Multiple endpoint analysis of BAC-preserved and unpreserved

antiallergic eye drops on a 3D-reconstituted corneal epithelial

model

A. Pauly,

1,2,3

E. Brasnu,

1,2,3,4

L. Riancho,

1,2,3

F. Brignole-Baudouin,

1,2,3,5

C. Baudouin

1,2,3,4,6

(The first two authors contributed equally to this work)

INSERM, UMR_S968, Institut de la Vision, Paris, France;

UPMC University Paris 06, UMR_S 968, Institut de la Vision, Paris,

France;

CNRS, UMR_7210, Paris,

France;

Department of Ophthalmology III, Quinze-Vingts National Ophthalmology Hospital,

Paris, France;

Department of Toxicology, Faculty of Biological and Pharmacological Sciences, Paris, France;

Ambroise Paré

Hospital, APHP, University of Versailles Saint-Quentin-en-Yvelines, Versailles, France

Purpose: To compare the effects of benzalkonium chloride (BAC)-preserved and unpreserved antiallergic eye drops on

the human 3D-reconstituted corneal epithelial model (3D-HCE).

Methods: 3D-HCE were treated for 24 h followed or not by a 24 h post-incubation recovery period (24 h+24 h) with

phosphate-buffered saline (PBS), 0.01% BAC, unpreserved formulations of ketotifen, N Acetyl-Aspartyl Glutamic Acid

(NAAGA), cromoglycate, or BAC-preserved commercial formulations of ketotifen, olopatadine, epinastine, and

levocabastine. The 3D-HCE viability was evaluated using the 3-(4,5-Dimethylthiazol-2-yl) -2,5-Diphenyltetrazolium

Bromide (MTT) test at 24 h and 24 h+24 h. At 24 h, the numbers of Cluster of Differentiation 54 (CD54)- and Ki67-

immunopositive cells as well as the number of apoptotic deoxynucleotidyl transferase-mediated dUTP nick-end labeling

(TUNEL)-positive cells were evaluated on 3D-HCE frozen sections. The expression of the tight junction-associated

protein occludin was also assessed using fluorescence confocal microscopy on flat-mounted 3D-HCE epithelia.

Results: The MTT and the TUNEL tests revealed a significant decrease of cell viability and an increased apoptosis in the

superficial layers of the 3D-HCE only when treated with BAC-containing formulations and in a BAC concentration-

dependent manner. The expression of CD54 and Ki67 in the basal layers was also increased in this group. A concentration-

dependent disorganization of occludin distribution in the epithelium treated with BAC-containing solutions was also

observed. The unpreserved formulations induced effects comparable to the control.

Conclusions: BAC-preserved solutions decreased cell viability and induced apoptosis in a concentration-dependent

manner. Moreover, they induced CD54 expression, proliferation in the basal layers, and changes in the distribution of

occludin, which is consistent with a disorganization of the tight-junctions and suggests the loss of the epithelial barrier

function. On the contrary, the unpreserved solutions did not impair cell structures and viability, suggesting a better

tolerance for the ocular surface. As allergic patients often exhibit impaired and inflammatory ocular surface, BAC-free

compounds should be the first choice when treating allergic conjunctivitis.

To limit and counteract the clinical manifestations of

allergic

diseases,

antiallergic

compounds can be used. One of

these molecules, ketotifen fumarate, has demonstrated both

H1-receptor antagonism and mast cell stabilizing properties

while inhibiting chemotaxis and eosinophil activation [1,2].

Moreover, ketotifen fumarate was shown to be well tolerated

and effective in reducing the signs and symptoms of allergic

conjunctivitis [3-6]. Allergic conjunctivitis, however, has

often a tendency to become chronic, due to repeated allergic

challenge or progressive impairment of the tear film and

ocular surface [7,8].

Correspondence to: Pr Christophe Baudouin, Department of

Ophthalmology III, Quinze-Vingts

National Ophthalmology

Hospital, 28, rue de Charenton, 75012, Paris, France ; Phone:

+33.1.40.02.13.01; FAX: +33.1.40.02.13.99 ; email:

[email protected]

As preservatives are usually used to prevent multidose

eyedrop microbial contamination,

their chronic

administration may cause further ocular surface changes, at

the levels of tear film and conjunctiva. They can induce

cytotoxic effects and deleterious reactions when used over

long-term periods. Indeed, the mostly used preservative

benzalkonium chloride (BAC) was already shown to exhibit

toxic and inflammatory effects in clinical, in vivo and in vitro

studies [9-20]. Chronic use of BAC in eye drops is known to

be responsible for apoptosis and oxidative stress on

conjunctival cells, and to induce conjunctival inflammation

that has demonstrated potentially harmful effects on glaucoma

outcome, e.g., on glaucoma surgery efficacy [17,21-25].

In this context, the implementation of very sensitive tools

to predict eye tolerance is critical for ophthalmologists, who

may be faced with long-term induced toxicity of substances

Molecular Vision 2011; 17:745-755 <http://www.molvis.org/molvis/v17/a85>

Received 1 February 2011 | Accepted 10 March 2011 | Published 16 March 2011

745

used at low concentration in ophthalmic preparations.

Supplied by SkinEthic

Laboratories (Nice, France), the

reconstructed three-dimensional (3D) model of human

corneal cells (3D-HCE) is an appropriate model for pre-

screening or investigating the undesirable effects of

ophthalmic drugs. It constitutes an interesting alternative to

animal testing that is time-consuming and often invasive and

may lack suited sensitive tools able to detect subclinical

reactions [26-28]. Multi-endpoint analyses using adapted and

improved techniques on such 3D-models have already proved

efficacy for the assessment of BAC toxicity [28] and eyedrop

tolerance [27].

The objective of this study was to investigate a large range

of commonly used antiallergic eye drops in this 3D-HCE

system and compare the tissue changes after treatment with

BAC-preserved commercial formulations of ketotifen,

olopatadine, epinastine or levocabastine, and unpreserved

commercial formulations of ketotifen, N Acetyl-Aspartyl

Glutamic Acid (NAAGA), or cromoglycate. Particularly, our

purpose was to determine the involvement of BAC in

epithelial cell damage induced after treatment with BAC-

preserved and unpreserved antiallergic eyedrops.

METHODS

Tissue model and antiallergic solution treatments: The 3D-

HCE model (SkinEthic

Laboratories, Nice, France) consists

of immortalized HCE cells grown vertically on a 0.5 cm

insert

permeable polycarbonate filter. All the experiments were

conducted as published previously [27-29]. Thirty microliters

of each solution were applied on the apical surface of 3D-

HCEs for 24 h and 24 h followed by 24 h additional recovery

time: sterile phosphate-buffered saline (PBS) used as negative

control solution, BAC solutions at 0.01% used as positive

control, the commercial solutions of 0.01% BAC-containing

ketotifen fumarate 0.025% (Zaditen

; Novartis Pharma SAS,

Rueil-Malmaison, France), 0.01% BAC-containing

olopatadine chlorhydrate 0.1% (Opatanol

; Patanol

; Alcon,

Ft. Worth, TX), 0.01% BAC-containing epinastine

chlorhydrate 0.05% (Purivist

; Allergan, Irvine, CA), 0.015%

BAC-containing levocabastine chlorhydrate 0.05%

(Levophta

; Chauvin Bausch & Lomb, Montpellier, France),

preservative-free ketotifen fumarate 0.025% (Zalerg

; Thea,

Clermont-Ferrand, France), preservative-free NAAGA 4.9%

(NAABAK

; Thea) and preservative-free sodium

cromoglycate 2% (Cromabak

; Thea; Table 1).

The recovery period (24 h) was chosen to assess the

potential reversibility of toxic effects on 3D-HCE. Six series

of 3D-HCE were used for each solution: two series for cell

viability 3-(4,5-Dimethylthiazol-2-yl) -2,5-

Diphenyltetrazolium Bromide (MTT) testing, two series for

histomorphologic analyses after hematoxylin and eosin

staining and immunohistological analyses on cryosections,

and two series for immunofluorescent labeling on the most

superficial layers of 3D-HCE by en-face confocal

microscopic analyses.

Modified MTT test: The modified MTT test was used to assess

cellular viability as described previously [27-29].

Experiments were conducted in duplicate. The 3D-HCEs were

transferred in 24-well plates containing 300 μl of the MTT

solution diluted at 0.5 mg/ml in culture medium and 300 µl of

the same MTT solution were applied on the apical surface of

the 3D-HCEs. Reconstituted tissues were incubated for 3 h.

Then, the 3D-HCEs

were transferred into 24-well plates

containing 750 µl isopropanol, and 750 µl isopropanol were

added to the apical surface of the 3D-HCEs. After a 2-h

agitation, solutions were vigorously homogenized before

reading the absorbance at 570 nm versus 690 nm. Results were

expressed as a percentage of cell viability compared to the

negative control, PBS. Analyses were performed using Safire

technology (Tecan, Lyon, France).

Confocal immunofluorescence analyses on cryosections and

entire epithelia: After incubation with the 9 different

solutions, the 3D-HCE samples were transferred into Petri

dishes containing PBS to be separated into two pieces using

a surgical scalpel. Each eceip fo tissue was e mbedded in

OCT

medium (Tissue-Tek, Miles Inc., Elkhart, IN), and

frozen at –80 °C. Vertical cryosections (10 μm thick) were

then cut using a cryotome (Leica CM 3050s, Leica

Microsystems AG, Wetzlar, Germany). The cryosections

were fixed in 4% paraformaldehyde (PFA) for 20 min

before immunofluorescent labeling of the tight junction

protein occludin.

Detection of apoptosis (TUNEL assay), inflammation (CD54)

and proliferation (Ki67) on 3D-HCE cryosections:

Apoptosis, TUNEL assay—Apoptosis in the tissue

layers was detected using a terminal deoxynucleotidyl

TABLE 1. BENZALKONIUM CHLORIDE (BAC) AND ACTIVE COMPOUND CONTENT OF THE ANTIALLERGIC EYE DROPS TESTED.

Eye drops

Active compound content BAC content

Ketotifen fumarate (Zaditen®; Novartis Pharma SAS,Rueil-Malmaison, France) 0.025% 0.01%

Olopatadine chlorhydrate (Opatanol®; Patanol®; Alcon, Ft. Worth, TX) 0.1% 0.01%

Epinastine chlorhydrate (Purivist®; Allergan, Irvin, CA) 0.05% 0.01%

Levocabastine chlorhydrate (Levophta®; Chauvin Bausch & Lomb, Montpellier, France) 0.05% 0.015%

Preservative-free ketotifen fumarate (Zalerg®; Thea, Clermont-Ferrand, France) 0.025% -

Preservative-free NAAGA (NAABAK®; Thea, Clermont-Ferrand, France) 4.9% -

Preservative-free sodium cromoglycate (Cromabak®; Thea, Clermont-Ferrand, France) 2% -

746

transferase-mediated dUTP-nick end labeling (TUNEL) kit

containing TUNEL enzyme and TUNEL label (Roche

Diagnostics, Meylan, France). Nuclei were stained with 4',6-

diamidino-2-phenylindole (DAPI) and the cryosections were

mounted in an anti-fade medium (Vectashield; Vector

Laboratories, Burlingame, CA).

CD54 (ICAM-1) and Ki67 immunostaining—First,

samples were fixed with 4% PFA for 10 min. Then, samples

were permeabilized with 0.01%-diluted Triton X100

(Sigma

Chemical Company, St. Louis, MO) for 5 min. Cells were

incubated in presence of the mouse anti-human cluster of

differentiation 54 (CD54) (IgG1; 1:100 final dilution; BD

Biosciences, PharMingen, San Diego, CA), the mouse anti-

human Ki67 (1:25 final dilution; Immunotech, Marseilles,

France) or with the isotypic control mouse IgG1 (BD

Biosciences) primary antibodies. Alexa 488 conjugated-goat

anti-mouse IgG (Invitrogen-Molecular Probes, Eugene, OR)

was used as second antibody at a 1:500 dilution. Nuclei were

labeled with propidium iodide (PI) and cryosections were

mounted in Vectashield. Samples were analyzed under a laser

confocal microscope equipped with a digital camera (E800;

PCM 2000; Nikon, Champigny-sur-Marne, France).

Immunopositive cells were then counted under the 20×

objective of the microscope in three different areas. Results

were calculated as the average of counts, and finally expressed

as cell numbers per mm of epithelial length (mm.E.L.) after

each treatment.

Confocal immunofluorescence on entire epithelia for

tight junction staining—The rabbit anti-human occludin

(IgG1; 1:100 dilution; Dako, Glostrup, Denmark) was used

for tight junction staining. Alexa 488-conjugated goat anti-

rabbit was used as second antibody. Samples were then

analyzed under a laser confocal microscope (E800; PCM

2000; Nikon) for detecting occludin expression.

Quantification and statistical analysis: Quantification of

TUNEL-, ICAM-1-, and Ki67-positive cells was performed

manually, using a microscopic grid on images under 400×

magnification. Results were expressed as mean cell numbers

per millimeter of epithelial length (mm.E.L). Standard

deviations were indicated. Statistical comparisons were

performed using two-way analysis of variance (ANOVA),

followed by multiple pairwise comparisons using the Fisher’s

adjustment (Statview V for Windows; SAS Institute, Cary,

NC).

RESULTS

Cell viability: MTT test: The PBS negative control did not

affect the cell viability neither at 24 h nor after the 24 h-

recovery period (24 h+24 h; Figure 1). The unpreserved

formulation of ketotifen fumarate KETO-BAC(-) showed the

same level of cell viability as PBS at 24 h (99,4%) and a slight

decrease of viability after 24 h+24 h (86.9% of the control).

The preservative-free formulations of NAAGA and

cromoglycate, NAA-BAC(-) and CRO-BAC(-), also showed

a weak decrease of cellular viability at 24 h (93.2% and 95.1%,

respectively) and after 24 h+24 h (87.7% for both; Figure 1).

Conversely, as expected according to previous studies

[28], 0.01% BAC showed a significant decrease of cell

viability at 24 h and after 24 h+24 h (59.6% and 55% viability,

respectively). Cell viability decreased in a BAC-

concentration dependant manner for the BAC-containing

antiallergic formulations, with a highest toxicity observed

with the 0.015% BAC-containing levocabastine chlorhydrate

0.05% [LEVO-BAC(+)]. Cell viability levels were 66.8% at

24 h and 55.3% at 24 h+24 h for 0.01% BAC-containing

ketotifen fumarate 0.025% [KETO-BAC(+)], 63.7% at 24 h

and 60.5% at 24 h+24 h for 0.01% BAC-containing

olopatadine chlorhydrate 0.1% [OPA-BAC(+)], 52.1% at 24

h and 56.6% at 24 h+24 h for 0.01% BAC-containing

epinastine chlorhydrate 0.05% [EPI-BAC(+)], and 46.3% at

24 h and 44.3% at 24 h+24 h for LEVO-BAC(+).

Immunofluorescence analyses and quantification of apoptosis

(TUNEL): Few apoptotic cells were observed after PBS

incubation (6.4 cells/mm.E.L.). Similar levels of apoptosis

were observed with the 3 unpreserved antiallergic

formulations (Figure 2): 9.2 cells/mm.E.L. for NAA-BAC(-),

12.2 cells/mm.E.L. for CRO-BAC(-), and 8.6 cells/mm.E.L.

for KETO-BAC(-), without any statistically significant

difference compared to PBS.

Consistent with previously published reports with the

same technique [28,29], BAC at 0.01% significantly increased

the number of TUNEL-positive cells compared to PBS

(p<0.0014). Apoptosis also increased on cells treated with all

BAC-containing antiallergic formulations, with a statistically

significant difference compared to PBS (p<0.0014): 28, 29.3,

46.6, and 75.5 cells/mm.E.L. for KETO-BAC(+), OPA-

BAC(+), EPI-BAC(+), and LEVO-BAC(+), respectively.

Immunofluorescence analyses and quantification of the

inflammation marker ICAM-1 (CD54): CD54 expression was

measured at 70 cells/mm.E.L. on PBS-treated 3D-HCE

cultures (Figure 3). The three unpreserved antiallergic

formulations NAA-BAC(-), CRO-BAC(-), and KETO-

BAC(-) expressed CD54 at low levels too, respectively, 60.3,

59.3, and 64 cells/mm.E.L., with no statistically significant

differences compared to PBS. Conversely, 0.01% BAC and

LEVO-BAC(+) showed increased levels of CD54 expression

with a statistically significant difference compared to PBS

(p<0.001): 150.4 cells/mm.E.L. for 0.01% BAC and 142 cells/

mm.E.L. for LEVO-BAC(+). KETO-BAC(+), OPA-BAC(+),

EPI-BAC(+) showed increased levels of CD54 expression

too, but no statistically significant difference was found

neither with PBS nor with 0.01% BAC i.e., 98.5, 114 and 92

cells/mm.E.L. for KETO-BAC(+), OPA-BAC(+) and EPI-

BAC(+), respectively.

Immunofluorescence analyses of cell proliferation marker

Ki67: After PBS treatment (Figure 4), few proliferating cells

were observed (29.2 cells/mm.E.L), scattered throughout the

747

entire epithelium. Similar findings were observed with the

unpreserved antiallergic treatments, with

no statistically

significant differences compared to PBS: 28.3, 30.0, and 27.3

cells/mm.E.L. for NAA-BAC(-), CRO-BAC(-), and KETO-

BAC(-), respectively (Figure 4). Conversely, numerous

proliferating cells, with a greater number located in the basal

layer, were found after 0.01% BAC, KETO-BAC(+), OPA-

BAC(+), and EPI-BAC(+) with a statistically significant

difference (p<0.04) compared to PBS: 55.3, 45, 45, and 55

cells/mm.E.L., respectively. With LEVO-BAC(+), no

proliferative cells were observed, most likely due to the deep

impairment of corneal cells as this group showed the most

important number of apoptotic cells.

En-face confocal microscopic analysis of the tight junction-

associated protein occludin: En-face confocal microscopic

analysis of 3D-HCE cultures treated with PBS, NAA-BAC(-),

CRO-BAC(-), and KETO-BAC(-) revealed a fine membrane

immunostaining of occludin in large superficial cells, forming

a ring around the cells (Figure 5). This kind of occludin

expression clearly disappeared after treatment with either

0.01% BAC, KETO-BAC(+), OPA-BAC(+), EPI-BAC(+), or

LEVO-BAC(+), all showing damaged cells with non-specific

staining.

DISCUSSION

In this study, the toxicological model of 3D-reconstructed

cornea was very helpful to demonstrate the effects of BAC-

preserved solutions on corneal cells in vitro, showing

increased

apoptosis, CD54 expression, proliferation in the

basal layers and changes in the distribution of occludin

induced with BAC-containing antiallergic treatments. On the

contrary, the unpreserved ketotifen, NAAGA and

cromoglycate solutions did not impair cell structures and

viability, suggesting a better tolerance for the ocular surface.

The highly differentiated, three-dimensional epithelial

system of human ocular origin is a desirable model for pre-

screening or investigating the effects of ophthalmic drugs. It

frees the experimenter from interspecies differences and

allows a better approach to the ocular epithelial physiology

than monolayer models and cells originating from other

organs. It also constitutes an interesting alternative to animal

testing, respecting the ethical guidelines of animal

experimentation, especially the 3R rule (refining, reducing

and replacing the use of animals) [30,31]. The reconstructed

three dimensional (3D) model of human corneal cells (3D-

HCE), supplied by SkinEthic

Laboratories, was found to

resemble the corneal epithelium of the human eye in

morphology and thickness [32]. Such a 3D-system models is

not only useful to demonstrate the different effects of toxic

substances on specific cell types, but also shows the

interactions between the cells and the spatial effects induced

by the toxic. Moreover, epithelium cultures at the air-liquid

interface are easy-to-handle and facilitate in vivo-like product

Figure 1. Cell viability MTT test.

Cellular viability of 3D-HCEs

treated

with PBS, 0.01% BAC, NAA-BAC(-),

CRO-BAC(-), KETO-BAC(-), KETO-

BAC(+), OPA-BAC(+), EPI-BAC(+),

or LEVO-BAC(+) for 24 h followed or

not by a 24 h post-incubation period (24

h+24 h-recovery). BAC induced a

concentration-dependent decrease of

cellular viability. At 24 h, the

unpreserved NAA-BAC(-), CRO-

BAC(-), and KETO-BAC(-)

formulations induced a slight or

insignificant decrease of cellular

viability, while the KETO-BAC

(+), OPA-BAC(+), EPI-BAC(+), and

LEVO-BAC(+) BAC-containing

formulations induced a marked decrease

of cellular viability compared to control.

After the 24-h recovery period, the

unpreserved formulations showed a

weak additional decrease of cellular

viability, while the BAC-containing

formulations still induced a strong

decrease of cellular viability compared

to control, showing irreversible damage

to 3D-HCE. Results are expressed as

percentage of cell viability compared to

the PBS control.

748

exposures. The 3D-HCEs were found to express cytokeratin-3

and include hemidesmosomes

within the basal layers.

Furthermore, they can inhibit the flow of ionic material such

as Na-fluorescein across their surface [32,33], suggesting the

presence of a functional epithelial barrier. Different types of

intercellular junctions have been identified in the corneal

epithelium ex vivo. Among them, adherens junctions,

comprising the E-cadherin protein, serve to anchor cells

together [34]. Also, the tight-junctions, originally defined as

zonula occludentes (ZO) and comprising occludin, ZO-1 and

other proteins, are thought to provide the hydrophobic barrier

preventing the free passage of molecules between adjacent

Figure 2. Apoptosis analysis (TUNEL).

Localization of TUNEL

positive cells

(green) on 3D-HCE samples after 24 h

of incubation with PBS (A), 0.01% BAC

(B), KETO-BAC(-) (C), KETO-

BAC(+) (D), NAA-BAC(-) (E), OPA-

BAC(+) (F), CRO-BAC(-) (G), EPI-

BAC(+) (H), LEVO-BAC(+) (I).

Nuclei were stained with DAPI (blue).

No or very rare apoptotic cells were

observed after PBS (A), KETO-BAC(-)

(C), KETO-BAC(+) (D) and NAA-

BAC(-) (E) treatments. KETO-BAC(+)

(D) and OPA-BAC(+) (F) induced

moderate expression of apoptosis, and

0.01% BAC (B), EPI-BAC(+) (H) and

LEVO-BAC(+) (I) induced a greater

number of TUNEL-positive cells

principally in the apical cell layers, and

also in the middle epithelial layers with

EPI-BAC(+) (H) and LEVO-BAC(+)

(I). Deeper modifications were

observed with 0.015% BAC-containing

LEVO-BAC(+) (I) compared to 0.01%

BAC (B), with a greater number of

TUNEL-positive cells in the middle

epithelial layers and a higher level of

vacuolization in the basal epithelial

layers observed with LEVO-BAC(+)

(I). The quantification of apoptotic cells

with the TUNEL assay (J) showed that

apoptotic cell number increased in a

BAC concentration-dependent manner.

BAC at 0.01% and the four BAC-

containing formulations KETO-

BAC(+), OPA-BAC(+), EPI-BAC(+)

and LEVO-BAC(+) showed much

higher expression of apoptotic TUNEL-

positive cells than did the unpreserved

formulations NAA-BAC(-), CRO-

BAC(-), KETO-BAC(-) at 24 h. Results

are expressed as cell number per mm of

epithelial length (mm.E.L.): Mean±SD

*Statistically significant compared to

PBS with p<0.0014. **Statistically

significant compared to 0.01% BAC

with p<0.0014. †Statistically significant

compared to EPI-BAC(+) with

p<0.0014. $Statistically significant

compared to LEVO-BAC(+) with

p<0.0014.

749

epithelial cells [35-37]. In a

previous study [28], we developed

a new procedure of the classical MTT test used on 3D-

reconstituted epidermal and corneal models to evaluate the

viability. This procedure showed increased sensitivity levels

and allowed detecting slight damage even in the most

superficial layers. Therefore, it is well suited to the prediction

of low to very low irritant potential, especially when products

are used repeatedly during long-term periods of time, like in

allergic conjunctivitis, when repeated allergenic challenge or

ocular surface impairment occur and require sustained

therapy.

Although the morphological relevance and sensitivity of

the 3D-HCE model allowed the modeling of cumulative

effects that may approach conditions obtained after long-term

application of eye-drops [27], our in vitro findings cannot

fully be extrapolated to in vivo conditions. Indeed, preserved

eye drops may be less toxic in vivo, according to the

continuous action of the eyelids, the permanent renewal of

ocular surface epithelia, and the presence of the preocular

mucin layer and glycocalyx. Conversely, the accumulation of

BAC-containing eye drops in the eye and the long-term use

of eye drops in allergic patients with ocular surface disorders

will emphasize the risk of toxic reactions and further

contribute to inflammatory stimulation throughout the ocular

surface, at least at a subclinical level [38].

In the present study, using our modified MTT procedure,

we evaluated the effects of either preserved or unpreserved

antiallergic formulations on cellular viability and correlated

Figure 3. Inflammation analysis.

Immunolocalization of CD54

(ICAM-1) positive cells (green) on 3D-

HCE samples after 24 h of incubation

with PBS (A), 0.01% BAC (B), KETO-

BAC(-) (C), KETO-BAC(+) (D), NAA-

BAC(-) (E), OPA-BAC(+) (F), CRO-

BAC(-) (G), EPI-BAC(+) (H), LEVO-

BAC(+) (I). Nuclei were stained with

propidium iodide (PI, red). PBS (A),

KETO-BAC(-) (C), NAA-BAC(-) (E)

and CRO-BAC(-) (G) showed a weak

expression of CD54. A significant

increase of CD54 expression was

observed after the treatments with

0.01% BAC (B) and LEVO-BAC(+)

(I), showing a green staining in all the

epithelial layers. LEVO-BAC(+) (I)

showed deeper modifications with a

higher loss of continuity between cells

and a higher level of vacuolization

observed in the basal epithelial layers.

KETO-BAC(+) (D), OPA-BAC(+) (F)

and EPI-BAC(+) (H) showed an

intermediate CD54 expression that was

localized in all epithelial layers.

Quantification of CD54-positive cells

(J) showed a higher CD54 expression

with BAC at 0.01% or the four BAC-

containing formulations KETO-

BAC(+), OPA-BAC(+), EPI-BAC(+)

and LEVO-BAC(+) than with the

unpreserved formulations NAA-

BAC(-), CRO-BAC(-), KETO-BAC(-)

at 24 h. *Statistically significant

compared to PBS with p<0.001.

**Statistically significant compared to

0.01% BAC with p<0.001. Results are

expressed as cell number per mm of

epithelial length (mm.E.L.): Mean±SD.

750

these results with those of a TUNEL assay performed on 3D-

HCE frozen sections.

Then, we investigated on entire 3D-

HCE and using en-face confocal microscopy the changes of

expression and spatial distribution of cellular markers

involved in intercellular junctions such as occludin after the

different antiallergic treatments.

Thus, with this procedure, we were able to demonstrate

concentration-dependent cytotoxic effects of BAC at 24 h, the

absence of significant cellular viability decrease following

treatment with the ketotifen, NAAGA and cromoglycate

BAC-free formulations, and a cell viability decrease similar

to that disclosed by the 0.01% BAC treatment with the BAC-

containing antiallergic formulations of ketotifen, olopatadine,

epinastine and levocabastatine. We confirmed the toxic and

proinflammatory effects of the BAC-containing solutions

using a TUNEL assay and CD54 immunostaining performed

on 3D-HCE frozen sections and found a significantly

increased number of apoptotic cells and an increased CD54

Figure 4. Proliferation analysis.

Immunolocalization of Ki67 positive

cells (green) on 3D-HCE samples after

24h of incubation with PBS (A), 0.01%

BAC (B), KETO-BAC(-) (C), KETO-

BAC(+) (D), NAA-BAC(-) (E), OPA-

BAC(+) (F), CRO-BAC(-) (G), EPI-

BAC(+) (H), LEVO-BAC(+) (I).

Nuclei were stained with propidium

iodide (PI, red). PBS (A), KETO-

BAC(-) (C), NAA-BAC(-) (E) and

CRO-BAC(-) (G) showed a weak

expression of Ki67 in all epithelial

layers. BAC at 0.01% (B), KETO-

BAC(+) (D), OPA-BAC(+) (F), and

EPI-BAC(+) (H) showed a higher Ki67

expression in all epithelial layers too.

With LEVO-BAC(+), no Ki67 positive

cells were observed, most likely due to

the deep impairment of corneal cells

with a most likely inhibition of

proliferative capabilities of 3D-HCE

submitted at this higher concentration in

BAC. Quantification of Ki67-positive

cells was concordant with these

observations (J). *Statistically

significant compared to PBS with

p<0.04. **Statistically significant

compared to 0.01% BAC with p<0.001.

$Statistically significant compared to

the other solutions tested. Results are

expressed as cell number per mm of

epithelial length (mm.E.L.): Mean±SD.

751

expression following exposure to 0.01% BAC and BAC-

containing solutions compared

to the control. Finally, we

examined the integrity of the structural and functional barrier

conferred by the tight-junctions by assessing the distribution

pattern of the occludin protein. The tight-junctions regulate

the passive movement of fluids, electrolytes, macromolecules

and cells through the paracellular pathway, thereby

contributing to the corneal defense system and to the

maintenance of the corneal homeostasis. In the mouse cornea,

the occludin distribution pattern was already described as

altered by a detergent treatment (Triton X100) using

immunohistochemistry [39]. In a previous study, Chuan et al.

[40] showed the effects of contact lens multipurpose solutions

on the corneal cells’ barrier function using fluorescein

permeability assay and immunofluorescent staining for tight

junctions proteins (ZO-1 and occludin). Recently, we also

demonstrated that occludin mRNA expression was correlated

to BAC early toxic effects [28]. Our results were consistent

with those studies, showing the disturbance of occludin tight-

junction protein distribution after BAC-containing

antiallergic treatments.

Currently, allergic conjunctivitis incidence is increasing

in developed countries. According to Manners T et al. [41],

15% of eye related consultations in general practice are due

to allergic conjunctivitis. Recommended topical treatments

for symptoms of allergic conjunctivitis include topical mast

cell stabilizers and/or topical antihistamines (H1-receptor

antagonists). Some of the new antiallergic drugs now

available may have both effects and sometimes additional

properties, such as the ability to inhibit the expression of cell

adhesion molecules (CAMs) on the cell surface or to attenuate

inflammatory mediator release [1,3-6,42-45].

Figure 5. Tight junction-associated protein occludin. Immunofluorescence analysis of occludin expressions using en-face confocal microscopy

after treatment with PBS (A), 0.01% BAC (B), NAA-BAC(-) (C) CRO-BAC(-) (D), KETO-BAC(-) (E), KETO-BAC(+) (F), OPA-BAC(+)

(G), EPI-BAC(+) (H), LEVO-BAC(+) (I). Bar=100 µm.

752

The panel of eye drops tested in the present study was

deemed to be fairly representative of the predominantly

prescribed therapeutic antiallergic molecules at the time of

these experiments in France. Indeed, among the commercial

antiallergic eye drops, one can distinguish between two types

of formulations, according to the presence of BAC as

preservative. Currently, the antihistamines olopatadine,

epinastine and levocabastine are available only as preserved

solutions whereas ketotifen is recently accessible in both

formulations. These four antihistamines constitute a group of

comparable products from the therapeutic use viewpoint and

all of them are available as preserved solutions. We deemed

it interesting to add the preservative-free ketotifen solution in

the comparison. Naaga and cromoglycate eye drops belong to

a different class of antiallergic agents, i.e., the mast cells

degranulation inhibitors. Although they were both available

as preserved and unpreserved formulations, we chose to only

test their unpreserved formulations in order not to weigh the

experiment down all the more since the preserved eye drop

forms of these two molecules are now much less used in

therapeutics than their unpreserved counterparts. Overall, our

panel choice was conducted by the actuality of the antiallergic

armamentarium that is available to the patients.

There is currently enough evidence from clinical, in vivo,

and in vitro studies that long-term use of preserved topical

drugs may induce several deleterious effects on ocular

surface, being responsible for ocular discomfort, tear film

instability, conjunctival inflammation, subconjunctival

fibrosis and epithelial apoptosis [9]. Several studies have

confirmed the participation of high concentrations of BAC-

preserved eye drops in induction of ocular surface

inflammation, allergy, fibrosis, punctate corneal staining, and

dry eye syndrome [9,38,46,47]. Three mechanisms have been

described: detergent effects inducing loss of tear film stability;

immunoallergic reactions; and direct toxic effects to epithelial

cells [48,49]. Other experimental and clinical studies have

shown that the long-term use of BAC-containing ophthalmic

solutions can induce conjunctival stroma infiltrates and

overexpression of inflammation- or apoptosis-related

molecules, such as class II antigen HLADR, ICAM-1, Fas

antigen, or the apoptotic marker Apo 2.7 [50-52].

In the present study, we showed that BAC-containing eye

antiallergic solutions may decrease cell viability, induce

apoptosis, ICAM-1 expression and proliferation in the basal

layers, and changes in the distribution of occludin.

Conversely, the unpreserved ketotifen, NAAGA and

cromoglycate formulations did not impair cell structures and

viability, suggesting a better tolerance for the ocular surface.

These findings were consistent with several previous in vitro

or ex vivo studies that demonstrated BAC toxicity and

potential advantages of BAC-free formulations [9].

Moreover, the present results support our earlier findings on

antiallergic preserved and unpreserved eye-drops. Indeed, in

a previous study, we showed that antiallergic eye drops

preserved with BAC induced high ICAM-1 expression levels,

apoptosis and oxidative stress and reduced cellular viability

in opposition to the unpreserved formulations of NAAGA and

cromoglycate [53].

In addition, our results were consistent with a recent study

by Ayaki et al. [54] on corneal and conjunctival cell lines that

showed that cell toxicity was mostly affected by the

concentration of BAC rather than the active component of

antiallergic ophthalmic solutions. The use of a large range of

commonly antiallergic eye drops in the present study was

useful to support this point, showing BAC-concentration

dependent toxic effects in all experiments. This may

emphasize the fact that epithelial toxicity was most likely

induced by the preservative (BAC) than by the active

antiallergic compound in the present study too.

These findings strongly support the use of preservative-

free solutions in patients with chronic eye diseases and

treatments over the long-term, especially in allergic

conjunctivitis or dry eye conditions. Definitely, preservative-

free antiallergic medications may decrease the adverse effects

of chronic topical medications, which could lead to better

tolerability, lower treatment discontinuations and improved

quality of life of patients with ocular allergic diseases.

ACKNOWLEDGMENTS

The study was supported by an unrestricted grant from Théa

(Clermont-Ferrand, France).

REFERENCES

1. Martín AP, Urrets-Zavalia J, Berra A, Mariani AL, Gallino N,

Gomez Demel E, Gagliardi J, Baena-Cagnani CE, Urrets-

Zavalia E, Serra HM. The effect of ketotifen on inflammatory

markers in allergic conjunctivitis: an open, uncontrolled

study. BMC Ophthalmol 2003; 3:2. [PMID: 12515585]

2. Schoch C. Effects of ketotifen 0.025% and lodoxamide 0.1%

on eosinophil infiltration into the guinea pig conjunctiva in a

model of allergic conjunctivitis. J Ocul Pharmacol Ther 2003;

19:153-9. [PMID: 12804060]

3. Borazan M, Karalezli A, Akova YA, Akman A, Kiyici H, Erbek

SS. Efficacy of olopatadine HCI 0.1%, ketotifen fumarate

0.025%, epinastine HCI 0.05%, emedastine 0.05% and

fluorometholone acetate 0.1% ophthalmic solutions for

seasonal allergic conjunctivitis: a placebo-controlled

environmental trial. Acta Ophthalmol 2009; 87:549-54.

[PMID: 18631332]

4. Avunduk AM, Tekelioglu Y, Turk A, Akyol N. Comparison of

the effects of ketotifen fumarate 0.025% and olopatadine HCl

0.1% ophthalmic solutions in seasonal allergic

conjunctivities: a 30-day, randomized, double-masked,

artificial tear substitute-controlled trial. Clin Ther 2005;

27:1392-402. [PMID: 16291412]

5. Kidd M, McKenzie SH, Steven I, Cooper C, Lanz R. Efficacy

and safety of ketotifen eye drops in the treatment of seasonal

allergic conjunctivitis. Br J Ophthalmol 2003; 87:1206-11.

[PMID: 14507747]

6. Ganz M, Koll E, Gausche J, Detjen P, Orfan N. Ketotifen

fumarate and olopatadine hydrochloride in the treatment of

753

allergic conjunctivitis: a real-world comparison of efficacy

and ocular

comfort. Adv Ther 2003; 20:79-91. [PMID:

12836808]

7. Bielory L, Friedlaender MH. Allergic conjunctivitis. Immunol

Allergy Clin North Am 2008; 28:43-58. [PMID: 18282545]

8. Wong AH, Barg SS, Leung AK. Seasonal and perennial allergic

conjunctivitis. Recent Pat Inflamm Allergy Drug Discov

2009; 3:118-27. [PMID: 19519588]

9. 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; 29:312-34. [PMID: 20302969]

10. Baudouin C, Liang H, Hamard P, Riancho L, Creuzot-Garcher

C, Warnet JM, Brignole-Baudouin F. The ocular surface of

glaucoma patients treated over the long term expresses

inflammatory markers related to both T-helper 1 and T-helper

2 pathways. Ophthalmology 2008; 115:109-15. [PMID:

17532048]

11. 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; 33:303-12. [PMID: 18398704]

12. 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; 48:4123-8.

[PMID: 17724196]

13. 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; 92:1275-82. [PMID:

18723745]

14. 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; 46:2444-50. [PMID: 15980234]

15. Ichijima H, Petroll WM, Jester JV, Cavanagh HD. Confocal

microscopic studies of living rabbit cornea treated with

benzalkonium chloride. Cornea 1992; 11:221-5. [PMID:

1587129]

16. Ishibashi T, Yokoi N, Kinoshita S. Comparison of the short-

term effects on the human corneal surface of topical timolol

maleate with and without benzalkonium chloride. J Glaucoma

2003; 12:486-90. [PMID: 14646684]

17. Buron N, Micheau O, Cathelin S, Lafontaine PO, Creuzot-

Garcher C, Solary E. Differential mechanisms of conjunctival

cell death induction by ultraviolet irradiation and

benzalkonium chloride. Invest Ophthalmol Vis Sci 2006;

47:4221-30. [PMID: 17003409]

18. Epstein SP, Ahdoot M, Marcus E, Asbell PA. Comparative

toxicity of preservatives on immortalized corneal and

conjunctival epithelial cells. J Ocul Pharmacol Ther 2009;

25:113-9. [PMID: 19284328]

19. Martone G, Frezzotti P, Tosi GM, Traversi C, Mittica V,

Malandrini A, Pichierri P, Balestrazzi A, Motolese PA,

Motolese I, Motolese E. An in vivo confocal microscopy

analysis of effects of topical antiglaucoma therapy with

preservative on corneal innervation and morphology. Am J

Ophthalmol 2009; 147:725-35. [PMID: 19181302]

20. Rolando M, Brezzo V, Giordano G, Campagna P, Burlando S,

Calabria G. The effect of different benzalkonium chloride

concentrations on human normal ocular surface. In: Van

Bijsterveld O, Lemp M, Spinelli D, editors. The Lacrimal

System. Amsterdam, Berkely, Milano: Kugler and Ghedini

Publications; 1991.

21. Baudouin C, Hamard P, Liang H, Creuzot-Garcher C,

Bensoussan L, Brignole F. Conjunctival epithelial cell

expression of interleukins and inflammatory markers in

glaucoma patients treated over the long term. Ophthalmology

2004; 111:2186-92. [PMID: 15582072]

22. Calonge M, Diebold Y, Saez V, Enriquez de Salamanca A,

Garcia-Vazquez C, Corrales RM, Herreras JM. Impression

cytology of the ocular surface: a review. Exp Eye Res 2004;

78:457-72. [PMID: 15106925]

23. Schwab IR, Linberg JV, Gioia VM, Benson WH, Chao GM.

Foreshortening of the inferior conjunctival fornix associated

with chronic glaucoma medications. Ophthalmology 1992;

99:197-202. [PMID: 1348114]

24. Broadway D, Grierson I, Hitchings R. Adverse effects of topical

antiglaucomatous medications on the conjunctiva. Br J

Ophthalmol 1993; 77:590-6. [PMID: 8218059]

25. Broadway DC, Grierson I, O'Brien C, Hitchings RA. Adverse

effects of topical antiglaucoma medication. II. The outcome

of filtration surgery. Arch Ophthalmol 1994; 112:1446-54.

[PMID: 7980134]

26. Doucet O, Lanvin M, Thillou C, Linossier C, Pupat C, Merlin

B, Zastrow L. 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. [PMID: 16243479]

27. 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; 24:276-85.

[PMID: 19729060]

28. 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; 50:1644-52. [PMID:

19168896]

29. Liang H, Pauly A, Riancho L, Baudouin C, Brignole-Baudouin

F. Toxicological evaluation of preservative-containing and

preservative-free topical prostaglandin analogs on a 3D-

reconstituted corneal epithelium system. Br J Ophthalmol.

2010In press

30. Manciocco A, Chiarotti F, Vitale A, Calamandrei G, Laviola G,

Alleva E. The application of Russell and Burch 3R principle

in rodent models of neurodegenerative disease: the case of

Parkinson's disease. Neurosci Biobehav Rev 2009;

33:18-32. [PMID: 18771685]

31. ECVAM Technical Report on the Status of Alternative Methods

for Cosmetics Testing (2008–2009). A report prepared in the

framework of Directive 2003/15/EC (7th Amendment to the

Cosmetics Directive): Publications Office of the European

Union; 2010.

754

32. Nguyen DH,

Beuerman RW, De Wever B, Rosdy M.

Alternative Toxicological Methods. Salem H, Katz S, Editors:

CRC Press; 2003. p. 147–59.

33. Kahn CR, Young E, Lee IH, Rhim JS. Human corneal epithelial

primary cultures and cell lines with extended life span: in vitro

model for ocular studies. Invest Ophthalmol Vis Sci 1993;

34:3429-41. [PMID: 7693609]

34. Gumbiner BM. Cell adhesion: the molecular basis of tissue

architecture and morphogenesis. Cell 1996; 84:345-57.

[PMID: 8608588]

35. Wang Y, Chen M, Wolosin JM. ZO-1 in corneal epithelium;

stratal distribution and synthesis induction by outer cell

removal. Exp Eye Res 1993; 57:283-92. [PMID: 8224016]

36. Langbein L, Grund C, Kuhn C, Praetzel S, Kartenbeck J,

Brandner JM, Moll I, Franke WW. Tight junctions and

compositionally related junctional structures in mammalian

stratified epithelia and cell cultures derived therefrom. Eur J

Cell Biol 2002; 81:419-35. [PMID: 12234014]

37. Ban Y, Dota A, Cooper LJ, Fullwood NJ, Nakamura T, Tsuzuki

M, Mochida C, Kinoshita S. Tight junction-related protein

expression and distribution in human corneal epithelium. Exp

Eye Res 2003; 76:663-9. [PMID: 12742348]

38. Baudouin C. Allergic reaction to topical eyedrops. Curr Opin

Allergy Clin Immunol 2005; 5:459-63. [PMID: 16131924]

39. Sosnová-Netuková M, Kuchynka P, Forrester JV. The

suprabasal layer of corneal epithelial cells represents the

major barrier site to the passive movement of small molecules

and trafficking leukocytes. Br J Ophthalmol 2007;

91:372-8. [PMID: 17020902]

40. Chuang EY, Li DQ, Bian F, Zheng X, Pflugfelder SC. Effects

of contact lens multipurpose solutions on human corneal

epithelial survival and barrier function. Eye Contact Lens

2008; 34:281-6. [PMID: 18779668]

41. Manners T. Managing eye conditions in general practice. BMJ

1997; 315:816-7. [PMID: 9345194]

42. Owen CG, Shah A, Henshaw K, Smeeth L, Sheikh A. Topical

treatments for seasonal allergic conjunctivitis: systematic

review and meta-analysis of efficacy and effectiveness. Br J

Gen Pract 2004; 54:451-6. [PMID: 15186569]

43. Abelson MB. A review of olopatadine for the treatment of

ocular allergy. Expert Opin Pharmacother 2004; 5:1979-94.

[PMID: 15330735]

44. Ciprandi G, Cosentino C, Milanese M, Tosca MA. Rapid anti-

inflammatory action of azelastine eyedrops for ongoing

allergic reactions. Ann Allergy Asthma Immunol 2003;

90:434-8. [PMID: 12722967]

45. Yanni JM, Weimer LK, Sharif NA, Xu SX, Gamache DA,

Spellman JM. Inhibition of histamine-induced human

conjunctival epithelial cell responses by ocular allergy drugs.

Arch Ophthalmol 1999; 117:643-7. [PMID: 10326962]

46. Fisher AA. Allergic contact dermatitis and conjunctivitis from

benzalkonium chloride. Cutis 1987; 39:381-3. [PMID:

3581909]

47. Baudouin C, Garcher C, Haouat N, Bron A, Gastaud P.

Expression of inflammatory membrane markers by

conjunctival cells in chronically treated patients with

glaucoma. Ophthalmology 1994; 101:454-60. [PMID:

7907416]

48. De Saint Jean M, Baudouin C, Di Nolfo M, Roman S, Lozato

P, Warnet JM, Brignole F. Comparison of morphological and

functional characteristics of primary-cultured human

conjunctival epithelium and of Wong-Kilbourne derivative of

Chang conjunctival cell line. Exp Eye Res 2004; 78:257-74.

[PMID: 14729358]

49. Debbasch C, Brignole F, Pisella PJ, Warnet JM, Rat P,

Baudouin C. Quaternary ammoniums and other preservatives'

contribution in oxidative stress and apoptosis on Chang

conjunctival cells. Invest Ophthalmol Vis Sci 2001;

42:642-52. [PMID: 11222522]

50. Becquet F, Goldschild M, Moldovan MS, Ettaiche M, Gastaud

P, Baudouin C. Histopathological effects of topical

ophthalmic preservatives on rat corneoconjunctival surface.

Curr Eye Res 1998; 17:419-25. [PMID: 9561834]

51. Baudouin C, Pisella PJ, Fillacier K, Goldschild M, Becquet F,

De Saint Jean M, Bechetoille A. Ocular surface inflammatory

changes induced by topical antiglaucoma drugs: human and

animal studies. Ophthalmology 1999; 106:556-63. [PMID:

10080214]

52. Mietz H, Niesen U, Krieglstein GK. The effect of preservatives

and antiglaucomatous medication on the histopathology of

the conjunctiva. Graefes Arch Clin Exp Ophthalmol 1994;

232:561-5. [PMID: 7959096]

53. Pauly A, Brignole-Baudouin F, Guenoun JM, Riancho L, Rat

P, Warnet JM, Baudouin C. Comparative study of topical anti-

allergic eye drops on human conjunctiva-derived cells:

responses to histamine and IFN gamma and toxicological

profiles. Graefes Arch Clin Exp Ophthalmol 2007;

245:534-46. [PMID: 16900358]

54. Ayaki M, Iwasawa A, Yaguchi S, Koide R. Preserved and

unpreserved 12 anti-allergic ophthalmic solutions and ocular

surface toxicity: in vitro assessment in four cultured corneal

and conjunctival epithelial cell lines. Biocontrol Sci 2010;

15:143-8. [PMID: 21212507]

Articles are provided courtesy of Emory University and the Zhongshan Ophthalmic Center, Sun Yat-sen University, P.R. China.

The print

version of this article was created on 11 March 2011. This reflects all typographical corrections and errata to the article

through that date. Details of any changes may be found in the online version of the article.

755