Free
Glaucoma  |   August 2012
Toxicity Evaluation of Antiglaucoma Drugs Using Stratified Human Cultivated Corneal Epithelial Sheets
Author Affiliations & Notes
  • Suguru Nakagawa
    From the Department of Ophthalmology, University of Tokyo School of Medicine, Bunkyo-ku, Tokyo, Japan;
    Tokyo Metropolitan Geriatric Hospital, Itabashi-ku, Tokyo, Japan;
  • Tomohiko Usui
    From the Department of Ophthalmology, University of Tokyo School of Medicine, Bunkyo-ku, Tokyo, Japan;
  • Seiichi Yokoo
    From the Department of Ophthalmology, University of Tokyo School of Medicine, Bunkyo-ku, Tokyo, Japan;
  • Sachiko Omichi
    From the Department of Ophthalmology, University of Tokyo School of Medicine, Bunkyo-ku, Tokyo, Japan;
    JR Tokyo General Hospital, Shibuya-ku, Tokyo, Japan;
  • Mikiko Kimakura
    From the Department of Ophthalmology, University of Tokyo School of Medicine, Bunkyo-ku, Tokyo, Japan;
  • Yosai Mori
    Department of Ophthalmology, Miyata Eye Hospital, Miyakonojo, Miyazaki, Japan; and
  • Kazunori Miyata
    Department of Ophthalmology, Miyata Eye Hospital, Miyakonojo, Miyazaki, Japan; and
  • Makoto Aihara
    From the Department of Ophthalmology, University of Tokyo School of Medicine, Bunkyo-ku, Tokyo, Japan;
  • Shiro Amano
    From the Department of Ophthalmology, University of Tokyo School of Medicine, Bunkyo-ku, Tokyo, Japan;
  • Makoto Araie
    From the Department of Ophthalmology, University of Tokyo School of Medicine, Bunkyo-ku, Tokyo, Japan;
    Kanto Central Hospital, Setagaya-ku, Tokyo, Japan.
  • Corresponding author: Tomohiko Usui, Department of Ophthalmology, University of Tokyo School of Medicine, Bunkyo-ku, Tokyo, Japan; tomohiko-tky@umin.ac.jp
Investigative Ophthalmology & Visual Science August 2012, Vol.53, 5154-5160. doi:10.1167/iovs.12-9685
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Suguru Nakagawa, Tomohiko Usui, Seiichi Yokoo, Sachiko Omichi, Mikiko Kimakura, Yosai Mori, Kazunori Miyata, Makoto Aihara, Shiro Amano, Makoto Araie; Toxicity Evaluation of Antiglaucoma Drugs Using Stratified Human Cultivated Corneal Epithelial Sheets. Invest. Ophthalmol. Vis. Sci. 2012;53(9):5154-5160. doi: 10.1167/iovs.12-9685.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose.: To investigate the toxicity profiles of seven antiglaucoma topical eye drops and benzalkonium chloride (BAC) using stratified cultivated human corneal epithelial cell sheets (HCES) in a serum-free culture system.

Methods.: A range of prostaglandin analogies and preservatives, including BAC, sofZia (SZ), sodium benzoate (SB), and polyquaternium-1 (PQ) were tested. The barrier function and cell viability were examined by a carboxyfluorescein permeability assay and WST-1 assay. Histological evaluation of the HCES was also performed after application of each solution.

Results.: The carboxyfluorescein permeability assay had a higher sensitivity for the detection of toxicity of test solutions than the WST-1 assay or histological examination. Latanoprost BAC, latanoprost/timolol BAC, and 0.02% or higher concentration of BAC were the most toxic, followed by latanoprost SB, latanoprost preservative-free, BAC 0.002%, and travoprost/ latanoprost PQ. Travoprost SZ and tafluprost BAC (preserved with 0.001% BAC) was the least toxic in our experimental conditions.

Conclusions.: The carboxyfluorescein permeability assay using HCES in a serum-free system was the most useful for the quantification of toxicity of ophthalmic solutions. Among the regimens examined, a BAC concentration of 0.001% or lower or non-BAC preservative sofZia was suggested to be the least toxic to the ocular surface.

Introduction
Medical therapy using topical eye drops is a mainstay for treatment of glaucoma patients; however, ophthalmologists often encounter corneal epithelial disturbances (toxic keratoepitheliopathy) due to side effects of antiglaucoma eye drops. The side effects of eye drops are attributable to the active components as well as to preservatives. Recently, special attention has been paid to the preservatives, and benzalkonium chloride (BAC), which is the most frequently used preservative in these preparations, has been reported to be especially toxic to the ocular surface. 118 The current strategy to develop new formulae is to reduce BAC concentrations or to adopt an alternative safe preservative. 
To evaluate the toxicity of topical ophthalmic formulations in vivo, the conventional method is to use live rabbit eyes (the Draize test); however, this model imposes discomfort and pain and quantification of toxicity in this model is limited. Furthermore, the responses of the rabbit eyes are not always comparable to those of human eyes. 19,20 A useful alternative to the reliance on the animal model is in vitro cell culture assays. In previous studies, corneal epithelial and conjunctival cell lines have been widely used for ocular toxicology, 5,10,12,14,15,2125 but these cell lines have a single layer and cannot mimic the stratified corneal and conjunctival epithelium. Barrier function is one of the important functions of the corneal epithelium, which is well known to be compromised in eye drop–induced keratopathy. 2629 However, it is impossible to evaluate barrier function using a cell monolayer. Although a three-dimensional stratified corneal epithelial cell model has been established, 17,3035 a difference from in vivo human corneal epithelial cells still exists because this model also relies on immortalized cells. 
Recently, we have established primary cultivated human corneal epithelial cell sheets (HCES) in a serum-free culture system, 36 and these HCES have been used for the reconstruction of ocular surface diseases such as limbal deficiency. HCES showed well-formed stratified epithelium with clear cell boundaries, microvilli, and hemidesmosomal/ desmosomal junctions and expressed cytokeratin-3 and −12. 36 Because the HCES form layers four to seven cells thick and mimics closely the in vivo stratified corneal epithelium in humans, they also provide us with a good system in which to estimate cellular toxicity and viability. Furthermore, as HCES are cultured in culture inserts, we can investigate barrier function quantitatively, as well as examine the morphology of HCES. The aim of the present study was to quantitatively examine the toxicity of commercially available antiglaucoma eye drops and that of various concentrations of BAC as controls, using HCES that mimic the in vivo stratified human corneal epithelium, focusing not only on histological effects, but also on barrier function. 
Materials and Methods
Preparation of Stratified Cultivated Human Epithelial Corneal Sheets
HCES were cultivated as previously described. 36 In brief, human corneas were obtained from an eye bank (SightLife, Seattle, WA). The limbal tissues were cut carefully from the corneoscleral tissues, which were incubated overnight at 37°C in basal culture medium (Dulbecco's modified Eagle's medium [DMEM]-F12; Wako, Tokyo, Japan) with 0.02% type 1A collagenase (Wako Co., Ltd., Tokyo, Japan). Cells were collected from incubated tissues in coated tubes (Sumilon Stem Full; Sumitomo Bakelite Co., Ltd., Tokyo, Japan), allowed to stand in 0.05% trypsin/EDTA for 10 minutes at 37°C, and then dissociated into single cells by pipetting. After washing the trypsin/EDTA with medium, the cells were resuspended in medium with 20 ng/mL of EGF and B27 supplement (Invitrogen, Inc., Tokyo, Japan). Next, epithelial cells (2.4 × 104 cells/well) were seeded onto the culture inserts of 12-well culture dishes (Corning Transwell catalog #3460; Corning, Inc., Lowell, MA). The cells were covered with medium in the culture insert and kept at 37°C in an atmosphere of 5% CO2/95% air without air lifting. The medium was changed daily and the HCES were grown for 3 weeks. 
To test a variation of cell sheets among wells, histological examination was performed every one well per dish for confirmation of the cell sheets. 
Antiglaucoma Drugs and BAC Preparation
We analyzed the effects of seven antiglaucoma drugs: preservative-free 0.005% latanoprost (Latanoprost PF; Nitten Laboratories, Inc., Nagoya, Japan), 0.005% latanoprost preserved with sodium benzoate (latanoprost SB; Latanoprost, Nitten), 0.004% travoprost preserved with sofZia (travoprost SZ; TravatanZ, Alcon Laboratories, Inc., Fort Worth, TX), 0.004% travoprost/0.5% timolol fixed combination preserved with polyquaternium-1 0.001% (travoprost/timolol PQ; DuoTrav, Alcon), 0.015% tafluprost preserved with 0.001% BAC (tafluprost BAC; Tapros, Santen Pharmaceutical Co., Ltd., Osaka, Japan), 0.005% latanoprost preserved with 0.02% BAC (latanoprost BAC; Xalatan, Pfizer, Inc., New York, NY), 0.005% latanoprost/0.5% timolol fixed combination preserved with 0.02% BAC (latanoprost/timolol BAC; Xalacom, Pfizer). Each drug was applied in its commercial presentation. We also used five concentrations (0.0002%, 0.002%, 0.01%, 0.02%, and 0.2%) of BAC (catalog #B6295; Sigma-Aldrich Japan, Inc., Tokyo, Japan) diluted in PBS, which contains predominantly C12H25N(CH3)2C7H7Cl and also contains C14 and C16 analogs according to the manufacturer's data sheet. 
Barrier Function by Carboxyfluorescein Permeability Assay
The barrier function of HCES was evaluated using the permeability to carboxyfluorescein. Carboxyfluorescein is similar in molecular size and fluorescence to fluorescein, but it is approximately 1000 times more hydrophilic and is thought to pass through the cell layer almost exclusively through the paracellular spaces, indicating that carboxyfluorescein is more sensitive to changes in barrier functions than fluorescein. 3744 We prepared HCES in the upper chamber of 12 well plates (Corning Transwell catalog #3460; Corning, Inc., Lowell, MA) with 500 μL of DMEM-F12 medium in the upper chamber and 1 mL in the lower chamber. After three PBS washes of the upper chamber, seven ophthalmic solutions and five concentrations of BAC (0.0002%, 0.002%, 0.01%, 0.02%, and 0.2%) were introduced to the upper chamber for 10 minutes or 6 hours. After three PBS washes, 500μL of 50 μM 5(6)-carboxyfluorescein (catalog #21877; Sigma-Aldrich Japan, Inc.) in DMEM-F12 medium (pH 7.4) was transferred to the upper chamber and incubated for 2 hours. The fluorescence intensity of the outer dish was measured 2 hours later at emission 532 nm on a plate reader (Victor 3V Multilable Counter model 1420; Perkin Elmer, Waltham, MA). The results are expressed as a percentage of fluorescence intensity compared with that of the PBS-treated control. The results are expressed as the mean ± SD. 
Viability by WST-1 Assay
Cell viability was evaluated by a nonradioactive colorimetric assay (WST-1; Takara Bio, Inc., Shiga, Japan), based on the cleavage of a tetrazolium salt, as recommended by the manufacturer's protocol. This assay detects only living cells, and the signal generated is directly proportional to the number of live cells. After exposure to seven ophthalmic solutions and two concentrations of BAC (0.02% and 0.2%) for 10 minutes and 6 hours, WST-1 reagent in DMEM-F12 medium was applied for 1 hour at 37°C, and the dye was measured at 450 nm by a plate reader (Victor 3V Multilable Counter model 1420: Perkin Elmer, Waltham, MA). The results were expressed as the mean ± SD of a percentage of cell viability compared with that of the PBS-treated control. 
Histological Examination
After exposure to seven ophthalmic solutions and PBS for 24 hours, HCES were fixed in 4% buffered paraformaldehyde, and the epithelial samples were cut and embedded in paraffin. Then, 5-μm sections were cut from each paraffin block and placed on microscope slides. These sections were stained with hematoxylin and eosin (HE) and observed under a light microscope (BX51; Olympus Optical Co., Ltd., Tokyo, Japan). 
ZO-1 Expression
After exposure to the seven ophthalmic solutions and PBS for 6 hours, HCES were fixed in 4% buffered paraformaldehyde, and then were incubated overnight at 4°C with rabbit anti-ZO-1 primary antibody at a 1:200 dilution (Zymed Laboratories, Inc., San Francisco, CA). After three washes in PBS, HCES were incubated again for 1 hour in the dark at room temperature with Alexa594-conjugated goat anti-rabbit antibody at a 1:500 dilution (Invitrogen Molecular Probes, Eugene, OR). After three washes in PBS, the nuclei were labeled with propidium iodide or 4′,6-diamidino-2-phenylindole and mounted in an antifade medium (Vectashield: Vector Laboratories, Inc., Burlingame, CA). The samples were analyzed under a fluorescence microscope (Biorevo BZ-9000; Keyence Co., Tokyo, Japan). 
Statistical Analysis
Statistical comparisons between control and treatment groups for carboxyfluorescein permeability data and WST-1 data were performed using a Mann-Whitney U test with a Bonferroni correction (12 comparisons in 13 groups for carboxyfluorescein permeability data and 9 comparisons in 10 groups for WST-1 data) (JMP software; Cary, NC). 
Results
Barrier Function of HCES as Assessed by the Carboxyfluorescein Permeability Test
In the control (PBS)-treated group, the fluorescent intensity of outer dish was 100% ± 13.02% (mean ± SD) and variance was 169.47 in the 10-minute PBS load group, and 100% ± 25.98% and 674.70, respectively, in the 6-hour PBS load group. 
Ten minutes after application of some of the test drugs, the fluorescence intensity of the outer dish had significantly increased. The barrier function of the HCES treated with 0.01%, 0.02%, and 0.2% BAC, and latanoprost BAC, and latanoprost/timolol BAC was significantly impaired compared with PBS-treated control sheets in a BAC-dependent manner (P < 0.05 versus control, Fig. 1a). The lowest membrane integrity was obtained with latanoprost/timolol BAC (278.18% ± 159.52% fluorescence intensity of control; N = 8 in each group; P = 0.01 versus control). 
Figure 1. 
 
Comparison of barrier function by a carboxyfluorescein permeability assay after 10 minutes (a) and 6 hours (b) treatment with commercially available antiglaucoma drugs and BAC. Barrier function of cell sheets was significantly decreased with 0.01% BAC, 0.02% BAC, 0.2% BAC, latanoprost BAC, and latanoprost/timolol BAC compared with PBS-treated control sheets in a BAC-dependent manner (*P < 0.05 versus control) after 10-minute exposure (a). Barrier function of cell sheets significantly decreased with 0.002% BAC, 0.01% BAC, 0.02% BAC, 0.2% BAC, travoprost/timolol PQ, latanoprost PF, latanoprost SB, latanoprost BAC, and latanoprost/timolol BAC compared with PBS-treated control sheets (*P < 0.05 versus control ) after 6-hour exposure (b). Each bar represents the fluorescence intensity of carboxyfluorescein in the outer dish that flows out of the insert culture, normalized to the mean value obtained using PBS (N = 8). The significance of differences between control (PBS) and test values was determined by Mann-Whitney U test with Bonferroni correction (*P < 0.05) (12 comparisons in 13 groups).
Figure 1. 
 
Comparison of barrier function by a carboxyfluorescein permeability assay after 10 minutes (a) and 6 hours (b) treatment with commercially available antiglaucoma drugs and BAC. Barrier function of cell sheets was significantly decreased with 0.01% BAC, 0.02% BAC, 0.2% BAC, latanoprost BAC, and latanoprost/timolol BAC compared with PBS-treated control sheets in a BAC-dependent manner (*P < 0.05 versus control) after 10-minute exposure (a). Barrier function of cell sheets significantly decreased with 0.002% BAC, 0.01% BAC, 0.02% BAC, 0.2% BAC, travoprost/timolol PQ, latanoprost PF, latanoprost SB, latanoprost BAC, and latanoprost/timolol BAC compared with PBS-treated control sheets (*P < 0.05 versus control ) after 6-hour exposure (b). Each bar represents the fluorescence intensity of carboxyfluorescein in the outer dish that flows out of the insert culture, normalized to the mean value obtained using PBS (N = 8). The significance of differences between control (PBS) and test values was determined by Mann-Whitney U test with Bonferroni correction (*P < 0.05) (12 comparisons in 13 groups).
At 6 hours after application of each drug, the fluorescence intensity of the outer dish had also significantly increased with 0.002%, 0.01%, 0.02%, and 0.2% BAC, and with travoprost/timolol PQ, latanoprost PF, latanoprost SB, latanoprost BAC, and latanoprost/timolol BAC compared with PBS-treated control sheets (P < 0.05 versus control, Fig. 1b). The lowest membrane integrity treated with commercially available ophthalmic solutions was obtained with latanoprost/timolol BAC (299.97% ± 46.34% of control; P = 0.01 versus control). 
Cell Viability Measured by WST-1 Assay
After 10 minutes, cellular viability significantly was decreased with latanoprost/timolol BAC, latanoprost BAC, and 0.02% BAC compared with PBS-treated control cells (P < 0.05 versus control, Fig. 2a). The lowest cell viability was obtained with latanoprost/timolol BAC (46.86% ± 11.21% of control; N = 8 in each group; P = 0.008 versus control). 
Figure 2. 
 
Comparison of viability with a WST-1 assay after 10 minutes (a) and 6 hours (b) of treatment with commercially available antiglaucoma drugs and BAC. Cellular viability significantly decreased with 0.02% BAC, latanoprost BAC, and latanoprost/timolol BAC after 10 minutes of exposure (*P < 0.05 versus control) (a). Cellular viability significantly decreased with 0.02% BAC, travoprost/timolol PQ, latanoprost BAC, and latanoprost/ timolol BAC after 6 hours of exposure (*P < 0.05 versus control) (b). Each bar represents the cellular viability normalized to the mean value obtained using PBS (N = 8). The significance of differences between control (PBS) and test values was determined by Mann-Whitney U test with Bonferroni correction (*P < 0.05) (9 comparisons in 10 groups).
Figure 2. 
 
Comparison of viability with a WST-1 assay after 10 minutes (a) and 6 hours (b) of treatment with commercially available antiglaucoma drugs and BAC. Cellular viability significantly decreased with 0.02% BAC, latanoprost BAC, and latanoprost/timolol BAC after 10 minutes of exposure (*P < 0.05 versus control) (a). Cellular viability significantly decreased with 0.02% BAC, travoprost/timolol PQ, latanoprost BAC, and latanoprost/ timolol BAC after 6 hours of exposure (*P < 0.05 versus control) (b). Each bar represents the cellular viability normalized to the mean value obtained using PBS (N = 8). The significance of differences between control (PBS) and test values was determined by Mann-Whitney U test with Bonferroni correction (*P < 0.05) (9 comparisons in 10 groups).
After 6 hours, cellular viability had significantly decreased with latanoprost/timolol BAC, latanoprost BAC, 0.02% BAC, and travoprost/timolol PQ compared with PBS-treated control cells (P < 0.05 versus control, Fig. 2b). The lowest cell viabilities were obtained with latanoprost/timolol BAC (1.05% ± 0.64% of control; P = 0.001 versus control), latanoprost BAC (0.088% ± 0.19% of control; P = 0.008 versus control), and 0.02% BAC (0.19% ± 0.31% of control; P = 0.008 versus control). 
Histological Analysis
After 24 hours of treatment with latanoprost BAC and latanoprost/timolol BAC, the epithelial thickness was reduced and the presence of an increasing number of dead cells from the apical to the basal layers, modification of the superficial layer continuity, nucleus condensation, and vacuole formation in the superficial epithelial layer was observed compared with the control (Fig. 3). The other five drugs, latanoprost SB, latanoprost PF, travoprost SZ, tafluprost BAC, and travoprost/ timolol PQ, apparently reduced epithelial thickness compared with controls at 24 hours, but the damage was less than that caused by latanoprost BAC and latanoprost/timolol BAC. 
Figure 3. 
 
Histomorphological analysis of HCES by HE staining after 24 hours of treatment with commercially available antiglaucoma drugs. Normal HCES morphology was observed before treatment: the cell sheets showed multilayered corneal epithelium composed of a basal layer with cuboidal cells, intermediate layers with lateral elongated cells similar to the wing cells, and a surface layer with flattened superficial cells on HE staining. After 24 hours of treatment, latanoprost BAC and latanoprost/timolol BAC induced modifications in the continuity of the superficial morphology with the appearance of tissue necrosis and a reduced thickness compared with the control. The other drugs produced less tissue necrosis and reduced thickness than did latanoprost BAC and latanoprost/timolol BAC. Bar = 50 μm.
Figure 3. 
 
Histomorphological analysis of HCES by HE staining after 24 hours of treatment with commercially available antiglaucoma drugs. Normal HCES morphology was observed before treatment: the cell sheets showed multilayered corneal epithelium composed of a basal layer with cuboidal cells, intermediate layers with lateral elongated cells similar to the wing cells, and a surface layer with flattened superficial cells on HE staining. After 24 hours of treatment, latanoprost BAC and latanoprost/timolol BAC induced modifications in the continuity of the superficial morphology with the appearance of tissue necrosis and a reduced thickness compared with the control. The other drugs produced less tissue necrosis and reduced thickness than did latanoprost BAC and latanoprost/timolol BAC. Bar = 50 μm.
ZO-1 Protein Expression
The epithelium showed a fine ZO-1 staining in the most superficial large cells, forming a continuous ring around them with PBS, latanoprost SB, latanoprost PF, travoprost SZ, tafluprost BAC, and travoprost/timolol PQ. Confocal microscopic analysis of the ZO-1 protein showed it to be significantly decreased with latanoprost BAC and latanoprost/timolol BAC 6 hours after treatment (Fig. 4). 
Figure 4. 
 
ZO-1 expression after 6h of treatment with commercially available antiglaucoma drugs. The epithelium showed a fine ZO-1 staining in the most superficial large cells, forming a continuous ring around them with PBS, latanoprost SB, latanoprost PF, travoprost SZ, tafluprost 0.001% BAC, and travoprost/timolol PQ. ZO-1 expression in the HCES disappeared after 6 hours of treatment with latanoprost BAC and latanoprost/timolol BAC. Bar = 50 μm.
Figure 4. 
 
ZO-1 expression after 6h of treatment with commercially available antiglaucoma drugs. The epithelium showed a fine ZO-1 staining in the most superficial large cells, forming a continuous ring around them with PBS, latanoprost SB, latanoprost PF, travoprost SZ, tafluprost 0.001% BAC, and travoprost/timolol PQ. ZO-1 expression in the HCES disappeared after 6 hours of treatment with latanoprost BAC and latanoprost/timolol BAC. Bar = 50 μm.
Discussion
According to the results of the carboxyfluorescein permeability assay and WST-1 assay, toxicity on the HCES appears to be dependent on BAC concentrations higher than 0.01%. Figures 1a and 2a illustrate that the cellular toxicity was apparent in latanoprost BAC (preserved with 0.02% BAC) and latanoprost/timolol BAC (preserved with 0.02% BAC) and in BAC concentrations of 0.01% or higher. Past studies have also demonstrated that toxic effects of antiglaucoma treatments increase, depending on BAC concentration, and BAC concentrations over 0.01% were associated with apparent cell death. 2,5,6,8,1012,14,31,3335 Otherwise, latanoprost PF, latanoprost SB, travoprost SZ, tafluprost BAC, travoprost/timolol PQ, and BAC concentration less than or equal to 0.002% showed no significant effects on the barrier function or cellular viability index at 10 minutes exposure (Fig. 1a). At 6 hours after application of each drug, the barrier function of the HCES significantly decreased with 0.002% BAC, travoprost/timolol PQ, latanoprost PF, and latanoprost SB compared with control (Fig. 1b). Cellular viability had significantly decreased with travoprost/timolol PQ compared with control at 6 hours (Fig. 2b). 
To develop less toxic alternatives to BAC, preservatives such as sofZia, sodium benzoate, or polyquad have been proposed and are commercially available. Travoprost/timolol PQ decreased barrier function and viability first after long-term exposure (Figs. 1b, 2b). Polyquad (polyquaternium) is a family of polycationic polymers. At least 37 different polymers exist under the polyquaternium designation but polyquanternium-1 is commonly used as a multipurpose solution for contact lens care, and has shown a good safety and tolerance profile compared with other multipurpose solutions. 45 Except at concentrations of 0.5% or above, PQ did not cause significant changes in the ocular surface compared with saline, whereas BAC induced apparent toxic effects at 0.1% and 0.5% concentrations, with a dramatic and extremely rapid destruction of ocular cells. 46 Although PQ is thought to be relatively safe, we note that this preservative dose have potent toxicity on the barrier function and cell viability after long-term exposure. 
SB is generally recognized as a safe preservative in foods. 47 Latanoprost SB did not decrease cellular viability (Fig. 2b) or alter histological integrity (Figs. 3, 4) after long-term exposure, although it did cause a mild disturbance in barrier function (Fig. 1b). 47 Thus SB should be recognized as a safer preservative for ocular formulations than BAC, but SB can have more toxic effects on the barrier function than sofZia. Furthermore, these results suggest that barrier function was more sensitive than WST-1 assay. 
Latanoprost PF decreased barrier function after long-term exposure (Fig. 1b), although latanoprost PF showed no or low toxicity on viability or morphology after long-term exposure (Figs. 2b, 3, 4). These unexpected results also suggest that other additives of latanoprost PF ophthalmic solution or latanoprost itself may be slightly toxic to the barrier function of corneal epithelium. 
We observed a discrepancy between barrier function and viability, where the barrier function with latanoprost PF and latanoprost SB was significantly decreased at 6 hours (Fig. 1b), although viability with latanoprost PF and latanoprost SB was not significantly decreased (Fig. 2b). This discrepancy likely indicates that barrier function and viability assess different toxicity profiles. 
Travoprost SZ did not show any toxicity even after long-term exposure (Figs. 1b, 2b). SofZia is composed of boric acid, propylene glycol, sorbitol, and zinc chloride and causes oxidative damage and subsequent death to organisms, such as most species of bacteria that lack the enzymes cytochrome oxidase or catalase, 48 which is compatible with the result currently obtained. Because human cells possess these enzymes, sofZia is apparently not toxic for humans. 48 Baudouin et al. 10 also demonstrated that travoprost SZ induced significantly less apoptosis and fewer changes in cell viability in a cultured conjunctival cell line, but this has not been demonstrated in corneal cells. In animal models, a once-daily application to live rabbit eyes also caused significantly less damage to the corneal epithelial cells and attenuated the reduction in goblet cells. 11  
Interestingly, tafluprost BAC (0.001%) did not show apparent toxicity even with long-term exposure (6 hours) despite containing BAC (0.001%; Figs. 1b, 2b). We speculate that this low toxicity of tafluprost BAC could be partly due to the low concentration of BAC in tafluprost (0.001%). In support of this speculation, low concentrations of BAC alone, such as 0.002% and 0.0002%, did not affect the HCES significantly compared with PBS (Figs. 1, 2). 
In this study, we examined the barrier function of HCEC against topical formulations and BAC solutions. To the best of our knowledge, this is the first study to compare effects of topical formulations on the barrier function of corneal epithelial cells. After a 6-hour exposure to each drug, the barrier function of HCES decreased with latanoprost PF, latanoprost SB, and BAC 0.002% also, although the cell viability determined with WST-1 and by histology was not altered (Figs. 1b, 2b, 3, 4). Thus, the barrier function assay using carboxyfluorescein may have a higher sensitivity than the other assays used here, and one advantage of this method is that the effect is quantitatively determined. 
Glaucoma patients in clinical settings often have abnormal corneal epithelium due to the complications of dry eyes and diabetes mellitus, the long-term use of eye drops preserved with BAC, and lower tear film disturbances due to the bleb formed by trabeculectomy. Thus, sensitive detection of the potential toxicity of eye drops is of clinical importance to avoid unnecessary insult to the ocular surface in such eyes. We suggest that the carboxyfluorescein permeability assay, which is more sensitive than the other assays in the present study, will be useful for this purpose. 
As reported previously by Friedlaender et al., 49 eye drops are washed out from the ocular surface within a few minutes. Thus, loading for a short time may be appropriate to represent the load of a single instillation. We performed the WST-1 assay and the carboxyfluorescein permeability assay with a 3-minute load at the beginning of the study, and obtained findings similar to those after a 10-minute load (data not shown), but a large variation occurred due to a technical problem of difficulty in treating all wells for strictly 3 minutes (104 wells in total). Because the results after a 3-minute load were similar to those after a 10-minute load, and the data variation was smaller, we assumed that the 10-minute load represents a single instillation. Because it was reported that BAC was retained in cornea after a single 30-μL drop of 0.01% BAC in rabbits 50 and glaucoma patients are usually treated for a prolonged period, a stress test may be also important with regard to the toxicity of long-term instillation. We also investigated the assay with 24-hour loading, but the barrier function and cell viability were severely damaged in some wells (data not shown). Thus, we considered that 6 hours is appropriate as a stress test in this experiment. 
There are several limitations to this study. First, we could not assess the effect of test solutions on the interaction between the tear film and HCES. Although BAC at concentrations of 0.001% or less was thought to not affect the barrier function or the viability of HCES, we could not exclude the possibility that at concentrations 0.001% or less, BAC may affect the tear film in vivo due to its detergent effect, and this would be also the case for other non-BAC preservatives. Second, we used eye drops in their commercial presentation in this study. Topical formulations contain active components and additives, such as preservatives and buffers. We could not identify the individual effects of each of these components in this study. Third, BAC is a mixture of N-alkyldimethylbenzylammonium chlorides with N-alkyl chain lengths ranging from C8 to C18. Commercially available BAC consists of C12, C14, and C16 homologs. Uematsu et al. 51 reported that C12 is the safest and C14 is more cytotoxic than C12 or C16 compounds; however, we were unable to obtain the same preservatives as those used in each commercially available product. According to the manufacturer's data sheet, the BAC used in this study consists of predominantly of C12, C14, and C16 compounds; however, the actual ratio of each compound is not shown. 
In conclusion, we found that HCES mimicking stratified human corneal epithelial cells was useful in quantitatively estimating functional and morphological effects of various ophthalmic formulations and preservatives on the corneal epithelium. BAC at a concentration of 0.001% or less was suggested to be safe. Ophthalmic solutions using non-BAC preservatives, such as polyquad, sodium benzoate, and sofZia, were safer than those using BAC, but these ophthalmic solutions were not free from toxicity. The ophthalmic solution using sofZia was thought to be the least toxic for ocular surfaces among those using the non-BAC preservatives presently studied. 
Acknowledgments
The authors thank IOVS volunteer editor Tim Corson (Indiana University School of Medicine, Indianapolis, IN) for editing the manuscript. 
References
Furrer P Mayer JM Gurny R. Ocular tolerance of preservatives and alternatives. Eur J Pharm Biopharm . 2002;53:263–280. [CrossRef] [PubMed]
Hamard P Blondin C Debbasch C Warnet JM Baudouin C Brignole F. In vitro effects of preserved and unpreserved antiglaucoma drugs on apoptotic marker expression by human trabecular cells. Graefes Arch Clin Exp Ophthalmol . 2003;241:1037–1043. [CrossRef] [PubMed]
Pisella PJ Debbasch C Hamard P Conjunctival proinflammatory and proapoptotic effects of latanoprost and preserved and unpreserved timolol: an ex vivo and in vitro study. Invest Ophthalmol Vis Sci . 2004;45:1360–1368. [CrossRef] [PubMed]
Whitson JT Cavanagh HD Lakshman N Petroll WM. Assessment of corneal epithelial integrity after acute exposure to ocular hypotensive agents preserved with and without benzalkonium chloride. Adv Ther . 2006;23:663–671. [CrossRef] [PubMed]
Yee RW Norcom EG Zhao XC. Comparison of the relative toxicity of travoprost 0.004% without benzalkonium chloride and latanoprost 0.005% in an immortalized human cornea epithelial cell culture system. Adv Ther . 2006;23:511–519. [CrossRef] [PubMed]
McCarey B Edelhauser H. In vivo corneal epithelial permeability following treatment with prostaglandin analogs [correction of analoges] with or without benzalkonium chloride. J Ocul Pharmacol Ther . 2007;23:445–451. [CrossRef] [PubMed]
Yee RW. The effect of drop vehicle on the efficacy and side effects of topical glaucoma therapy: a review. Curr Opin Ophthalmol . 2007;18:134–139. [CrossRef] [PubMed]
Guo Y Satpathy M Wilson G Srinivas SP. Benzalkonium chloride induces dephosphorylation of Myosin light chain in cultured corneal epithelial cells. Invest Ophthalmol Vis Sci . 2007;48:2001–2008. [CrossRef] [PubMed]
Malvitte L Montange T Vejux A Measurement of inflammatory cytokines by multicytokine assay in tears of patients with glaucoma topically treated with chronic drugs. Br J Ophthalmol . 2007;91:29–32. [CrossRef] [PubMed]
Baudouin C Riancho L Warnet JM Brignole F. In vitro studies of antiglaucomatous prostaglandin analogues: travoprost with and without benzalkonium chloride and preserved latanoprost. Invest Ophthalmol Vis Sci . 2007;48:4123–4128. [CrossRef] [PubMed]
Kahook MY Noecker RJ. Comparison of corneal and conjunctival changes after dosing of travoprost preserved with sofZia, latanoprost with 0.02% benzalkonium chloride, and preservative-free artificial tears. Cornea . 2008;27:339–343. [CrossRef] [PubMed]
Ayaki M Yaguchi S Iwasawa A Koide R. Cytotoxicity of ophthalmic solutions with and without preservatives to human corneal endothelial cells, epithelial cells and conjunctival epithelial cells. Clin Experiment Ophthalmol . 2008;36:553–559. [CrossRef] [PubMed]
Liang H Baudouin C Pauly A Brignole-Baudouin F. Conjunctival and corneal reactions in rabbits following short- and repeated exposure to preservative-free tafluprost, commercially available latanoprost and 0.02% benzalkonium chloride. Br J Ophthalmol . 2008;92:1275–1282. [CrossRef] [PubMed]
Epstein SP Chen D Asbell PA. Evaluation of biomarkers of inflammation in response to benzalkonium chloride on corneal and conjunctival epithelial cells. J Ocul Pharmacol Ther . 2009;25:415–424. [CrossRef] [PubMed]
Ammar DA Kahook MY. The effects of combination glaucoma medications on ocular surface epithelial cells. Adv Ther . 2009;26:970–975. [CrossRef] [PubMed]
Liang H Baudouin C Faure MO Lambert G Brignole-Baudouin F. Comparison of the ocular tolerability of a latanoprost cationic emulsion versus conventional formulations of prostaglandins: an in vivo toxicity assay. Mol Vis . 2009;15:1690–1699. [PubMed]
Baudouin C Labbe A Liang H Pauly A Brignole-Baudouin F. Preservatives in eyedrops: the good, the bad and the ugly. Prog Retin Eye Res . 2010;29:312–334. [CrossRef] [PubMed]
Whitson JT Trattler WB Matossian C Williams J Hollander DA. Ocular surface tolerability of prostaglandin analogs in patients with glaucoma or ocular hypertension. J Ocul Pharmacol Ther . 2010;26:287–292. [CrossRef] [PubMed]
Wilhelmus KR. The Draize eye test. Surv Ophthalmol . 2001;45:493–515. [CrossRef] [PubMed]
Curren RD Harbell JW. Ocular safety: a silent (in vitro) success story. Altern Lab Anim . 2002;30:69–74. [PubMed]
Guenoun JM Baudouin C Rat P Pauly A Warnet JM Brignole-Baudouin F. In vitro comparison of cytoprotective and antioxidative effects of latanoprost, travoprost, and bimatoprost on conjunctiva-derived epithelial cells. Invest Ophthalmol Vis Sci . 2005;46:4594–4599. [CrossRef] [PubMed]
Guenoun JM Baudouin C Rat P Pauly A Warnet JM Brignole-Baudouin F. In vitro study of inflammatory potential and toxicity profile of latanoprost, travoprost, and bimatoprost in conjunctiva-derived epithelial cells. Invest Ophthalmol Vis Sci . 2005;46:2444–2450. [CrossRef] [PubMed]
Brasnu E Brignole-Baudouin F Riancho L Guenoun JM Warnet JM Baudouin C. In vitro effects of preservative-free tafluprost and preserved latanoprost, travoprost, and bimatoprost in a conjunctival epithelial cell line. Curr Eye Res . 2008;33:303–312. [CrossRef] [PubMed]
Kahook MY Ammar DA. In vitro toxicity of topical ocular prostaglandin analogs and preservatives on corneal epithelial cells. J Ocul Pharmacol Ther . 2010;26:259–263. [CrossRef] [PubMed]
Ammar DA Noecker RJ Kahook MY. Effects of benzalkonium chloride-preserved, polyquad-preserved, and sofZia-preserved topical glaucoma medications on human ocular epithelial cells. Adv Ther . 2010;27:837–845. [CrossRef] [PubMed]
DeRosa AJ. Toxic keratopathy. Int Ophthalmol Clin . 1998;38:15–22. [CrossRef] [PubMed]
Chetoni P Burgalassi S Monti D Saettone MF. Ocular toxicity of some corneal penetration enhancers evaluated by electrophysiology measurements on isolated rabbit corneas. Toxicol In Vitro . 2003;17:497–504. [CrossRef] [PubMed]
Nakamura T Yamada M Teshima M Electrophysiological characterization of tight junctional pathway of rabbit cornea treated with ophthalmic ingredients. Biol Pharm Bull . 2007;30:2360–2364. [CrossRef] [PubMed]
Nakashima M Nakamura T Teshima M Breakdown evaluation of corneal epithelial barrier caused by antiallergic eyedrops using an electrophysiologic method. J Ocul Pharmacol Ther . 2008;24:43–51. [CrossRef] [PubMed]
Doucet O Lanvin M Thillou C Reconstituted human corneal epithelium: a new alternative to the Draize eye test for the assessment of the eye irritation potential of chemicals and cosmetic products. Toxicol In Vitro . 2006;20:499–512. [CrossRef] [PubMed]
Pauly A Meloni M Brignole-Baudouin F Warnet JM Baudouin C. Multiple endpoint analysis of the 3D-reconstituted corneal epithelium after treatment with benzalkonium chloride: early detection of toxic damage. Invest Ophthalmol Vis Sci . 2009;50:1644–1652. [CrossRef] [PubMed]
Khoh-Reiter S Jessen BA. Evaluation of the cytotoxic effects of ophthalmic solutions containing benzalkonium chloride on corneal epithelium using an organotypic 3-D model. BMC Ophthalmol . 2009;9:5. [CrossRef] [PubMed]
Meloni M Pauly A Servi BD Varlet BL Baudouin C. Occludin gene expression as an early in vitro sign for mild eye irritation assessment. Toxicol In Vitro . 2010;24:276–285. [CrossRef] [PubMed]
Liang H Pauly A Riancho L Baudouin C Brignole-Baudouin F. Toxicological evaluation of preservative-containing and preservative-free topical prostaglandin analogues on a three-dimensional-reconstituted corneal epithelium system. Br J Ophthalmol . 2011;95:869–875. [CrossRef] [PubMed]
Pauly A Brasnu E Riancho L Brignole-Baudouin F Baudouin C. Multiple endpoint analysis of BAC-preserved and unpreserved antiallergic eye drops on a 3D-reconstituted corneal epithelial model. Mol Vis . 2011;17:745–755. [PubMed]
Yokoo S Yamagami S Usui T Amano S Araie M. Human corneal epithelial equivalents for ocular surface reconstruction in a complete serum-free culture system without unknown factors. Invest Ophthalmol Vis Sci . 2008;49:2438–2443. [CrossRef] [PubMed]
Grimes PA Stone RA Laties AM Li W. Carboxyfluorescein. A probe of the blood-ocular barriers with lower membrane permeability than fluorescein. Arch Ophthalmol . 1982;100:635–639. [CrossRef] [PubMed]
Araie M. Carboxyfluorescein. A dye for evaluating the corneal endothelial barrier function in vivo. Exp Eye Res . 1986;42:141–150. [CrossRef] [PubMed]
Araie M. Barrier function of corneal endothelium and the intraocular irrigating solutions. Arch Ophthalmol . 1986;104:435–438. [CrossRef] [PubMed]
Tsuboi S Pederson JE. Permeability of the isolated dog retinal pigment epithelium to carboxyfluorescein. Invest Ophthalmol Vis Sci . 1986;27:1767–1770. [PubMed]
Blair NP Rusin MM. Blood-retinal barrier permeability to carboxyfluorescein and fluorescein in monkeys. Graefes Arch Clin Exp Ophthalmol . 1986;224:419–422. [CrossRef] [PubMed]
Grimes PA. Carboxyfluorescein transfer across the blood-retinal barrier evaluated by quantitative fluorescence microscopy: comparison with fluorescein. Exp Eye Res . 1988;46:769–783. [CrossRef] [PubMed]
Grimes PA. Carboxyfluorescein distribution in ocular tissues of normal and diabetic rats. Curr Eye Res . 1988;7:981–988. [CrossRef] [PubMed]
Kimura M Araie M Koyano S. Movement of carboxyfluorescein across retinal pigment epithelium-choroid. Exp Eye Res . 1996;63:51–56. [CrossRef] [PubMed]
Lipener C. A randomized clinical comparison of OPTI-FREE EXPRESS and ReNu MultiPLUS multipurpose lens care solutions. Adv Ther . 2009;26:435–446. [CrossRef] [PubMed]
Labbe A Pauly A Liang H Comparison of toxicological profiles of benzalkonium chloride and polyquaternium-1: an experimental study. J Ocul Pharmacol Ther . 2006;22:267–278. [CrossRef] [PubMed]
Nair B. Final report on the safety assessment of benzyl alcohol, benzoic acid, and sodium benzoate. Int J Toxicol . 2001;20:23–50. [CrossRef] [PubMed]
Kaur IP Lal S Rana C Kakkar S Singh H. Ocular preservatives: associated risks and newer options. Cutan Ocul Toxicol . 2009;28:93–103. [CrossRef] [PubMed]
Friedlaender MH Breshears D Amoozgar B Sheardown H Senchyna M. The dilution of benzalkonium chloride (BAK) in the tear film. Adv Ther . 2006;23:835–841. [CrossRef] [PubMed]
Champeau EEH. Effect of Ophthalmic Preservatives on the Ocular Surface: Conjunctival and Corneal Uptake and Distribution of Benzalkonium Chloride and Chlorhexidine Digluconate . Lubbock, TX.: Dry Eye Institute, Inc; 1986.
Uematsu M Kumagami T Shimoda K Influence of alkyl chain length of benzalkonium chloride on acute corneal epithelial toxicity. Cornea . 2010;29:1296–1301. [CrossRef] [PubMed]
Footnotes
 Disclosure: S. Nakagawa, None; T. Usui, None; S. Yokoo, None; S. Omichi, None; M. Kimakura, None; Y. Mori, None; K. Miyata, None; M. Aihara, None; S. Amano, None; M. Araie, None
Figure 1. 
 
Comparison of barrier function by a carboxyfluorescein permeability assay after 10 minutes (a) and 6 hours (b) treatment with commercially available antiglaucoma drugs and BAC. Barrier function of cell sheets was significantly decreased with 0.01% BAC, 0.02% BAC, 0.2% BAC, latanoprost BAC, and latanoprost/timolol BAC compared with PBS-treated control sheets in a BAC-dependent manner (*P < 0.05 versus control) after 10-minute exposure (a). Barrier function of cell sheets significantly decreased with 0.002% BAC, 0.01% BAC, 0.02% BAC, 0.2% BAC, travoprost/timolol PQ, latanoprost PF, latanoprost SB, latanoprost BAC, and latanoprost/timolol BAC compared with PBS-treated control sheets (*P < 0.05 versus control ) after 6-hour exposure (b). Each bar represents the fluorescence intensity of carboxyfluorescein in the outer dish that flows out of the insert culture, normalized to the mean value obtained using PBS (N = 8). The significance of differences between control (PBS) and test values was determined by Mann-Whitney U test with Bonferroni correction (*P < 0.05) (12 comparisons in 13 groups).
Figure 1. 
 
Comparison of barrier function by a carboxyfluorescein permeability assay after 10 minutes (a) and 6 hours (b) treatment with commercially available antiglaucoma drugs and BAC. Barrier function of cell sheets was significantly decreased with 0.01% BAC, 0.02% BAC, 0.2% BAC, latanoprost BAC, and latanoprost/timolol BAC compared with PBS-treated control sheets in a BAC-dependent manner (*P < 0.05 versus control) after 10-minute exposure (a). Barrier function of cell sheets significantly decreased with 0.002% BAC, 0.01% BAC, 0.02% BAC, 0.2% BAC, travoprost/timolol PQ, latanoprost PF, latanoprost SB, latanoprost BAC, and latanoprost/timolol BAC compared with PBS-treated control sheets (*P < 0.05 versus control ) after 6-hour exposure (b). Each bar represents the fluorescence intensity of carboxyfluorescein in the outer dish that flows out of the insert culture, normalized to the mean value obtained using PBS (N = 8). The significance of differences between control (PBS) and test values was determined by Mann-Whitney U test with Bonferroni correction (*P < 0.05) (12 comparisons in 13 groups).
Figure 2. 
 
Comparison of viability with a WST-1 assay after 10 minutes (a) and 6 hours (b) of treatment with commercially available antiglaucoma drugs and BAC. Cellular viability significantly decreased with 0.02% BAC, latanoprost BAC, and latanoprost/timolol BAC after 10 minutes of exposure (*P < 0.05 versus control) (a). Cellular viability significantly decreased with 0.02% BAC, travoprost/timolol PQ, latanoprost BAC, and latanoprost/ timolol BAC after 6 hours of exposure (*P < 0.05 versus control) (b). Each bar represents the cellular viability normalized to the mean value obtained using PBS (N = 8). The significance of differences between control (PBS) and test values was determined by Mann-Whitney U test with Bonferroni correction (*P < 0.05) (9 comparisons in 10 groups).
Figure 2. 
 
Comparison of viability with a WST-1 assay after 10 minutes (a) and 6 hours (b) of treatment with commercially available antiglaucoma drugs and BAC. Cellular viability significantly decreased with 0.02% BAC, latanoprost BAC, and latanoprost/timolol BAC after 10 minutes of exposure (*P < 0.05 versus control) (a). Cellular viability significantly decreased with 0.02% BAC, travoprost/timolol PQ, latanoprost BAC, and latanoprost/ timolol BAC after 6 hours of exposure (*P < 0.05 versus control) (b). Each bar represents the cellular viability normalized to the mean value obtained using PBS (N = 8). The significance of differences between control (PBS) and test values was determined by Mann-Whitney U test with Bonferroni correction (*P < 0.05) (9 comparisons in 10 groups).
Figure 3. 
 
Histomorphological analysis of HCES by HE staining after 24 hours of treatment with commercially available antiglaucoma drugs. Normal HCES morphology was observed before treatment: the cell sheets showed multilayered corneal epithelium composed of a basal layer with cuboidal cells, intermediate layers with lateral elongated cells similar to the wing cells, and a surface layer with flattened superficial cells on HE staining. After 24 hours of treatment, latanoprost BAC and latanoprost/timolol BAC induced modifications in the continuity of the superficial morphology with the appearance of tissue necrosis and a reduced thickness compared with the control. The other drugs produced less tissue necrosis and reduced thickness than did latanoprost BAC and latanoprost/timolol BAC. Bar = 50 μm.
Figure 3. 
 
Histomorphological analysis of HCES by HE staining after 24 hours of treatment with commercially available antiglaucoma drugs. Normal HCES morphology was observed before treatment: the cell sheets showed multilayered corneal epithelium composed of a basal layer with cuboidal cells, intermediate layers with lateral elongated cells similar to the wing cells, and a surface layer with flattened superficial cells on HE staining. After 24 hours of treatment, latanoprost BAC and latanoprost/timolol BAC induced modifications in the continuity of the superficial morphology with the appearance of tissue necrosis and a reduced thickness compared with the control. The other drugs produced less tissue necrosis and reduced thickness than did latanoprost BAC and latanoprost/timolol BAC. Bar = 50 μm.
Figure 4. 
 
ZO-1 expression after 6h of treatment with commercially available antiglaucoma drugs. The epithelium showed a fine ZO-1 staining in the most superficial large cells, forming a continuous ring around them with PBS, latanoprost SB, latanoprost PF, travoprost SZ, tafluprost 0.001% BAC, and travoprost/timolol PQ. ZO-1 expression in the HCES disappeared after 6 hours of treatment with latanoprost BAC and latanoprost/timolol BAC. Bar = 50 μm.
Figure 4. 
 
ZO-1 expression after 6h of treatment with commercially available antiglaucoma drugs. The epithelium showed a fine ZO-1 staining in the most superficial large cells, forming a continuous ring around them with PBS, latanoprost SB, latanoprost PF, travoprost SZ, tafluprost 0.001% BAC, and travoprost/timolol PQ. ZO-1 expression in the HCES disappeared after 6 hours of treatment with latanoprost BAC and latanoprost/timolol BAC. Bar = 50 μm.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×