971

INTRODUCTION

The ocular surface is a mucosal structure directly exposed to a great variety of environmental agents, some of them noxious, such as pathogens, allergens or irritants. In this structure the corneal epithelium forms a barrier that not only regulates the passive movement of molecules through the paracellular pathway, but also prevents foreign material from entering the eye.1 An altered corneal epithelium result in a vulnerable cornea2 and it is frequently associated with an increased risk of sterile or infectious corneal ulceration or persistent epithelial defects, for instance.

An intact corneal epithelium is therefore essential to maintain ocular surface homeostasis, and intercellular junctions, such as tight junctions (TJs) and adherens junctions (AJs), play a key role in the formation and maintenance of this epithelial barrier.

TJs are cell–cell junctions that seal adjacent cells together, preventing the passage of most solute molecules from one side of the epithelial layer to the other. They form a complex structure that mainly consists of transmembrane proteins, such as claudins and occludin, as well as cytoplasmic proteins, such as the zonula occludens protein family.3–5 Their molecular composition plays a major role in the

Current Eye Research, 37(11), 971–981, 2012© 2012 Informa Healthcare USA, Inc.ISSN: 0271-3683 print/1460-2202 onlineDOI: 10.3109/02713683.2012.700756

Received 02 April 2012; revised 23 May 2012; accepted 03 June 2012

Correspondence: Yolanda Diebold, Ph.D. IOBA-University of Valladolid, Edificio IOBA, Campus Miguel Delibes, Paseo de Belén 17, E-47011 Valladolid (Spain). Tel: +34–983-18 47 50. Fax: +34–983-18 47 62. E-mail: yol@ioba.med.uva.es

02April2012

23May2012

03June2012

© 2012 Informa Healthcare USA, Inc.

2012

Current Eye Research

1460-2202

10.3109/02713683.2012.700756

37

971

981

11

Original article

Structural and Functional Alteration of Corneal Epithelial Barrier Under Inflammatory Conditions

Laura Contreras-Ruiz1,2, Ute Schulze3, Laura García-Posadas1,2, Isabel Arranz-Valsero1,2, Antonio López-García1,2, Friedrich Paulsen3,4, and Yolanda Diebold1,2

1Ocular Surface Group-IOBA, University of Valladolid, Valladolid, Spain, 2Networking Research Centre on Bioengineering, Biomaterials and Nanomedicine (CIBER-BBN), Spain, 3Department of Anatomy and Cell Biology,

Martin Luther University Halle-Wittenberg, Halle/Saale, Germany, and 4Department of Anatomy II, Friedrich Alexander University Erlangen-Nürnberg, Germany

ABSTRACT

Purpose: The aim of the study was to determine the effect of inflammatory conditions on the expression of tight junction (TJ) and adherens junction (AJ) proteins between human corneal epithelial cells and, consequently, on corneal epithelial barrier integrity.

Materials and methods: Zonula occludens proteins ZO-1 and ZO-2, claudin-1 and -2 (CLDN-1 and CLDN-2), occludin (OCLN) as well as E-cadherin (E-cad) expression were analyzed in a human corneal epithelial cell line (HCE) at basal conditions and after stimulation with inflammatory cytokines (TNFα, TGFβ, IL-10, IL-13, IL-17, IL-6), using real time RT-PCR, Western blotting and immunofluorescence. Actin cytoskeleton staining was performed after all stimulations. Transepithelial electrical resistance (TER) and fluorescein transepithelial permeability (TEP) were measured as barrier integrity functional assays.

Results: ZO-1, ZO-2, CLDN-1, CLDN-2, OCLN and E-cad were detected in HCE cell membranes at basal conditions. Cytokine stimulation resulted in significant changes in the expression of TJ and AJ proteins, both at mRNA and protein level, a remarkable change in their localization pattern, as well as a reorganization of actin cytoskeleton. Pro-inflammatory cytokines TNFα, TGFβ, IL-13, IL-17 and IL-6 induced a structural and functional disruption of the epithelial barrier, while IL-10 showed a barrier protective effect.

Conclusion: Simulated inflammatory conditions lead to an alteration of corneal barrier integrity by modulating TJ, and to a lesser extent also AJ, protein composition, at least in vitro. The observed barrier protective effects of IL-10 support its well-known anti-inflammatory functions and highlight a potential therapeutic perspective.

KEYWORDS: Corneal epithelium, Corneal barrier, Inflammation, Cytokines, Tight junctions

0271-3683

corneal Barrier Under inflammatory conditions

l. contreras-ruiz et al.

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Lav

al o

n 02

/27/

13Fo

r pe

rson

al u

se o

nly.

972 L. Contreras-Ruiz et al.

Current Eye Research

regulation of the epithelial barrier function.6 The expression of most of these proteins has been previously described in the cornea,7 although their regulation is not completely understood, yet.

AJs occur in most epithelia as adherens contacts encircling the cell, also designated as zonula adherens. Based on this girdle, a cell is connected to all surround-ing cells.8 E-cadherin (E-cad) is the best-studied mol-ecule. It is a transmembrane protein whose extracellular segment binds to those of adjacent cells.

The barrier function of the corneal epithelium can be disrupted as a result of several factors; one of the most important is inflammation. Inflammation constitutes the biological process underlying a great deal of ocular diseases, including complex processes such as dry eye syndrome,9 allergic diseases, or the more severe cica-trizing conjunctivitis,10 in addition to ocular infections, traumas, and surgery. Barrier function is compromised in many of these conditions. For example, corneal epi-thelial permeability to fluorescein dye in patients with dry eye syndrome was reported to be 2.7 to 3 times greater than in healthy eyes.11

It is thought that epithelial cells, including corneal epithelial cells, actively participate in inflammatory processes, expressing adhesion and co-stimulatory molecules in response to different cytokines, such as TNFα, TGFβ, IL-6 or IL-10.12 These cytokines play a key role in the coordination and persistence of ocular inflammation. Altered levels of several cytokines have been found in different types of inflammatory ocular surface diseases.10 Tears from allergic patients showed a significant decrease in levels of IL-10, which is consid-ered an important anti-inflammatory cytokine, whereas TNFα and IL-13, two pro-inflammatory cytokines that contribute to ocular inflammation, were significantly increased.10 It has also been reported that desiccating stress in C57BL/6 mice induces a significant progressive increase of several cytokines such as (1) IL-6, which is a potent pro-inflammatory cytokine of Th1 lymphocytes; (2) IL-17, a pro-inflammatory cytokine that presents an interesting bridge between the adaptive and innate immune system; and (3) TGFβ, a pleiotropic molecule that works as a pro-inflammatory or anti-inflammatory cytokine according to surrounding conditions.13

The aim of the present study was to gain insights into the mechanism by which inflammation can result in dis-ruption of the barrier function of the corneal epithelium. To accomplish this, we investigated the effect of several inflammatory mediators on corneal barrier structure and functionality using a corneal epithelial cell line.

MATERIALS AND METHODS

Human Corneal Epithelial Cells

The SV40-immortalized HCE human corneal epithe-lial cell line14 was used. Cells from passages 42 to 52

were cultured in DMEM/F-12 supplemented with 15% fetal bovine serum (FBS), 100 U/ml penicillin, 0.1 mg/ml streptomycin, 10 ng/ml EGF, 0.5% DMSO, 5 µg/ml insulin, and 0.1 µg/ml cholera toxin (all from Invitrogen-Gibco, Inchinnan, UK). Cells 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.

In Vitro Inflammation Model

An in vitro inflammation model was used as previously described.12 Briefly, HCE cells were plated in 24-well plates (80,000 cells/well), 96-well plates (10,000 cells/well) or 8-well multichamber Permanox™ slides (15,000 cells/well) (Nunc, Roskilde, Denmark) and grown for 48–72 h. Cells were then maintained for 24 h in serum-free, non supplemented medium, before treating them with IL-10 (20 ng/ml), IL-13 (20 ng/ml), TNFα (25 ng/ml), TGFβ (10 ng/ml), IL-17 (10 ng/ml), IL-6 (10 ng/ml), or a combination of IL-10 and TNFα (10 ng/ml each) (all from PeproTech, London, UK) for 48 h. At that time, a tight confluent cell monolayer had been already formed. Doses of the stimulatory cytokines used were chosen based on previous publication.12 Controls were untreated cells. At least three independent experiments were performed for each cytokine stimulation.

Real-Time RT-PCR

Isolation of total RNA from cells was performed using TRIZOL® reagent (Invitrogen-Gibco) following the manufacturer’s protocol. The RNA was pretreated with 1 U/1 µg RNase-free DNase I (Promega, Madison, WI, USA) to prevent DNA contaminations. The cDNA synthesis was performed using Revertaid First Strand cDNA Synthesis Kit (Fermentas GmbH, Madrid, Spain), according to manufacturer’s instructions.

Conventional RT-PCR was performed to study the expression of TJ complex proteins zonula occludens 1 and 2 (ZO-1 and ZO-2), claudins 1 and 2 (CLDN-1 and CLDN-2) and occludin (OCLN), and E-cad in unstimu-lated conditions. Table 1 shows primer sequences used for RT-PCR. PCR products were sequenced to confirm that the amplified products were the target sequences.

Real Time RT-PCR (RT2-PCR) was performed on the iQ5 system (BioRad, Hercules, CA, USA) using SYBR Green (BioRad) to quantify levels of TJ mRNAs in HCE cells. The same primer sequences as in conventional RT-PCR (Table 1) were used for RT2-PCR. The cycle profile was 50°C for 2 min, 95°C for 2 min, 40 cycles at 95°C for 20 s, annealing temperature for 30 s and 72°C for 40 s. To verify the specificity of the amplification reaction, a melting curve analysis was performed. Relative quantification of the signals was done by normalizing the signal of the studied TJ genes with

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Lav

al o

n 02

/27/

13Fo

r pe

rson

al u

se o

nly.

Corneal Barrier Under Inflammatory Conditions 973

© 2012 Informa Healthcare USA, Inc.

the 18S rRNA signal. The difference in expression for each target in the treated samples was compared to the amount in the control group. Three independent experiments, in triplicate, were performed.

Immunofluorescence Analysis

Treated and control cells were fixed in ice-cold metha-nol and incubated at room temperature (RT) for 50 min with blocking buffer composed of phosphate-buffered saline (PBS) with 4% donkey serum, 0.3% Triton X-100, and 1% bovine serum albumin (all from Sigma-Aldrich, St. Louis, MO, USA) to block non-specific binding. Afterwards, they were incubated with the primary antibodies (Abs) (Table 2) for 1 h at RT [ZO-2, CLDN-1, and OCLN] or overnight at 4°C [ZO-1 CLDN-2 and E-cad]. Alexa Fluor 488-conjugated secondary Abs (Table 2) were incubated for 1 h at RT. Preparations were visualized under a Leica DMI 6000B fluorescence micro-scope (Leica Microsystems Wetzlar GmbH, Mannheim, Germany). Specificity of Abs had been previously tested using cornea tissues.

In addition, cells were stained with rhodamine phal-loidin, (Invitrogen), which targets cytoskeletal actin network. Briefly, cells were fixed in ice-cold methanol, permeabilized in 5% Triton X-100 and stained with 10 U/ml phallodin. After 30 min of incubation, cells

were washed in PBS before microscopy examination. Fluorescence was visualized with a Leica TCS SPE confocal laser microscope equipped with a solid state laser with a 488 nm excitation filter. Cells were observed under a dry 40× objective with a pinhole of 1 (optical slice thickness was 1.7 µm). Four independent experi-ments, in duplicate, were performed, and negative controls included omission of primary Abs.

Electrophoresis and Western Blotting

HCE cells were homogenized in ice-cold radioim-munoprecipitation assay (RIPA) buffer plus protease inhibitors (10 µl/ml phenylmethylsulfonyl fluoride, 6 µl/ml aprotinin and 100 nM sodium orthovanadate). Total protein homogenates were separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) on 10% acrylamide gels according to Laemmli,15 transferred to nitrocellulose membranes according to Towbin,16 and blocked for 1 h in tris-buffered saline (TBS) containing 0.05% Tween-20, 5% milk and 4% donkey serum. Membranes were incu-bated with primary Abs (Table 2) overnight at 4°C and with HRP-conjugated secondary Abs (Table 2) for 1 h at RT. Glyceraldehyde 3-phosphate dehy-drogenase (GAPDH) (Abcam, Camgridge, UK) was used as loading control. Immunoreactive bands were

TABLE 1 Primer sequences of TJ complex genes.Gene Description Forward primer (5′–3′) Reverse primer (5′–3′) Length (bp)*ZO-1 Zonula Occludens Protein 1 TCTGCCCGACCATTTGAACG AGGCTGCAAACTGTGCG 149ZO-2 Zonula Occludens Protein 2 CGAACGGGTCTGGCAACTAAAG AATGACGGGATGTTGATGAGGG 185CLDN-1 Claudin 1 AGCTGTTGGGCTTCATTCTCGC TGGGCGGTCACGATGTTGTC 112CLDN-2 Claudin 2 AGCTCTCCAAGGCCTGGTCAAC GCAGCACCTTCTGACACGATCC 186OCLN Occludin ATTGCCATCTTTGCCTGTGTGG GCCATAGCCATAGCCACTTCCG 13518S 18S ACTCAACAGGGGAAACCTCAGC CGCTCCACCAACTAAGAACGG 125E-cad E-cadherin Unknow (SABioscience, Frederick, MD, USA). Ref. NM_004360.3 174*Source for all primers (aside from E-cad): Invitrogen-Gibco, Inchinnan, UK.

TABLE 2 Antibodies (Abs) of TJ complex proteins and E-cadherin.Immunofluorescence Western blotting

1st Abs 2nd Abs 1st Abs 2nd AbsZO-1 2.5 µg/ml 20 µg/ml* 1 µg/ml 0.2 µg/ml‡

Invitrogen Invitrogen Invitrogen Sta. Cruz BiotechZO-2 4 µg/ml 20 µg/ml* 1 µg/ml 0.2 µg/ml‡

Sta. Cruz Biotech Invitrogen Sta. Cruz Biotech Sta. Cruz Biotech.CLDN-2 2.5 µg/ml 20 µg/ml* 1 µg/ml 0.2 µg/ml‡

Zymed-Invitrogen Invitrogen Zymed-Invitrogen Sta. Cruz Biotech.OCLN 5 µg/ml 20 µg/ml† 1 µg/ml 0.05 µg/ml§

Zymed-Invitrogen Invitrogen Zymed-Invitrogen Jackson Inmuno Res.CLDN-1 2 µg/ml 20 µg/ml† 1 µg/ml 0.05 µg/ml§

Sta. Cruz Biotech Invitrogen Sta. Cruz Biotech Jackson Inmuno Res.E-cadherin 2.5 µg/ml 10 µg/ml† 0.5 µg/ml 0.05 µg/ml§

BD Bioscience Invitrogen BD Bioscience Jackson Inmuno Res.*AlexaFluor488 Donkey anti-Rabbit (Invitrogen, Inchinnan, UK); †AlexaFluor488 Donkey anti-Mouse (Invitrogen); ‡HRP Goat anti-Rabbit (Santa Cruz Biotechnolgy Inc., Santa Cruz, CA, USA); §HRP Donkey anti-Mouse (Jackson Inmuno Research Laboratories, Inc., West Grove, PA, USA).

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Lav

al o

n 02

/27/

13Fo

r pe

rson

al u

se o

nly.

974 L. Contreras-Ruiz et al.

Current Eye Research

visualized by a chemiluminescence method using the ChemiDoc gel documentation system (BioRad), and images were analyzed with the Quantity One soft-ware (BioRad), performing a densitometry analysis of the bands. Three independent experiments, in trip-licate, were performed.

Functional Studies

We performed functional assays to evaluate the effect of the different inflammatory conditions on corneal barrier function. The evaluation included measurements of transepithelial electrical resistance (TER) and transepithelial permeability (TEP). Polycarbonate membrane Transwell® Inserts (24-well size; Corning Inc., Acton, MA, USA) with 0.4 µm pore size, were used. Because HCE cells needed more time to reach confluence on polycarbonate membranes than in conventional plastic support, a proliferation curve was done to decide at what day an optimal cell monolayer was formed to initiate experiments (data not shown). Based on those data, cells seeded on Transwell® Inserts were allowed to grow for 8 days before starting experiments. Four independent experiments, in triplicate, were performed.

TER: TER was measured with Millicel-ERS system (Millipore, Billerica, MA, USA). Results were calculated from the measured resistance and normalized for the area of the insert. The background of the filters was sub-stracted. Results were expressed as relative resistance to the unstimulated control.

TEP: TEP measures the permeability of an epithe-lial monolayer to the diffusion of sodium fluorescein through the cellular barrier. 200 µl of 0.02% sodium fluorescein (W/V) (Sigma-Aldrich) was added to the upper chamber and cultures were incubated at 37°C for 30 min. Afterwards, absorbance (490 nm) of quadru-plicate samples of the medium in the lower chamber was measured using a SpectraMAX®M5 multidetection microplate reader.

Retention of sodium fluorescein was calculated as follows:

1Abs 49 nm

Abs 49 nm of Inserts without cells1–

( )( )

00

00

×

The % of relative fluorescein retention (RFR) was cal-culated after designating the control cultures as having 100% fluorescein retention.

Statistical Analysis

Two-way ANOVA was performed, followed by a Bonferroni test. Results were expressed as mean ± stan-dard error of the mean (SEM). Differences were consid-ered to be significant when p ≤ 0.05 (*).

RESULTS

Expression of TJ and AJ Proteins at Basal Conditions

ZO-1, ZO-2, CLDN-1, CLDN-2 and OCLN were detected in unstimulated HCE cells by RT-PCR, Western blotting and immunofluorescence (Figure 1). E-cad was also detected by RT-PCR, Western blotting and immuno-fluorescence. Both, RT-PCR amplification products and Western blotting bands, corresponded to their expected molecular weights. All these proteins, aside from CLDN-2 that showed a diffuse staining, were always present in cell–cell contact points of plasma membrane.

In vitro Inflammatory Conditions Modify the Expression of TJ mRNA and Proteins

The study of mRNA and protein expression of TJs under inflammatory conditions showed an alteration in the expression of almost all the studied genes and proteins after different cytokine stimulations (Figure 2 and Table 3).

TNFα and IL-13 stimulations led to a significantly increased expression of ZO-2 and CLDN-1 mRNA and a significant decrease in OCLN mRNA expression. At the protein level, ZO-2, CLDN-1 and CLDN-2 showed significantly increased expression after stimulation with either cytokines.

Cells exposed to IL-17 or IL-6 showed a significant decrease in the expression of ZO-2 mRNA and a signifi-cant increase in CLDN-1 and OCLN mRNA expression. Western blot analysis revealed significantly increased expression of all studied TJ proteins, aside from the sig-nificantly decreased expression of ZO-2.

Stimulation with IL-10 resulted in a significantly increased expression of ZO-2 and CLDN-1 mRNA, and a significant increase in the expression level of all studied TJ proteins.

Regarding TGFβ, RT2-PCR revealed significantly increased expression of ZO-2 and CLDN-1 mRNA and a significant decrease in CLDN-2 and OCLN mRNA expression. Similar results were obtained at the protein level with significant upregulation of ZO-1, ZO-2 and CLDN-1 proteins and significant downregulation of OCLN protein.

In vitro Inflammatory Conditions Affect TJ Protein Distribution

Immunofluorescence revealed a remarkable change in the localization pattern of TJ proteins after cytokine stimula-tion (Figure 3). In control cells, aside from CDLN-2 that showed a diffuse staining, all TJ proteins were located in the plasma membrane at cell–cell contact points.

Cells exposed to IL-10 showed the same localization pattern as in controls, aside from ZO proteins with slightly

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Lav

al o

n 02

/27/

13Fo

r pe

rson

al u

se o

nly.

Corneal Barrier Under Inflammatory Conditions 975

© 2012 Informa Healthcare USA, Inc.

distribution changes. However, after the stimulation with TNFα and IL-13, TJ proteins showed a more predominant cytosolic localization, inducing the disappearance of the TJ proteins from the interfaces of adjacent corneal epithelial cells. Similar results were obtained in TGFβ exposed-cells. After IL-17 and IL-6 stimulation, all studied TJ proteins were identified in the plasma membrane, although a more irregular distribution was observed.

Some In Vitro Inflammatory Conditions Affect AJs Complexes

The study of E-cad expression after cytokine stimulation at the protein level showed an alteration in AJ compo-sition under inflammatory conditions, although to a lesser extent than that was observed in TJ proteins. No

significant changes in E-cad expression were detected after TNFα, TGFβ, IL-10 and IL-13 stimulation, while IL-17 and IL-6-exposed cells showed significantly increased expression (Figure 4).

In unstimulated cells, E-cad was located in the plasma membrane, at cell–cell contacts. After cytokine stimulation no change in E-cad localization was visible with remarkable membrane localization in all cytokine-exposed cells.

In Vitro Inflammatory Conditions Induce Reorganization of the Actin Cytoskeleton

Phalloidin staining revealed changes in the distribution of actin cytoskeleton in HCE cells after cytokine

FIGURE 1 Expression of TJ (a–c) and AJ (d–f) mRNA and proteins in HCE cells under basal conditions. TJ proteins ZO-1, ZO-2, CLDN-1, CLDN-2 and OCLN and the AJ protein E-cad were detected in HCE cells by RT-PCR (a, d), Western blotting (b, e) and immunofluorescence (c, f). All of these proteins, aside from CLDN-2 that shows a diffuse staining, are present in cell–cell contact points of plasma membrane. Representative images of four different experiments are shown. Scale bar = 30 µm.

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Lav

al o

n 02

/27/

13Fo

r pe

rson

al u

se o

nly.

976 L. Contreras-Ruiz et al.

Current Eye Research

stimulation (Figure 5). In control cells, actin was localized mainly at the cell border in a dense peripheral band. After cytokine stimulation, the peripheral band of actin begins to ruffle and fibrillar material appears in the cytoplasm. A relative increase in newly-formed stress fibers became apparent upon TNFα, TGFβ, IL-10 and IL-13 treatment. The reorganization effect was not so evident after IL-17 and IL-6 stimulation.

In Vitro Inflammatory Conditions Functionally Alter the Corneal Epithelial Barrier

To evaluate the barrier function integrity in HCE cells exposed to cytokines, we measured the TER and the TEP of transwell cultures. We had previously observed that the TER of HCE cells increased in a time-dependent manner achieving a plateau at day 7 that represents the

FIGURE 2 Changes in the expression of TJ proteins under inflammatory conditions at mRNA and protein levels, evaluated by RT2-PCR, and densitometry analysis of Western blot bands, respectively. The anti-inflammatory cytokine IL-10 shows upregulation of all TJ protein expressions. TNFα and IL-13, both pro-inflammatory cytokines, as well as TGFβ, produce a decrease in OCLN expression. In IL-17 and IL-6-exposed cells, ZO-2 expression is significantly decreased. Stimulated cells: black bars; unstimulated control cells: grey bars (n = 3).

TABLE 3 Changes in expression of TJ proteins under inflammatory conditions at mRNA and protein levels.Cytokine mRNA Protein Functional testsTNFα ↑ ZO-2 and CLDN-1 ↑ ZO-2, CLDN-1 and CLDN-2

= ZO-1 and CLDN-2 = ZO-1 and OCLN ↓TER and TEP↓OCLN ↓

TGFβ ↑ ↑ZO-2 and CLDN-1 ↑ ZO-1, ZO-2 and CLDN-1= ZO-1 = ↓TER and TEP↓ CLDN-2 and OCLN ↓OCLN and CLDN-2

IL-10 ↑ ZO-2 and CLDN-1 ↑ ZO-1, ZO-2, CLDN-1, CLDN-2, OCLN= ZO-1, CLDN-2 and OCLN = = TER and TEP↓ ↓

IL-13 ↑ ZO-2 and CLDN-1 ↑ ZO-2, CLDN-1 and CLDN-2= ZO-1 and CLDN-2 = ZO-1 and OCLN ↓TER and TEP↓OCLN ↓

IL-17 ↑ CLDN-1 and OCLN ↑ ZO-1, CLDN-1, CLDN-2 and OCLN= ZO-1 = ↓TER and TEP↓ZO-2 and CLDN-2 ↓ZO-2

IL-6 ↑ CLDN-1 and OCLN ↑ ZO-1, CLDN-1, CLDN-2 and OCLN= ZO-1, CLDN-2 and ZO-2 = ↓TER and TEP↓ ↓ZO-2

↑: Significantly increase; =: Without significant changes; ↓: Significantly decrease.

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Lav

al o

n 02

/27/

13Fo

r pe

rson

al u

se o

nly.

Corneal Barrier Under Inflammatory Conditions 977

© 2012 Informa Healthcare USA, Inc.

establishment of barrier function (data not shown). HCE were therefore stimulated with the different cytokines after the establishment of barrier function.

Cell stimulation with TNFα, IL-17, IL-13, IL-6 and TGFβ resulted in a significant decrease in TER values. The decrease was significantly higher in TNFα- and TGFβ-exposed cells than in IL-13-, IL-17- and IL-6-exposed cells. IL-10 did not induce significant TER changes (Figure 6a).

After TER measurements, TEP measurements were done in the same HCE transwell culture (Figure 6b). HCE monolayer stimulated with TNFα, TGFβ, IL-13 IL-17 and IL-6 retained around 10% less fluorescein than control HCE monolayer, indicating that HCE cells exposed to any of these cytokines were about 10% more permeable than control cells. Interestingly, no changes in perme-ability were detected after IL-10 stimulation (Figure 6b).

FIGURE 3 Distribution of TJ proteins in HCE cells under in vitro inflammatory conditions. Under basal conditions and in IL-10, IL-17 and IL-6-exposed cells, TJ proteins are present in the plasma membrane. After TNFα, TGFβ and IL-13 stimulation, TJ proteins show a predominantly cytosolic localization. Representative images of four different experiments are shown. Scale bar = 30 µm.

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Lav

al o

n 02

/27/

13Fo

r pe

rson

al u

se o

nly.

978 L. Contreras-Ruiz et al.

Current Eye Research

In order to evaluate a potential protective effect of IL-10 in the corneal barrier function, TER and TEP of HCE cells were studied after IL-10 and TNFα co-stimu-lation (Figure 7). Both TER and TEP values significantly increased after the co-stimulation when compared with those obtained after TNFα stimulation alone. This result would indicate that TNFα-exposed cells are less perme-able and show more resistance when IL-10 is present.

DISCUSSION

In this study, we evaluated the effect of in vitro inflam-matory conditions on the structure and function of corneal epithelial barrier. We found that the tested cytokines modify the expression of TJ and AJ proteins in human corneal epithelial cells, induce reorganization of the actin cytoskeleton, and consequently alter the corneal epithelial barrier function.

It was previously reported that the molecular composition of intercellular junction complexes plays a major role in the regulation of the corneal barrier function.3 It is important to point out that TJ permeability may be regulated through modifications on the cytoskeleton, too.17 It is also known that certain inflammatory conditions, such as inflammatory bowel disease, airway inflammation in asthma or cystic fibrosis, affect these junctions and consequently the epithelial barrier.18 However, this is the first in vitro study that evaluates a set of inflammatory cytokines of relevance in ocular surface inflammatory conditions and their effect on both structure and function of the corneal epithelial barrier. We believe that a good understanding of intercellular and molecular mechanisms that mediate cytokine modulation of corneal epithelial barrier function may be important in the development of future therapeutic strategies to preserve this barrier during inflammatory conditions.

Our results indicate that studied cytokines alter the expression of ZO-1, ZO-2, CLDN-1, CLDN-2 and OCLN at mRNA and protein levels, in addition to their cellu-lar localization. There were some differences between mRNA and protein expression levels of TJ components after several stimulations, which might be due to post-transcriptional regulation mechanisms previously reported for TJ proteins.19 Additionally, E-cad expression levels after cytokine stimulation showed no significant alteration, aside from IL-17 and IL-6, which did produce a significantly increased expression of E-cad. No changes in E-cad localization were visible after cytokine stimulation, showing in all cases remarkable membrane localization. Similar results were obtained in other epithelia, such as nasal epithelium.20 We consider that n in this work could be acceptable for the experimental design, as our results have variability enough to permit error calculation but not so high that induces to doubt about the truth of results.

FIGURE 4 Changes in the expression and distribution of E-cad in HCE cells under in vitro inflammatory conditions. (a) No significant changes are detected in E-cad expression after TNFα, TGFβ, IL-10 and IL-13 stimulation, as determined by Western blotting, while IL-17 and IL-6-exposed cells show significantly increased expression. (b) E-cad reveals remarkable membrane localization under all conditions. Representative images of three different experiments are shown. Scale bar = 30 µm.

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Lav

al o

n 02

/27/

13Fo

r pe

rson

al u

se o

nly.

Corneal Barrier Under Inflammatory Conditions 979

© 2012 Informa Healthcare USA, Inc.

In order to explain the altered TJ complex composi-tion we observed, cytokine effects can be grouped in three different groups: (1) effects of pro-inflammatory cytokines, such as TNFα and IL-13, and TGFβ; (2) effects

of the anti-inflammatory cytokine IL-10; and (3) pro-inflammatory cytokines, such as IL-17 and IL-6.

Our data show that TNFα, IL-13, and TGFβ, induce (1) a decrease in OCLN mRNA expression, probably related to a decrease in the number of TJ strands; (2) increased CLDNs expression that does not apparently

FIGURE 5 Reorganization of actin cytoskeleton in HCE cells under in vitro inflammatory conditions. Actin cytoskeleton was labeled with rhodamin-conjugated phalloidin. Representative images of four different experiments are shown. Scale bar = 50 µm.

FIGURE 6 Alteration of barrier function in HCE cells under in vitro inflammatory conditions, evaluated by TER (a) and TEP (b) measurements. TER results are expressed as resistance relative to an unstimulated control and TEP as % of relative fluorescein retention (RFR). HCE monolayers stimulated with TNFα, TGFβ, IL-13, IL-17, and IL-6 show significantly decreased resistance and increased permeability (n = 4).

FIGURE 7 Abrogation by IL-10 of the TNFα-induced decrease in TER and TEP values in HCE cells. TER and TEP values sig-nificantly increase after the co-stimulation when compared with those obtained after TNFα stimulation alone (n = 3). (*p < 0.05 vs. C; #p < 0.05 vs. TNFα).

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Lav

al o

n 02

/27/

13Fo

r pe

rson

al u

se o

nly.

980 L. Contreras-Ruiz et al.

Current Eye Research

affect the function of TJs; and (3) increased ZO pro-teins expression. Also, TJ proteins changed their localization after stimulation from cell–cell contacts to a predominant cytosolic localization. To explain all these changes several mechanisms can be taken into account. Endocytosis may be at least partially responsible for such localization changes, because pro-inflammatory cytokines induce selective endo-cytosis of TJ, but not AJ proteins, in epithelial cells.21 In addition, phosphorylation regulates TJ protein cel-lular localization, and activation of protein kinase C (PKC) may also mediate TJ disassembly in confluent epithelial cell monolayers.17 Cytokines such as TNFα and TGFβ can activate the PKC activity through their intracellular signaling pathways.22–25 Also, those cytokines produce reorganization of the actin cytoskeleton.26,27 As a consequence, a loss of organiza-tion in TJ complexes may occur due to a sequence of events such as low OCLN levels, translocation of all TJ proteins from the membrane to the cytosol, reorganization of the cytoskeleton and finally, the incapability of ZO proteins to scaffold the complexes. The final consequence of all these changes would be a loss of barrier function, evaluated as decreased TER measurement and increased cell permeability. This possible explanation agrees with other reports in that IL-13 (Calonge M, et al. IOVS 2006;43:ARVO E-Abstract 4938) and TNFα28 are involved in the func-tional disruption of the epithelial barrier of the ocular surface. Regarding TGFβ, it is described as a pleiotro-pic molecule, with a tissue specific effect on epithelial barriers. In intestinal epithelial cells, TGFβ protects against barrier disruption,29 whereas in breast30 or Sertoli cells31 TGFβ disrupts the barrier, as in corneal epithelial cells.

Our results showed that IL-10 up-regulates all TJ proteins, including OCLN, with membrane localization, pointing out a good assembly of TJ complexes with no changes in barrier function. The potential protective effect of IL-10 in corneal barrier function was further evaluated after IL-10 and TNFα co-stimulation. Corneal barrier disruption produced by TNFα was significantly abrogated by IL-10. Therefore, according to our in vitro results, a protective role against TJ barrier disruption may be assigned to IL-10. IL-10 is an important immunoregulatory cytokine implicated in the limitation and termination of inflammatory responses.32 Recombinant human IL-10 is currently being tested in different clinical trials for rheumatoid arthritis, inflammatory bowel disease and psoriasis.32 Therefore, a therapeutic strategy based on IL-10 might be an option to preserve the corneal barrier during inflammatory conditions.

Finally, IL-17 and IL-6 induced an overexpression of all TJ proteins and E-cad. Only ZO-2 expression was significantly decreased. All proteins showed membrane localization. According to structural data, TJ complexes were well organized; however, functional

assays showed a loss of barrier function. IL-17 has already been associated with corneal epithelial barrier disruption in dry eye (DE).13 Antibody neutralization of IL-17 ameliorated experimental DE-induced corneal epithelial barrier dysfunction and decreased the expression of matrix metalloproteinases.13 ZO-2 has been previously shown to be essential for correct TJ and AJ assembly in epithelial cells. Targeted disruption of the ZO-2 protein in mouse epithelial cells resulted in deficient TJ formation.33 Therefore, the IL-17- and IL-6-induced decrease in ZO-2 expression might be responsible, at least in part, for loss of organization in TJ and AJ complexes and the consequent loss of barrier function.

It is important to point out that, in addition to ZO-1, ZO-2, CLDN-1, CLDN-2 and OCLN, there are more proteins implicated in TJ complex structure. Other transmembrane proteins, such as junctional adhesion molecules or claudins 3-16, or cytoplas-mic proteins, such as ZO-3, might play a role in the establishment and maintenance of corneal barrier function. In addition, this study was performed on corneal epithelial cell monolayers, where junctions are restricted to a single layer of cells and might behave in a different way from the real corneal stratified epithe-lium.34 However, the information in the literature so far comes mainly from monolayer models.18,28,29 More complex and representative in vitro study models are warranted.

In conclusion, the tested cytokines alter TJ and, to a lesser extent, AJ composition in human corneal epithelial cells and, consequently, alter the corneal epithelial barrier function in vitro. TNFα, TGFβ, IL-13, IL-17 and IL-6 induce a structural and functional disruption of the epithelial bar-rier, whereas IL-10 reveals barrier protective effects. More studies, including the evaluation of phosphorylation in TJ and AJ complexes components, are necessary to clarify the regulation of the structure, and consequently the func-tion, of the corneal epithelial barrier.

ACKNOWLEDGMENTS

We thank Itziar Fernández, M.Sc. for statistical analysis advice, Susan Möschter for technical assistance, and the confocal microscopy service from the University of Valladolid, especially Sagrario Callejo for her technical support.

Declaration of interest: This study was supported by grants from the Ministry of Education and Science, Spain (FEDER-CICYT MAT2007–64626-C02-01), also by FPU-MEC, JCyL-Regional, and FPI-University of Valladolid Scholarships Programs, Spain, by DFG grant PA738/9-2 and BMBF Roux program grants FKZ 9/18,12/08,13/08, Germany, and by Bilateral Research Grant Spain/Germany DE2009-0085 and DAAD ID6234017.

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Lav

al o

n 02

/27/

13Fo

r pe

rson

al u

se o

nly.

Corneal Barrier Under Inflammatory Conditions 981

© 2012 Informa Healthcare USA, Inc.

REFERENCES

[1] Kinoshita S, Adachi W, Sotozono C et al. Characteristics of the human ocular surface epithelium. Prog Retin Eye Res 2001;20:639–673.

[2] Hemady R, Chu W, Foster CS. Keratoconjunctivitis sicca and corneal ulcers. Cornea 1990;9:170–173.

[3] Matter K, Balda MS. Occludin and the functions of tight junc-tions. Int Rev Cytol 1999;186:117–146.

[4] Morita K, Furuse M, Fujimoto K et al. Claudin multigene family encoding four-transmembrane domain protein components of tight junction strands. Proc Natl Acad Sci USA 1999;96:511–516.

[5] Denker BM, Nigam SK. Molecular structure and assembly of the tight junction. Am J Physiol 1998;274:F1–F9.

[6] Tsukita S, Furuse M, Itoh M. Multifunctional strands in tight junctions. Nat Rev Mol Cell Biol 2001;2:285–293.

[7] Ban Y, Dota A, Cooper LJ et al. Tight junction-related protein expression and distribution in human corneal epithelium. Exp Eye Res 2003;76:663–669.

[8] Meng W, Takeichi M. Adherens junction: molecular archi-tecture and regulation. Cold Spring Harb Perspect Biol 2009;1:a002899.

[9] Pflugfelder SC. Anti-inflammatory therapy of dry eye. Ocul Surf 2003;1:31–36.

[ 10] Enriquez-de-Salamanca A, Calonge M. Cytokines and che-mokines in immune-based ocular surface inflammation. Expert Rev Clin Immunol 2008;4:457–467.

[ 11] Göbbels M, Spitznas M. Corneal epithelial permeability of dry eyes before and after treatment with artificial tears. Ophthalmology 1992;99:873–878.

[ 12] Enriquez-de-Salamanca A, Calder V, Gao J et al. Cytokine responses by conjunctival epithelial cells: an in vitro model of ocular inflammation. Cytokine 2008;44:160–167.

[ 13] De Paiva CS, Chotikavanich S, Pangelinan SB et al. IL-17 dis-rupts corneal barrier following desiccating stress. Mucosal Immunol 2009;2:243–253.

[ 14] Araki-Sasaki K, Ohashi Y, Sasabe T et al. An SV40-immortalized human corneal epithelial cell line and its char-acterization. Invest Ophthalmol Vis Sci 1995;36:614–621.

[ 15] Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970;227:680–685.

[ 16] Towbin H, Staehelin T, Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 1979;76:4350–4354.

[ 17] Harhaj NS, Antonetti DA. Regulation of tight junctions and loss of barrier function in pathophysiology. Int J Biochem Cell Biol 2004;36:1206–1237.

[ 18] Capaldo CT, Nusrat A. Cytokine regulation of tight junctions. Biochim Biophys Acta 2009;1788:864–871.

[ 19] Ohkubo T, Ozawa M. The transcription factor Snail down-regulates the tight junction components independently of E-cadherin downregulation. J Cell Sci 2004;117:1675–1685.

[ 20] Carayol N, Vachier I, Campbell A et al. Regulation of E-cadherin expression by dexamethasone and tumour necro-sis factor-α in nasal epithelium. Eur Respir J 2002;20:1430–1436.

[ 21] Ivanov AI, Nusrat A, Parkos CA. The epithelium in inflam-matory bowel disease: potential role of endocytosis of junc-tional proteins in barrier disruption. Novartis Found Symp 2004;263:115–24; discussion 124.

[ 22] Jain N, Zhang T, Kee WH et al. Protein kinase C delta associ-ates with and phosphorylates Stat3 in an interleukin-6-de-pendent manner. J Biol Chem 1999;274:24392–24400.

[ 23] Sakaguchi M, Miyazaki M, Sonegawa H et al. PKCα medi-ates TGFβ-induced growth inhibition of human kerati-nocytes via phosphorylation of S100C/A11. J Cell Biol 2004;164:979–984.

[ 24] Wyatt TA, Ito H, Veys TJ, et al. Stimulation of protein kinase C activity by tumor necrosis factor-α in bovine bronchial epi-thelial cells. Am J Physiol 1997;273:L1007–L1012.

[ 25] Xu B, Bhattacharjee A, Roy B et al. Role of protein kinase C isoforms in the regulation of interleukin-13-induced 15-lipox-ygenase gene expression in human monocytes. J Biol Chem 2004;279:15954–15960.

[ 26] Koukouritaki SB, Vardaki EA, Papakonstanti EA, et al. TNF-α induces actin cytoskeleton reorganization in glomerular epi-thelial cells involving tyrosine phosphorylation of paxillin and focal adhesion kinase. Mol Med 1999;5:382–392.

[ 27] Moustakas A, Stournaras C. Regulation of actin organisation by TGF-β in H-ras-transformed fibroblasts. J Cell Sci 1999;112 (Pt 8):1169–1179.

[ 28] Kimura K, Teranishi S, Fukuda K, et al. Delayed disruption of barrier function in cultured human corneal epithelial cells induced by tumor necrosis factor-α in a manner dependent on NF-κB. Invest Ophthalmol Vis Sci 2008;49:565–571.

[ 29] Planchon SM, Martins CA, Guerrant RL et al. Regulation of intestinal epithelial barrier function by TGF-β 1. Evidence for its role in abrogating the effect of a T cell cytokine. J Immunol 1994;153:5730–5739.

[ 30] Woo PL, Cha HH, Singer KL et al. Antagonistic regulation of tight junction dynamics by glucocorticoids and transforming growth factor-β in mouse mammary epithelial cells. J Biol Chem 1996;271:404–412.

[ 31] Lui WY, Lee WM, Cheng CY. Transforming growth factor-β3 perturbs the inter-Sertoli tight junction permeability bar-rier in vitro possibly mediated via its effects on occlu-din, zonula occludens-1, and claudin-11. Endocrinology 2001;142:1865–1877.

[ 32] Asadullah K, Sterry W, Volk HD. Interleukin-10 therapy–review of a new approach. Pharmacol Rev 2003;55:241–269.

[ 33] Umeda K, Ikenouchi J, Katahira-Tayama S et al. ZO-1 and ZO-2 independently determine where claudins are polymerized in tight-junction strand formation. Cell 2006;126:741–754.

[ 34] Bhabra G, Sood A, Fisher B et al. Nanoparticles can cause DNA damage across a cellular barrier. Nat Nanotechnol 2009;4:876–883.

Cur

r E

ye R

es D

ownl

oade

d fr

om in

form

ahea

lthca

re.c

om b

y U

nive

rsity

of

Lav

al o

n 02

/27/

13Fo

r pe

rson

al u

se o

nly.