to the manufacturer (Bio-Rad, Inc., Hercules, CA, USA). Cells were irradiated for 15 seconds from the bottom of the well plate to avoid UV-B absorption by polyphenols in the culture media. After 15 seconds, UV-B radiant exposure was 107.25 mJ/cm², as calculated with the formula H = E · t, where H is the radiant exposure (J/cm²), E is the irradiance (W/cm²), and t is the exposure time (seconds). Control cells were not irradiated. After UV-B exposure, cells were cultured for 1 hour, and then intracellular fluorescence intensity was measured at 488 nm_ex /522 nm_em using the SpectraMax M5 UV/Vis spectrophotometer. Fluorescence data from each sample were normalized to the corresponding total protein content, determined in adherent cells by the BCA protein assay kit after measuring intracellular fluorescence intensity. Three different experiments were carried out, and samples were performed in duplicate. 

Statistical Analysis

All data were expressed as mean ± standard error of the mean (SEM). Statistics were analyzed using the SPSS software package (SPSS version 15.0 for Windows; SPSS, Inc., Chicago, IL, USA). Homogeneity of variances was analyzed using Levene's test. The t-test or t-test with Welch correction was used for comparison of unstimulated cells versus stimulated cells, and one-way analysis of variance (ANOVA) with Dunnett's post hoc test or Games-Howell test was used for intergroup comparisons. Two-sided P values equal to or less than 0.05 were considered statistically significant. 

Results

Cytotoxicity of Polyphenols

In order to test the cytotoxicity of both polyphenols, cells were exposed to different concentrations of QCT and RES for 24 hours. Figure 1 shows the cytotoxicity of QCT and RES in IOBA-NHC (Figs. 2A, 2B) and HCE (Figs. 2C, 2D) cell lines. Quercetin did not decrease cell viability in either conjunctival or corneal epithelial cells at any concentration tested. Regarding RES concentrations tested, only 300 μM RES provoked a significant decrease of cell viability in both cell lines, with this effect higher in HCE cells than in IOBA-NHC cells. 

Figure 1

 

Cytotoxicity effect of quercetin (QCT) and resveratrol (RES) on IOBA-NHC and HCE cells. Cells were treated with QCT (1, 5, 10, 15, 20, and 25 μM), RES (1, 5, 10, 25, 50, 100, 150, and 300 μM) or vehicle (0.5% ethanol) and incubated for 24 hours. Cytotoxicity was determined using the alamarBlue test. None of the QCT concentrations tested was toxic for either IOBA-NHC (A) or HCE (C) cells, whereas only 300 μM RES was toxic for both IOBA-NHC (B) and HCE (D) cells. Data are presented as relative fluorescence units (RFU) of three independent experiments ± SEM. *P < 0.05, ***P < 0.001, compared to cells treated with vehicle.

Figure 2

 

Effect of quercetin (QCT) and resveratrol (RES) on TNF-α-induced cytokine release by IOBA-NHC. Cells were pretreated with QCT (0.5, 1, 5, 10, 15, 20, and 25 μM), RES (0.5, 1, 5, 10, 25, and 50 μM), or vehicle (0.5% ethanol) for 2 hours. Subsequently, cells were stimulated with 25 ng/mL TNF-α and treated with QCT, RES, or vehicle and incubated for 24 hours (black squares). Unstimulated cells, treated with polyphenols but without TNF-α, were used as control (white circles). Supernatants were further collected and analyzed for IL-6, IL-8, IP-10, and VEGF levels using x-MAP multiplexing bead technology. QCT decreased TNF-α-stimulated IL-6, IL-8, and IP-10 secretion at 15 μM (A), 20 μM (B), and 10 μM (C), respectively. RES decreased TNF-α-stimulated IL-6, IL-8, and IP-10 secretion at 25 μM (E), 50 μM (F), and 10 μM (G), respectively. TNF-α did not stimulate VEGF secretion by IOBA-NHC cells (D, H). Data are presented as picograms (pg) of cytokine normalized to micrograms of total protein (μg protein) from three independent experiments ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, compared to control cells; ⁺P < 0.05, ⁺⁺P < 0.01, ⁺⁺⁺P < 0.001, compared to vehicle-treated stimulated cells.

Effect of Polyphenols on Cytokine/Chemokine Secretion Induced by TNF-α

In order to investigate the anti-inflammatory properties of nontoxic concentrations of QCT (0.5–25 μM) and RES (0.5–50 μM) on conjunctival and corneal epithelial cells, their effect on the cytokine/chemokine secretion induced by TNF-α was determined. 

Figure 2 shows the effect of polyphenols on IL-6, IL-8, IP-10, and VEGF secretion by IOBA-NHC cells. Quercetin significantly decreased IL-6, IL-8, and IP-10 secretion stimulated by TNF-α from 15, 20 and 10 μM, respectively (Figs. 2A–2C), whereas RES decreased TNF-α-stimulated IL-6, IL-8, and IP-10 secretion from 25, 50, and 10 μM, respectively (Figs. 2E–2G). 

Figure 3 shows the effect of QCT and RES on IL-6, IL-8, IP-10, and VEGF secretion by HCE cells. In these cells, QCT significantly decreased IL-6-, IL-8-, and IP-10-stimulated secretion from 1, 5, and 1 μM, respectively (Figs. 3A–3C), whereas RES decreased IL-6-, IL-8-, and IP-10-stimulated secretion from 10, 25, and 0.5 μM, respectively (Figs. 3E–3G). 

Figure 3

 

Effect of quercetin (QCT) and resveratrol (RES) on TNF-α-induced cytokine release by HCE. Cells were pretreated with QCT (0.5, 1, 5, 10, 15, 20, and 25 μM), RES (0.5, 1, 5, 10, 25, and 50 μM), or vehicle (0.5% ethanol) for 2 hours. Subsequently, cells were stimulated with 25 ng/mL TNF-α and treated with QCT, RES, or vehicle and incubated for 24 hours (black squares). Unstimulated cells, treated with polyphenols but without TNF-α, were used as control (white circles). Supernatants were further collected and analyzed for IL-6, IL-8, IP-10, and VEGF levels using x-MAP multiplexing bead technology. QCT decreased TNF-α-stimulated IL-6, IL-8, and IP-10 secretion at 1 μM (A), 5 μM (B), and 1 μM (C), respectively. RES decreased TNF-α-stimulated IL-6, IL-8, and IP-10 secretion at 10 μM (E), 50 μM (F), and 0.5 μM (G), respectively. TNF-α did not stimulate VEGF secretion by HCE cells (D, H). Data are presented as pictograms (pg) of cytokine normalized to micrograms of total protein (μg protein) from three independent experiments ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, compared to control cells; ⁺P < 0.05, ⁺⁺P < 0.01, ⁺⁺⁺P < 0.001, compared to vehicle-treated stimulated cells.

A concentration of 25 ng/mL TNF-α did not stimulate VEGF secretion by either IOBA-NHC or HCE cells. 

The anti-inflammatory effect of both compounds in combination was also tested (Fig. 4). A combination of QCT+RES (0.5 μM QCT and 5 μM RES) was chosen based on dose–response curves because both concentrations separately did not have any significant effect on cytokine/chemokine release in both cell lines. 

Figure 4

 

Effect of the combination of quercetin (QCT) and resveratrol (RES) on TNF-α-induced cytokine release by IOBA-NHC and HCE cells. Both cell lines were pretreated with 0.5 μM QCT, 5 μM RES, 0.5 μM QCT + 5 μM RES, or vehicle (0.5% ethanol) for 2 hours. Subsequently, cells were stimulated with 25 ng/mL TNF-α and treated with QCT, RES, QCT+RES, or vehicle and incubated for 24 hours. Unstimulated cells, treated with polyphenols but without TNF-α, were used as control. Supernatants were further collected and analyzed for IL-6, IL-8, IP-10, and VEGF levels using x-MAP multiplexing bead technology. For IOBA-NHC cells, QCT+RES decreased TNF-α-stimulated IL-6 secretion more than QCT and RES separately (A). There were no significant differences between the effect of QCT+RES and both compounds separately on both TNF-α-stimulated IL-8 and IP-10 secretion by IOBA-NHC cells (B, C). Although TNF-α did not stimulate VEGF secretion, QCT and QCT+RES significantly decreased VEGF levels in the presence of TNF-α in IOBA-NHC cells (D). For HCE cells, there were also no significant differences between QCT+RES and both compounds separately for any cytokine level analyzed. Data are presented as picograms (pg) of cytokine normalized to micrograms of total protein (μg protein) from three independent experiments ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, compared to control cells; ⁺P < 0.05, ⁺⁺P < 0.01, compared to vehicle-treated stimulated cells.

Regarding IOBA-NHC cells, there was an inhibitory effect of the combination of both compounds QCT+RES on TNF-α-stimulated IL-6 secretion, stronger than that observed for 0.5 μM QCT and 5 μM RES separately (Fig. 4A). The mix QCT+RES also significantly decreased TNF-α-stimulated IL-8 production, but this decrease was similar to that found in cells treated with 0.5 μM QCT (Fig. 4B). No significant differences in TNF-α-stimulated IP-10 levels were found when IOBA-NHC cells were treated with QCT+RES with respect to stimulated cells (Fig. 4C). Referring to HCE cells, the combination QCT+RES did not have a significant effect on TNF-α-induced IL-6, IL-8, IP-10, and VEGF secretion (Figs. 4E–4H). 

Reactive Oxygen Species Production Induced by Ultraviolet B Radiation

In order to test the antioxidant effect of both polyphenols separately and in combination, UV-B-induced intracellular ROS production was determined in both epithelial cell lines (Fig. 5). None of the concentrations of QCT tested significantly decreased UV-B-induced intracellular ROS production in IOBA-NHC cells (Fig. 5A). However, it should be noted that from 1 to 25 μM QCT there were not significant differences in ROS levels between unexposed cells and UV-B exposed cells, suggesting prevention of UV-B-induced ROS production by QCT. When UV-B-exposed IOBA-NHC cells were treated with RES, only 50 μM RES significantly decreased intracellular ROS production (Fig. 5B). Prevention of intracellular ROS production by cells was found at RES concentrations of 25 and 50 μM. The combination QCT+RES did not have a significant effect on UV-B-induced ROS production (Fig. 5C). However, no significant differences in ROS production were found when compared to baseline levels in unexposed cells, showing an antioxidant preventive effect. 

Figure 5

 

Effect of quercetin (QCT), resveratrol (RES) and QCT+RES on UV-B-induced intracellular reactive oxygen species (ROS) production. IOBA-NHC and HCE cells were pretreated with QCT (0.5, 1, 5, 10, 15, 20, and 25 μM), RES (0.5, 1, 5, 10, 25, and 50 μM), 0.5 μM QCT + 5 μM RES, or vehicle (0.5% ethanol) for 1 hour. After that, cells were loaded with 10 μM H₂DCF-DA solution for 30 minutes and subsequently treated with QCT, RES, QCT+RES, or vehicle and exposed to 107.25 mJ/cm² UV-B light (black squares). Control cells were not irradiated (white circles). After 1 hour of culture, intracellular fluorescence intensity was measured. QCT did not decrease UV-B-stimulated ROS production significantly (although ROS levels were similar to those in unexposed IOBA-NHC cells) (A), whereas 50 μM RES decreased significantly UV-B-induced ROS production by conjunctival epithelial cells (B). On the other hand, both QCT and RES decreased UV-B-induced ROS production by HCE cells at 0.5 and 25 μM, respectively (D, E). The combination of QCT+RES did not decrease ROS production significantly for either IOBA-NHC or HCE cells (C, F). Nevertheless, there were not significant differences between UV-B-exposed cells and unexposed cells (both treated with QCT+RES) for either IOBA-NHC or HCE cells. Data are presented as relative fluorescence units (RFU) normalized to micrograms of total protein (μg protein) from three independent experiments ± SEM. *P < 0.05, **P < 0.01, ***P < 0.001, compared to control cells; ⁺P < 0.05, ⁺⁺P < 0.01, ⁺⁺⁺P < 0.001, compared to vehicle-treated stimulated cells.

Regarding HCE cells, QCT significantly decreased intracellular ROS production from 0.5 μM and prevented ROS production at 25 μM (Fig. 5D) in UV-B-exposed cells, whereas RES decreased ROS levels starting from 25 μM, although RES did not show a preventive effect because there were significant differences between exposed and unexposed cells treated with RES (Fig. 5E). The combination QCT+RES did not have any effect on decreasing ROS production, but the combination was effective in preventing ROS production (no significant differences when compared to unexposed cells) in corneal epithelial cells (Fig. 5F). 

Discussion

In this work, we have reported the anti-inflammatory and antioxidant effects of two naturally occurring polyphenols, QCT and RES, on human ocular surface corneal and conjunctival epithelial cells in vitro. 

The results from cytotoxicity assay showed that QCT and RES were nontoxic for both epithelial cell lines up to 25 and 300 μM, respectively. Concentrations of QCT above 25 μM were not tested in this work due to the very low solubility of QCT in EtOH (2 mg/mL). 

Our cytotoxicity results agree with previous studies regarding polyphenol cytotoxicity to ocular surface epithelial cells. For example, Stoddard et al.³⁶ studied the antioxidant effect of some polyphenols in stratified human corneal limbal epithelial cells and found that QCT is nontoxic at 56.5 μM in these cells. Chen et al.³⁷ found that curcumin, a natural substance derived from the rhizome of the plant Curcuma longa, is nontoxic up to 30 μM for immortalized human corneal epithelial cells. Also Cavet et al.³⁸ found that the green tea polyphenol epigallocatechin gallate is not toxic up to 30 μM for a human corneal epithelial cell line. Taken together, these studies suggest that polyphenols in general, and QCT and RES in particular, are not toxic for ocular surface epithelial cell lines at concentrations up to 50 μM. 

In this study, QCT and RES have been demonstrated to have an anti-inflammatory effect on ocular surface epithelial cells, as they decreased in a dose-dependent manner IL-6, IL-8, and IP-10 secretion by cells stimulated by the proinflammatory cytokine TNF-α. All of these cytokines are involved in inflammatory ocular surface diseases, including DED. For example, IL-6 has been suggested as a possible biomarker for DED because it is increased in the conjunctival epithelium and tears of DED patients.^39–43 Moreover, IL-6 concentration in tears correlates with some clinical parameters in patients with DED, such as pain, tear film breakup time, Schirmer test, tear clearance, goblet cell density, tear lysozyme levels, and conjunctival staining.^42–45 Interleukin-8 is involved in ocular inflammation and angiogenesis in the conjunctiva and cornea,⁴⁶ and has been also identified in tears and conjunctiva of patients with DED.^42,45 High concentration of IP-10 is detected in tears of DED patients⁴⁵ and also in the conjunctiva and tears of Sjögren and non-Sjögren syndrome DED patients.⁴⁷ In addition, an increased IP-10 expression in both corneal and conjunctival epithelia has been described in mice exposed to desiccating stress.⁴⁸ All these secreted cytokines by conjunctival and corneal epithelial cells were stimulated by TNF-α, a major proinflammatory cytokine involved in DED.^41,45 The effect of QCT and RES was also studied in VEGF secretion, as it has been demonstrated that normal and inflamed ocular surface epithelial cells can secrete VEGF.^49,50 For example, VEGF levels are increased in tears of mild to moderate DED patients⁴⁵ and in conjunctival biopsies and tears of patients suffering vernal keratoconjunctivitis or atopic keratoconjunctivitis,⁵¹ playing a crucial role in the remodelling process of these severe allergic conjunctival disorders.⁵² However, in our study, TNF-α failed to induce its secretion in any of both epithelial cell lines. Maybe other proinflammatory stimuli in vivo, different from TNF-α, either alone or in combination with TNF-α, are responsible for the increased VEGF secretion by ocular surface epithelial cells observed in those studies. More studies about the effect of QCT and RES on VEGF secretion by cells are warranted. Nevertheless, results from this work have demonstrated that QCT and RES can modulate the inflammatory response of conjunctival and corneal epithelial cells in vitro, inhibiting cytokine secretion stimulated by the proinflammatory cytokine TNF-α. However, further studies are necessary to elucidate how QCT and RES modulate TNF-α action and cytokine production (for instance, if polyphenols act on the release of preformed mediators or interfere with their expression at the mRNA level) and the downstream mechanisms involved on the inflamed ocular surface epithelial cells. 

Although there are very few studies regarding the effects of natural compounds on the ocular surface, similar results to those described in this work have been reported for other polyphenols in ocular surface epithelial cells. For example Chen et al.³⁷ demonstrated that curcumin has an anti-inflammatory effect on HCE cells exposed to hyperosmolarity. They found that 5 μM curcumin can abolish the hyperosmolarity-induced production of cytokines such as IL-1β, IL-6, and TNF-α, as well as IL-1β mRNA and activation of p38. Therefore, they suggested that curcumin inhibits hyperosmolarity-induced increase of IL-1β production in HCE cells through inhibition of p38, which leads to nuclear factor κB (NF-κB) p65 inhibition by curcumin. Cavet et al.³⁸ also reported that the green tea polyphenol epigallocatechin gallate has an anti-inflammatory effect on HCE cells using two stimuli, IL-1β and hyperosmolarity with sucrose. The authors found that IL-1β increases the release of IL-6, IL-8, monocyte chemoattractant protein (MCP)-1, granulocyte colony-stimulating factor (G-CSF), granulocyte macrophage colony-stimulating factor (GM-CSF), whereas hyperosmolarity increases only IL-6 and MCP-1. The polyphenol epigallocatechin gallate decreases IL-1β- and hyperosmolarity-induced cytokine/chemokine secretion by HCE cells in a dose-dependent manner from 3 to 30 μM. In addition, this polyphenol inhibits the IL-1β- and hyperosmolarity-induced phosphorylation/activation of both p38 and c-jun N-terminal kinase (JNK) in HCE cells. Their results also indicated that inhibition of IL-1β-induced cytokine expression is, at least in part, mediated by inhibition of both activator protein (AP)-1 and NF-κB transcription activities by epigallocatechin gallate. 

Our work also demonstrated for the first time that QCT and RES protect conjunctival and corneal epithelial cells from UV-B radiation-induced oxidative stress in a dose-dependent manner. Oxidative stress can be defined as a persistent imbalance between the production of free radicals and the biological defense mechanisms that eliminate the stress, in favor of the free radical production (i.e., ROS). Oxidative stress is present in several diseases, such as rheumatoid arthritis, diabetes, and cancer,⁵³ in which inflammation also plays a key role. Whether it is a primary cause or merely a downstream consequence of the inflammatory process is still an open question. Nevertheless, it is well known that oxidative stress is implicated in the pathogenesis of inflammatory ocular surface disorders such as DED. For example, an increase of oxidative markers and ROS production has been found in a blink-suppressed mouse model of DED.⁵⁴ The biochemical study by Augustin et al.⁵⁵ confirmed an increase of oxidative markers in tears of patients with DED. Moreover, Cejkova et al.⁵⁶ found that there is a decrease in the expression of antioxidant enzymes in the conjunctival epithelium of patients suffering from Sjögren DED. In addition, Wakamatsu et al.⁶ have recently reported a correlation between oxidative stress markers and staining scores and the number of inflammatory cells in patients suffering from Sjögren syndrome DED, concluding that there may be a close relationship between oxidative stress and the inflammatory process in DED. 

Quercetin and RES have been shown to possess two different antioxidant effects because they reduce or prevent ROS production, depending on concentration tested. These two mechanisms represent different therapeutic effects that should be considered as either before or after the onset of the disease. Nevertheless, our results are in agreement with other studies regarding the effect of polyphenols as antioxidant compounds on the ocular surface. For example, Larrosa et al.⁵⁷ found that hydrocaffeic acid alone and a mixture of two compounds (hydrocaffeic acid plus p-coumaric acid) reduce oxidation damage in human conjunctival cells exposed to UV-B radiation. They also found that topical treatment with hydrocaffeic acid and hydrocaffeic acid plus p-coumaric acid reduces corneal and scleral DNA oxidation damage, xanthine oxidase activity, and malondialdehyde levels in the corneal tissues of rabbit eyes exposed to UV-B radiation. Similar results have been reported recently by Chen et al.,⁵⁸ who found that topical treatment with epigallocatechin gallate eye drops ameliorates corneal damage, as well as increasing superoxide dismutase and catalase, and reducing glutathione activity in rabbit corneas exposed to UV-B radiation compared to the UV-B-treated group. 

The present work has some limitations. The effect of the combination of QCT and RES was tested using only one combination of one concentration of each compound. Further studies are thus necessary to elucidate whether other combinations of QCT and RES concentrations different from the one used here could be more effective. In addition, while both epithelial cell lines used here have been shown to be good in vitro models and to behave similarly to their corresponding human tissues of origin,^24,26,27 we are aware that the cellular responses in vivo of the human ocular surface epithelia are much more complex. Also, although both TNF-α and UV-B exposure are well-known inflammatory and oxidant stimuli, these may not be the main activators in an in vivo model. Further in vivo studies are warranted to corroborate our in vitro findings, and our group is currently working in this topic. 

In conclusion, we have demonstrated that the two polyphenols QCT and RES decrease inflammation and oxidative stress on human conjunctiva and corneal epithelial cells. These in vitro results show the therapeutic potential of QCT and RES as promising candidates for the treatment of ocular surface inflammatory diseases such as DED. 

Acknowledgments

Disclosure: A. Abengózar-Vela, P; M. Calonge, Allergan, Inc. (C), P; M.E. Stern, Allergan, Inc. (E), P; M.J. González-García, P; A. Enríquez-De-Salamanca, P 

Supported by ayudas FPI-UVa, University of Valladolid, Spain (AA-V).