Interleukin-1β–Induced Disruption of Barrier Function in Cultured Human Corneal Epithelial Cells

Abstract

purpose. The barrier function of the corneal epithelium contributes to corneal homeostasis and is impaired by inflammation. Adherens junctions (AJs) and tight junctions (TJs) of the corneal epithelium are essential for cell adhesion and barrier function. We examined the effects of the proinflammatory cytokine interleukin (IL)-1β on AJs and TJs as well as on barrier function in simian virus 40-transformed human corneal epithelial (HCE) cells.

methods. Barrier function was evaluated by measurement of transepithelial electrical resistance (TER). The subcellular distributions of the AJ proteins E-cadherin and β-catenin, the TJ proteins ZO-1 and occludin, and the p65 subunit of nuclear factor (NF)-κB were determined by immunofluorescence staining. The expression of junctional proteins as well as the phosphorylation and degradation of the NF-κB-inhibitory protein IκB-α were examined by immunoblot analysis.

results. IL-1β induced the disappearance of ZO-1 and occludin from the interfaces of neighboring HCE cells without affecting the localization of E-cadherin or β-catenin. It also reduced the TER of HCE cells in a concentration- and time-dependent manner. The overall abundance of TJ and AJ proteins was not affected by IL-1β. IL-1β induced the phosphorylation and downregulation of IκB-α as well as the translocation of p65 to the nucleus. The NF-κB inhibitor curcumin blocked the effects of IL-1β on both TER and the subcellular localization of ZO-1 and occludin.

conclusions. IL-1β induced the redistribution of ZO-1 and occludin from TJs of HCE cells and thereby disrupted the barrier function of these cells in a manner dependent on NF-κB. These effects of IL-1β may contribute to the loss of corneal epithelial barrier function associated with ocular inflammation.

The corneal epithelium contributes to the transparency of the cornea and provides a barrier to prevent external material such as dust and bacteria from entering the eye. ^{1 2 3} The barrier function of the corneal epithelium is susceptible to disruption by bacterial infection or inflammation, resulting in various pathologic conditions of the epithelium as well as in stromal edema and melting. ^{4 5 6} Intercellular junctions such as adherens junctions (AJs) and tight junctions (TJs) play an important role in the formation and maintenance of epithelial barriers. TJs form at the apical plasma membrane of epithelial cells ^{7 8} and contribute to the physiological barrier function of the corneal epithelium. ⁹ TJs consist of transmembrane proteins, including occludin, claudin, and junctional adhesion molecules, as well as cytoplasmic proteins such as zonula occludens (ZO)-1, -2, and -3, that interact with the transmembrane proteins at the internal face of the cell membrane. ^{10 11 12} AJs are positioned below tight junctions and consist of a complex of cadherins and catenins (α-catenin, β-catenin, p120 catenin). ^{13 14} Both AJs and TJs are tightly linked to the actin cytoskeleton. ^{15 16} E-cadherin, β-catenin, ZO-1, and occludin have been detected in the corneal epithelium ^{17 18 19 20} and contribute to its barrier function. ^{6 21 22 23}  

The tissue inflammation that results from insults such as trauma and infection is mediated by resident and infiltrated cells. Various cytokines and growth factors are secreted by these cells, and the inflammatory response is controlled by autocrine and paracrine signaling. Interleukin (IL)-1β is a proinflammatory cytokine ^{24 25} and is released at sites of ocular inflammation. ²⁶ It plays an important role in ocular inflammation associated with trauma, infection, and dry eye. ^{27 28 29} Ocular inflammation alters the structure and function of the corneal epithelium in diseases of the ocular surface. ^{30 31 32}  

To clarify the mechanism by which inflammation results in damage to the corneal epithelium, we have now investigated the effects of IL-1β on the structure and function of intercellular junctions both by analysis of the distribution of E-cadherin, β-catenin, ZO-1, and occludin, and by measurement of transepithelial electrical resistance (TER) in cultured human corneal epithelial cells. Given that nuclear factor (NF)-κB is a key mediator of IL-1β action, ^{33 34 35} we also examined whether the NF-κB signaling pathway might also contribute to the effects of this cytokine on corneal epithelial cells. 

Methods

Materials

Dulbecco’s modified Eagle’s medium–nutrient mixture F12 (DMEM-F12), phosphate-buffered saline (PBS), fetal bovine serum, trypsin-EDTA, and gentamicin were obtained from Invitrogen-Gibco (Carlsbad, CA). Bovine serum albumin (BSA), bovine recombinant insulin, cholera toxin, human recombinant epidermal growth factor, and a protease inhibitor cocktail were obtained from Sigma-Aldrich (St. Louis, MO). Recombinant human IL-1β was from R&D Systems (Minneapolis, MN), and curcumin was from Merck (Darmstadt, Germany). Six- or 24-well transwell plates as well as 24- or 96-well culture plates were obtained from Corning (Corning, NY). Rabbit polyclonal antibodies to ZO-1 or to occludin were obtained from Zymed-Invitrogen, and mouse monoclonal antibodies to E-cadherin or to β-catenin were from BD Bioscience (Carlsbad, CA). Rabbit polyclonal antibodies to the p65 subunit of NF-κB as well as mouse monoclonal antibodies to IκB-α or to phosphorylated IκB-α were from Santa Cruz Biotechnology (Santa Cruz, CA). Mouse monoclonal antibodies to myosin light chain (MLC) or to actin were obtained from Sigma-Aldrich, and rabbit polyclonal antibodies to phosphorylated MLC were from Cell Signaling (Danvers, MA). Horseradish peroxidase–conjugated goat antibodies to mouse or rabbit immunoglobulin G (IgG) as well as an enhanced chemiluminescence detection reagent (ECL Plus) were from Amersham Biosciences GE Healthcare (Little Chalfont, UK). A counterstain (TOTO-3 iodide) and AlexaFluor 488-labeled goat antibodies to rabbit or mouse IgG were obtained from Invitrogen. A cytotoxicity assay kit (CytoTox96 Non-Radioactive Cytotoxity Assay kit) was obtained from Promega (Madison, WI). 

Cell Culture

Simian virus 40-immortalized human corneal epithelial (HCE) cells ³⁶ were obtained from RIKEN Biosource Center (Tokyo, Japan) and were passaged in supplemented hormonal epithelial medium (SHEM), which comprises DMEM-F12 supplemented with 15% heat-inactivated fetal bovine serum, bovine insulin (5 μg/mL), cholera toxin (0.1 μg/mL), recombinant human epidermal growth factor (10 ng/mL), and gentamicin (40 μg/mL). For experiments, HCE cells were plated at a density of 5 × 10⁴ cells or 5 × 10⁵ cells per well in 24- or six-well transwell plates, respectively, or at densities of 1 × 10⁴ or 5 × 10⁴ cells per well in 96- or 24-well culture plates, respectively. They were then cultured for 4 days in SHEM and for 24 hours in unsupplemented DMEM-F12 before exposure to IL-1β or curcumin in the latter medium. 

Immunofluorescence Analysis

HCE cells were cultured in 24-well culture plates. For staining of E-cadherin, β-catenin, ZO-1, or occludin, the cells were fixed with 100% methanol for 20 minutes at room temperature. For staining of the p65 subunit of NF-κB, the cells were fixed with 4% paraformaldehyde in PBS, washed with PBS, and then permeabilized with 100% methanol for 5 minutes at −20°C. All cells were then washed with PBS and incubated at room temperature first for 1 hour with 1% BSA in PBS and then for 1 hour with primary antibodies or with normal rabbit or mouse IgG as a control at a 1:100 dilution in PBS containing 1% BSA. After washing with PBS, the cells were incubated at room temperature for 1 hour with AlexaFluor 488-labeled secondary antibodies at a 1:1000 dilution in PBS containing 1% BSA and then for 10 minutes with TOTO-3 for staining of nuclei. They were finally examined with a laser confocal microscope (LSM5; Carl Zeiss, Wexford, Germany). 

Immunoblot Analysis

HCE cells were cultured in six-well transwell plates. They were lysed in 300 μL of a solution containing 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1 mM EDTA, 5 mM NaF, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% SDS, 1 mM Na₃VO₄, and 1% protease inhibitor cocktail. The cell lysates were centrifuged at 15,000g for 10 minutes at 4°C, and the resulting supernatants were subjected to SDS-polyacrylamide gel electrophoresis on 7.5% or 10% gels. The separated proteins were transferred to a nitrocellulose membrane, which was then incubated at 4°C first for 16 hours with blocking solution [20 mM Tris-HCl (pH 7.4), 5% dried skim milk, 0.1% Tween 20] and then for 16 hours with primary antibodies at a 1:1000 dilution in blocking solution. After washing with a solution containing 20 mM Tris-HCl (pH 7.4) and 0.1% Tween 20, the membrane was incubated for 1 hour at room temperature with horseradish peroxidase-conjugated secondary antibodies at a 1:1000 dilution in the same solution, washed again, incubated with ECL Plus detection reagents for 5 minutes, and then exposed to film. 

Measurement of TER

HCE cells were cultured in 24-well transwell plates on filters with a pore size of 0.4 μm. Resistance was measured with the use of a volt-ohm meter (EVOM; World Precision Instruments, Sarasota, FL), and TER (ohms · cm²) was calculated by multiplying the measured resistance by the area of the transwell filter as previously described. ³⁵ The background resistance due to the filter alone was subtracted from the experimental values. 

Cytotoxicity Assay

HCE cells were cultured in 96-well culture plates. The cytotoxicity of IL-1β or curcumin was evaluated by determination of the activity of lactate dehydrogenase (LDH) released into the culture medium during 24 hours with the use of a CytoTox96 nonradioactive assay. Absorbance at 490 nm was measured with a microplate reader. Data were compared with the amount of LDH activity released from nontreated cells exposed to lysis solution and with baseline LDH release from nontreated cells exposed to culture medium only. 

Statistical Analysis

Quantitative data are presented as means ± SE. Differences were analyzed by Dunnett’s multiple comparison test. A P value of <0.05 was considered statistically significant. 

Results

To evaluate the barrier function of HCE cells, we measured the TER of transwell cultures. We previously showed that the TER of HCE cells increases in a time-dependent manner, eventually achieving a plateau that represents establishment of barrier function. ³⁷ HCE cells were therefore exposed to IL-1β after the establishment of barrier function. Exposure of the cells to IL-1β (0 to 10 ng/mL) for 24 hours resulted in a concentration-dependent decrease in TER, with this effect being significant at concentrations of ≥0.3 ng/mL and maximal at ≥3 ng/mL (Fig. 1A) . The effect of IL-1β (1 ng/mL) was also time dependent, being significant between 6 and 24 hours and maximal at 12 to 24 hours (Fig. 1B) . 

To investigate the effects of IL-1β on the components of AJs and TJs in HCE cells, we first examined the distributions of the AJ proteins E-cadherin and β-catenin and of the TJ proteins ZO-1 and occludin by immunofluorescence microscopy. E-cadherin, β-catenin, ZO-1, and occludin were localized at the interfaces of adjacent HCE cells in the absence of IL-1β. Exposure of the cells to IL-1β (1 ng/mL) for 24 hours resulted in a loss of ZO-1 and occludin immunoreactivity from the cellular borders, but it did not affect the distribution of E-cadherin or β-catenin (Fig. 2) . 

To examine the possible role of the NF-κB signaling pathway in disruption of the barrier function of HCE cells by IL-1β, we first determined whether IL-1β induces the phosphorylation and degradation of the NF-κB–inhibitory protein IκB-α. Immunoblot analysis revealed that incubation of HCE cells with IL-1β (1 ng/mL) resulted in the appearance of the phosphorylated form of IκB-α within 5 minutes and in a decrease in the amount of this protein within 15 minutes (Fig. 3A) . We next examined whether IL-1β might affect the subcellular localization of the p65 subunit of NF-κB. Immunofluorescence microscopy revealed that p65 was localized to the cytoplasm of HCE cells incubated in the absence of IL-1β. Exposure of the cells to IL-1β (1 ng/mL) for 30 minutes induced the translocation of p65 from the cytoplasm to the nucleus (Fig. 3B) . Furthermore, this effect of IL-1β was prevented by incubation of the cells with the NF-κB inhibitor curcumin (10 μM) for 1 hour before exposure to the cytokine (Fig. 3B) . 

We next investigated the effect of curcumin (10 μM) on the apparent redistribution of the TJ proteins ZO-1 and occludin induced by exposure of HCE cells to IL-1β (1 ng/mL) for 24 hours. Immunofluorescence microscopy revealed that the loss of ZO-1 and occludin immunoreactivity from the interfaces of adjacent HCE cells induced by IL-1β was blocked by curcumin (Fig. 4) . Curcumin (10 μM) also significantly reduced the extent of the decrease in TER of HCE cells induced by exposure to IL-1β (1 ng/mL) for 24 hours (Fig. 5) . 

We also examined the effects of IL-1β on the expression of AJ and TJ proteins by immunoblot analysis. Exposure of the cells to IL-1β (1 ng/mL) for 24 hours had no effect on the abundance of ZO-1, occludin, E-cadherin, or β-catenin (Fig. 6) . Immunoblot analysis also showed that incubation of HCE cells with IL-1β (1 ng/mL) for 24 hours resulted in the appearance of the phosphorylated form of myosin light chain (MLC) and that this effect was inhibited by curcumin at 10 μM (Fig. 7) . The amount of MLC was not affected by IL-1β or curcumin. 

Finally, by measuring the release of LDH, we investigated whether IL-1β or curcumin might exert a cytotoxic effect on HCE cells. Treatment of the cells for 24 hours with IL-1β (1 or 10 ng/mL) or curcumin (10 μM) did not significantly increase the release of this cytosolic enzyme (Fig. 8) . 

Discussion

We have shown that IL-1β, a proinflammatory cytokine, reduced the TER of HCE cells in a time- and concentration-dependent manner. Immunofluorescence analysis revealed that IL-1β induced the disappearance of the TJ proteins ZO-1 and occludin from the interfaces of adjacent HCE cells without affecting the localization of the AJ proteins E-cadherin and β-catenin. We found that IL-1β activated the NF-κB signaling pathway in HCE cells, and that the NF-κB inhibitor curcumin blocked the effects of IL-1β on both the localization of ZO-1 and occludin and the TER of HCE cells. IL-1β did not affect the overall abundance of ZO-1, occludin, E-cadherin, or β-catenin. Neither IL-1β nor curcumin had a cytotoxic effect on HCE cells. Our results thus suggest that IL-1β induced the redistribution of ZO-1 and occludin from the borders of adjacent HCE cells in a manner dependent on the NF-κB signaling pathway, resulting in disruption of barrier function. 

IL-1β has previously been shown to disrupt the barrier function of other types of epithelial cells, including retinal pigment epithelial cells and intestinal epithelial cells. ^{38 39} The disruption of the barrier function of HCE cells induced by IL-1β was accompanied by the disappearance of ZO-1 and occludin from the interfaces of neighboring cells. The distribution of ZO-1 and occludin at apical cell-cell junctions is thought to reflect the formation of a tight barrier in corneal epithelial cells. ^{17 40} Both ZO-1 and occludin are expressed in the superficial corneal epithelium in vivo. ^{17 41} We have previously shown that tumor necrosis factor-α, another proinflammatory cytokine, also disrupted the barrier function of HCE cells in association with a redistribution of ZO-1 from the interfaces of neighboring cells. ³⁷ Our results thus suggest that the loss of ZO-1 and occludin from the borders of adjacent HCE cells underlies the disruption of barrier function induced by IL-1β. We found that the localization of the AJ proteins E-cadherin and β-catenin at the interfaces of HCE cells was not affected by IL-1β. In contrast, disruption of barrier function in intestinal epithelial cells by certain agents was accompanied by redistribution not only of the TJ proteins ZO-1 and occludin but also of E-cadherin and β-catenin. ^{42 43} TJ proteins have been shown to be closely associated with AJ proteins at the interfaces of cells. ^{44 45} The mechanism by which IL-1β changes the localization of TJ proteins without affecting that of AJ proteins in HCE cells remains to be determined. 

The transcription factor NF-κB plays a central role in the regulation of cell growth, apoptosis, and inflammation by controlling gene expression. ^{46 47 48} It has also been implicated in the ubiquitin-proteasome pathway of protein degradation. ^{49 50} We have now shown that IL-1β activated the NF-κB signaling pathway in HCE cells, and that the NF-κB inhibitor curcumin blocked the effects of IL-1β both on the TER of HCE cells and on the localization of ZO-1 and occludin. Given that the disappearance of ZO-1 and occludin from the interfaces of adjacent HCE cells induced by IL-1β was not accompanied by downregulation of the expression of these proteins, IL-1β does not appear to induce their degradation. Previous study showed that MMP-9 activity on the ocular surface in response to dryness is involved in the disruption of corneal epithelial barrier function because of loss of tight junction in corneal epithelial cells, perhaps by proteolytic cleavage of occludin. ⁵¹ Moreover, MMP-1, -3, -9, -10, -11, and -13 in human corneal epithelial cells were up-regulated by IL-1β in dose dependent manner. ^{52 53} It remains possible that IL-1β induces the degradation of other TJ proteins or accessory proteins that interact with ZO-1 or occludin in HCE cells. 

TJ proteins interact with actin filaments and such interactions contribute to regulation of epithelial barrier function. ^{54 55} Reorganization of the actin cytoskeleton accompanied by MLC phosphorylation has thus been shown to disrupt barrier function. ^{56 57} Tumor necrosis factor-α induces the expression of MLC kinase or MLC phosphorylation in corneal epithelial cells and intestinal epithelial cells, and these effects are thought to underlie the disruption of barrier function by this proinflammatory cytokine. ^{37 58 59} We have now shown that IL-1β induced the phosphorylation of MLC in HCE cells, suggesting that reorganization of the actin cytoskeleton may contribute to barrier disruption by IL-1β in these cells. 

In this study, we used SV40 transformed human corneal epithelial cell line, which is one of most commonly investigated immortalized human corneal epithelial cell line. To extrapolate our current findings, it must be reminded that our results were based on the virus transformed cells. However, primary cultured human corneal epithelial cells have the restriction of finite lifespan, and the availability of corneal material is uncertain about human corneal epithelial cells. Because this cell line has been well characterized as to its morphology, and the physiological and biochemical aspects of the cells, ³⁶ we used this transformed cells. 

IL-1β has been detected in the tear fluid of individuals with ocular inflammation, keratoconjunctivitis sicca, or corneal disease. ^{60 61 62} IL-1β also increases the production of several cytokines, growth factors, and adhesion molecules as well as promoting collagen degradation in the cornea. ^{63 64 65 66 67 68} Together with these observations, our present results showing that IL-1β induced disruption of barrier function in HCE cells suggest that this cytokine plays a key role in ocular disease. Maintenance of TJs in the corneal epithelium by blocking the action of IL-1β may thus represent a new approach to the treatment of corneal epithelial disorders associated with ocular inflammation. 

 Supported in part by a Grant (19791271) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.

 Submitted for publication July 23, 2008; revised August 25, 2008; accepted November 3, 2008.

 Disclosure: K. Kimura, None; S. Teranishi, None; T. Nishida, None

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Effect of IL-1β on the barrier function of HCE cells. Cells were cultured in transwell plates and incubated with the indicated concentrations of IL-1β for 24 hours (A) or in the absence (open circles) or presence (closed circles) of IL-1β at 1 ng/mL for the indicted times (B), after which TER was determined. Data are means ± SE from four independent experiments. *P < 0.05, **P < 0.01 versus the corresponding value for cells incubated without IL-1β.

Figure 2.

 

Effects of IL-1β on the distribution of junctional proteins in HCE cells. Cells were incubated in the absence or presence of IL-1β (1 ng/mL) for 24 hours, fixed, and subjected to immunofluorescence analysis with antibodies to ZO-1, to occludin, to E-cadherin, or to β-catenin (green signals). Nuclei were stained with TOTO-3 (blue signal). Scale bar, 10 μm. Data are representative of three independent experiments.

Figure 3.

 

Activation of the NF-κB signaling pathway by IL-1β in HCE cells. (A) Cells were incubated for the indicated times in the absence or presence of IL-1β (1 ng/mL), after which cell lysates were prepared and subjected to immunoblot analysis with antibodies to phosphorylated IκB-α (p-IκB-α) or to total IκB-α. (B) Cells were incubated for 1 hour in the absence or presence of curcumin (10 μM) and then for 30 minutes in the additional absence or presence of IL-1β (1 ng/mL). The cells were then fixed, permeabilized, and subjected to immunofluorescence analysis with antibodies to the p65 subunit of NF-κB (green signal). Nuclei were detected by staining with TOTO-3 (blue signal). Scale bar, 10 μm. Data in (A) and (B) are representative of three independent experiments.

Figure 4.

 

Inhibition by curcumin of the IL-1β–induced redistribution of ZO-1 and occludin in HCE cells. Cells were incubated in the absence or presence of curcumin (10 μM) for 1 hour and then in the additional absence or presence of IL-1β (1 ng/mL) for 24 hours. The cells were then fixed and subjected to immunofluorescence analysis with antibodies to ZO-1 or to occludin (green signals). Nuclei were stained with TOTO-3 (blue signal). Scale bar, 10 μm. Data are representative of three independent experiments.

Figure 5.

 

Inhibition by curcumin of the IL-1β–induced decrease in TER of HCE cells. Cells were incubated for 1 hour in the absence or presence of curcumin (10 μM) and then for 24 hours in the additional absence or presence of IL-1β (1 ng/mL), after which TER was determined. Data are means ± SE from four independent experiments. *P < 0.05.

Figure 6.

 

Lack of effect of IL-1β on the abundance of TJ and AJ proteins in HCE cells. Cells were incubated for 24 hours in the absence or presence of IL-1β (1 ng/mL), after which cell lysates were prepared and subjected to immunoblot analysis with antibodies to ZO-1, to occludin, to E-cadherin, to β-catenin, or to actin (loading control). Data are representative of three independent experiments.

Figure 7.

 

Effect of IL-1β on the phosphorylation of MLC in HCE cells. Cells were incubated in the absence or presence of curcumin (10 μM) for 1 hour and then in the additional absence or presence of IL-1β (1 ng/mL) for 24 hours. Cell lysates were prepared and subjected to immunoblot analysis with antibodies to phosphorylated MLC, to total MLC, or to actin. Data are representative of three independent experiments.

Figure 8.

 

Lack of a cytotoxic effect of IL-1β or curcumin on HCE cells. Cells were incubated in the absence or presence of IL-1β (1 or 10 ng/mL) or curcumin (10 μM) for 24 hours, after which culture supernatants were assayed for LDH activity with a colorimetric assay and measurement of absorbance at 490 nm. The amount of LDH released from cells by cell lysis solution was determined as a positive control. Data are means ± SE from four independent experiments.

The authors thank Makoto Asashima (ICORP Organ Regeneration Project, Japan Science and Technology Agency) for discussion as well as Yasumiko Akamatsu and the staff of the Yamaguchi University Center for Gene Research for technical assistance.