Dietary Alcohol Consumption Elicits Corneal Toxicity Through the Generation of Cellular Oxidative Stress

Abstract

Purpose: Clinical data suggest that alcohol use is associated with the development of signs and symptoms of dry eye disease. However, preclinical data investigating ocular toxicity after dietary alcohol consumption are lacking. In this study, we investigated the effects of alcohol on the ocular surface, in human corneal epithelial cells (HCE-T) in vitro and in C57BL/6JRj mice in vivo. Methods: HCE-T were exposed to clinically relevant doses of ethanol. To determine the effects of dietary alcohol consumption in vivo, wild-type mice were administered the Lieber-DeCarli liquid diet (5% vol/vol ethanol or isocaloric control) for 10 days ad libitum. Corneal fluorescein staining was performed to assess ocular surface damage. Histopathological and gene expression studies were performed on cornea and lacrimal gland tissue. Results: Sublethal doses of ethanol (0.01%-0.5%) resulted in a dose-dependent increase of cellular oxidative stress in corneal epithelial cells and a significant increase in NFE2L2 and downstream antioxidant gene expression, as well as an increase in NFκB signaling; short-term exposure (0.5%, 4 h) triggered significant corneal epithelial cell barrier breakdown. Exposure to the alcohol-containing diet caused a 3-fold increase in corneal fluorescein staining, with no effect on tear volumes. Corneal thickness was significantly reduced in the alcohol diet group, and corneal tissue revealed dysregulated antioxidant and NFκB signaling. Our data provide the first published evidence that alcohol exposure causes ocular toxicity in mice. Conclusions: Our results are consistent with clinical studies linking past alcohol consumption to signs of ocular surface disease.

Keywords: alcohol; conjunctivitis; cornea; dry eye; ocular surface disease; oxidative stress.

FIG. 1.

Sublethal ethanol exposure induces ROS and activates Nrf2-mediated endogenous antioxidant signaling. (A) Acute ethanol exposure of concentrations up to 1 μM did not result in increased LDH release from HCE-T. In contrast, 5 μM alcohol caused significant increases in LDH release from HCE-T (n = 3). Data are shown as mean ± SEM from 3 separate experiments each derived 8 technical replicates. Data were analyzed by 2-way ANOVA with Dunnett's multiple comparisons test. (B) MTT assay revealed no significant cytotoxicity of HCE-T when exposed up to 5% ethanol vol/vol for 48 h (n = 3). Data are shown as mean ± SEM from 3 separate experiments each derived 8 technical replicates. Data were analyzed by 2-way ANOVA with Dunnett's multiple comparisons test. (C) Ethanol caused a dose-dependent generation of ROS and oxidative stress in HCE-T (n = 8). Data are shown from a representative experiment as mean ± SD from 8 replicates. Data were analyzed by 2-way ANOVA with Holm-Šídák multiple comparisons test. (D) HCE-T were exposed to 0.5% ethanol for 2, 4, 6, or 12 h, which cause a time-dependent increase in NFE2L2 mRNA levels that were different to untreated control at 2, 4, and 6 h (n = 5). Data are shown as mean ± SEM from 5 separate experiments each derived from 3 technical replicates. Data were analyzed by 2-way ANOVA with Dunnett's multiple comparisons test. (E–J) HCE-T were treated with 0.5% ethanol for 6 h, and mRNA levels of (E) CAT, (F) HO-1, (G) SOD1, (H) SOD2, (I) NOX2, and (J) NOX3. Expression levels of CAT, HO-1, and SOD2 were significantly increased by ethanol exposure, and NOX3 levels were significantly decreased (n = 3–5). qPCR data are shown as mean ± SEM derived from 3 separate experiments with 3 technical replicates each. Data were analyzed by Student's t-test. *P < 0.05, **P < 0.01, ***P < 0.001. ANOVA, analysis of variance; CAT, catalase; HCE-T, human corneal epithelial cells; HO-1, heme oxygenase-1; LDH, lactate dehydrogenase; mRNA, messenger RNA; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; NOX2, NADPH oxidase 2; NOX3, NADPH oxidase 3; Nrf2, nuclear factor erythroid 2-related factor 2; qPCR, quantitative polymerase chain reaction; ROS, reactive oxygen species; SD, standard deviation; SEM, standard error of the mean; SOD1, superoxide dismutase 1; SOD2, superoxide dismutase 2.

FIG. 2.

Gene and protein expression of canonical NF-κB signaling components is increased in ethanol-treated HCE-T. HCE-T were treated with 0.5% ethanol for 6 h, and RNA and protein were extracted for qPCR and western blot analyses, respectively. (A) qPCR revealed significant upregulation of (A) NFKB1 and (B) RELA in ethanol-treated HCE-T (n = 5). (C) Nuclear protein expression of RELA was also significantly reduced (n = 3). (D) Representative western blot bands for RELA and endogenous nuclear control, Lamin B1. Data are shown as mean ± SEM, with each circle representing the mean of a separate experiment. Data were analyzed by Student's t-test. *P < 0.05.

FIG. 3.

Secretion of pro-angiogenic cytokines is significantly increased in ethanol-treated HCE-T. Multiplex analysis of cytokine secretion from HCE-T revealed a significant increase in 4 pro-angiogenic cytokines: (A) IL-8, (B) MCP-1, (C) PDGF-AA, and (D) VEGF (n = 3). Data shown as mean ± SEM with each symbol representing a separate experiment with duplicate quantification. *P < 0.05, **P < 0.01, ***P < 0.001.

FIG. 4.

Ethanol exposure causes barrier dysfunction due to tight junction protein network disorganization in HCE-T. (A–C) Stratified HCE-T in transwells were treated with 0.5% ethanol for 4 h, then permeability through the cells and TEER were measured. (A) The cumulative amount of low permeability standard, 6-CF, in the receiver chamber over time was increased in ethanol-treated cells compared to control. (B) Similarly, the cumulative amount of high permeability standard, RhoB, in the receiver chamber increased significantly over time in ethanol-treated cells compared to control. Data were analyzed by nonparametric Mann–Whitney test. (C) The apparent permeability of 6-CF and RhoB through stratified HCE-T was significantly higher in ethanol-treated cells (n = 3). (D) Representative images of ZO-1 and occludin immunostaining in control and ethanol-treated HCE-T. (E) The TiJOR for ZO-1 was significantly disrupted in ethanol-treated cells compared to control (n = 3). (F) ZO-1 immunoreactivity was significantly lower in ethanol-treated cells, further indicative of disorganization of tight junctions (n = 3). (G) Similarly, occludin TiJOR was significantly decreased (n = 3). (H) Occludin immunoreactivity was significantly lower compared with control cells, further supporting the deleterious effect of ethanol on tight junction organization. Data are shown as mean ± SEM from 3 biological replicates, each derived from quantification of five images. Statistical significance was assessed by Student's t-test. *P < 0.05, **P < 0.01. 6-CF, 6-carboxyfluorescein; RhoB, rhodamine B; TEER, transepithelial electrical resistance; TiJOR, tight junction organization rate.

FIG. 5.

Alcohol consumption elicits ocular surface damage in vivo. (A) Graphical depiction of the experimental design. Mice were acclimated to the Lieber–DeCarli liquid diet for 5 days, before receiving either alcohol (5% vol/vol) or control isocaloric (n = 10 mice per group). Tear volumes were measured at baseline (0 day) and 10 days. Corneal fluorescein was measured at 10 days, then mice were euthanized for downstream tissue collection, histopathological assessment, and molecular analyses. (B) Tear volumes were not significantly affected by alcohol and increased similarly in both control- and alcohol-fed mice. Data were analyzed by 2-way mixed effects analysis (P < 0.001 for experimental group, P < 0.001 for time, P = 0.86 for interaction between the variables), followed by Holm-Šídák multiple comparisons test. (C) Intensity of corneal fluorescein staining on the cornea increased significantly in alcohol-fed mice compared with control mice (P < 0.001, n = 18–20). Data are presented as mean ± SEM with individual data points representing the measurement from a single eye. Data were analyzed by Student's t-test. (D) Representative images of corneal fluorescein staining from 3 different animals per group acquired using an epifluorescent microscope (left) or Heidelberg Spectralis^® (right). Images show significantly more severe corneal fluorescein staining in alcohol- versus control liquid diet-fed animals. ***P < 0.001.

FIG. 6.

Corneal expression of several antioxidant and Nf-κB genes was significantly decreased in alcohol-receiving mice. Gene expression of the following redox (n = 4 per group) and NF-κB (n = 8 per group) signaling proteins was quantified by qPCR from the corneas of control- and alcohol-fed mice: (A) Nfe2l, (B) Ho-1, (C) Cat, (D) Sod1, (E) Nox1, (F) Nox2, (G) Nox4, (H) Sod2, (I) Nfkb1, (J) RelA, (K) Nfkb2, (L) RelB. Data are shown as mean ± SEM, with each data point derived from a single cornea measured in triplicates. Data were analyzed by Student's t-test. *P < 0.05, **P < 0.01, ***P < 0.001.

FIG. 7.

Lack of histopathological abnormalities, but dysregulation of antioxidant gene expression is suggestive of oxidative stress in the lacrimal gland. (A) Intraorbital lacrimal glands were histopathologically normal, without presence of immune cell infiltration (n = 6 animals). (B) Similarly, no abnormalities were identified in extraorbital lacrimal glands (n = 9–10 animals). Data are shown as median ± interquartile range and were analyzed by Mann–Whitney test. Representative images from control- and alcohol-fed animals are shown. (C) qPCR analysis revealed a significant upregulation of Sod1 gene expression and significant reductions of Sod2 and Nox1 in the extraorbital lacrimal glands of alcohol versus control animals (n = 4 animals). Data are shown as mean ± SEM and were analyzed by Student's t-test. *P < 0.05, **P < 0.01.