Benzalkonium chloride-induced dry eye disease animal models: Current understanding and potential for translational research

Abstract

Dry eye disease (DED) is an emerging health issue affecting people worldwide. There have been rapid advances in the development of novel molecules and targeted therapies for the treatment of DED in the recent past. For testing and optimizing these therapies, it is necessary to have reliable experimental animal models of DED. One such approach is the use of benzalkonium chloride (BAC). Several BAC-induced DED models of rabbits and mice have been described in literature. BAC induces high levels of proinflammatory cytokines in the cornea and conjunctiva, along with epithelial cell apoptosis and reduction of mucins, which leads to tear film instability, thereby successfully simulating human DED. The stability of these models directs whether the treatment is to be applied while BAC is being instilled or after its cessation. In this review, we summarize the previously described BAC animal models of DED and present original data on rabbit DED models created using 0.1%, 0.15%, and 0.2% BAC administration twice daily for two consecutive weeks. The 0.2% BAC model sustained DED signs for 3 weeks, while 0.1% and 0.15% models sustained DED signs for 1–2 weeks after BAC discontinuation. Overall, these models look promising and continue to be used in various studies to investigate the efficacy of therapeutic drugs for DED treatment.

Dry eye disease (DED), which affects millions of people every year, is a common ocular surface disease characterized by tear film instability and hyperosmolarity, resulting in ocular discomfort and foreign body sensation in patients.[1,2] Animal models mimicking the human DED are an imperative tool for understanding the pathophysiology of DED as well as evaluating the therapeutic efficacy of new treatments for dry eye.[3] Therefore, several animal models have been fabricated over the past years based on various etiologies of dry eye, including evaporative model,[4] meibomian gland dysfunction model,[5] dacryoadenectomy,[6] induced autoimmune dacryoadenitis,[7] drainage duct injury model,[8] and topical application of benzalkonium chloride (BAC).[9] This publication focuses on BAC-induced dry eye animal model.

BAC, a quaternary ammonium compound, is a commonly used preservative in ophthalmic preparations for ocular surface diseases at a concentration ranging from 0.005% to 0.02%.[10,11] It dissolves the intercellular junctions within the corneal epithelium to boost drug delivery. BAC possesses detergent-like properties, which disrupt the lipid layer of the tear film and damage the integrity of epithelial cell membrane, resulting in corneal epithelial cell loss, conjunctival metaplasia, infiltration of inflammatory cytokines, and disruption of tear film [Fig. 1].[12,13] All these events are the hallmarks of dryness on the ocular surface. Therefore, several mice,[14] rabbit,[9] and rat[15] dry eye models have been successfully developed using topical administration of BAC. Among these animals, rabbits are considered to be more suitable for ophthalmic examinations due to their larger eye size which provides better accessibility of the ocular surface. Therefore, all the dry eye diagnostic tests such as Schirmer’s, corneal fluorescein staining, corneal smoothness, and tear breakup time (TBUT) can be easily conducted in rabbits than in mice or rats.[16] On the other hand, mice are feasible due to small size, low expense, and ease of creating various gene knockout models. However, their anatomical and physiological structures still create hindrance in developing murine dry eye model.[17]

Figure 1.

Comparison of tear film stability and corneal anatomy between normal and BAC induced DED animal models. BAC = benzalkonium chloride, DED = dry eye disease

This study intends to summarize the existing BAC-induced DED animal models and its effect on the ocular surface, along with various therapeutic studies which uses this model. In addition, it also encompasses the comparison of animal models with different BAC concentrations.

Methods

A thorough literature search of the articles published on BAC dry eye animal model was performed on PubMed. The search used the terms “animal models for dry eye,” “benzalkonium chloride and dry eye animal model,” “benzalkonium chloride and rabbit dry eye animal model,” “benzalkonium chloride and mouse dry eye animal model,” and “benzalkonium chloride and rat dry eye animal model.” We shortlisted the articles in English language that were relevant to BAC DED animal model.

BAC-induced DED animal models

Effect of BAC concentration and time period on the ocular surface of various animals

Various dry eye markers associated with inflammation and tear quality can be assessed with the help of BAC-induced animal models. Fig. 2 shows various pathological changes in the cornea and conjunctiva of BAC-induced dry eye animal models. Table 1 summarizes the conditions used in various studies that use BAC to establish a dry eye model. In 2008, Xiong et al.[9] first proposed the use of BAC to induce dry eye in rabbits. The effects of different concentrations of BAC (0.01%, 0.05%, 0.1%, 0.2%, 0.3%, 0.5%, and 1.0%) with varying frequencies of application (two to four times a day) and administration period (1–4 weeks) were examined on the ocular surface of rabbits, and administration of 0.1% BAC twice daily for two consecutive weeks was chosen to be optimum for dry eye induction in rabbits. It was found that lower concentrations were not able to cause tear deficiency, while higher concentrations were toxic to the ocular surface, resulting in scarring, corneal ulcers, and vascularization. Administration of 0.1% BAC demonstrated significant decrease in tear volume, mucin secretion, and conjunctival goblet cell density. In addition, scanning electron microscopy (SEM) and transmission electron microscopy (TEM) images of conjunctival epithelial cells exhibited a significant loss of microvilli structures that aid in the attachment of mucin to the cornea. Therefore, this model simulated human DED in terms of both aqueous and mucin deficiency. However, there was no evidence of continuation of DED symptoms after BAC stoppage. In the subsequent years, Tseng et al.[18] made modifications to the existing protocol and administered 0.1% BAC thrice daily for 4 weeks. The results demonstrated inflamed cornea with higher levels of inflammatory cytokines such as interleukin (IL)-8, IL-6, and tumor necrosis factor-alpha (TNF-α), reduced tear secretion, and fluorescent green stain patches on the cornea due to epithelial defects. This model established a stable DED state with moderate pathology of DED and showed no sign of automatic healing of the ocular surface after BAC discontinuation.

Figure 2.

BAC-induced pathological changes in the corneal and conjunctival layers of animal model. BAC = benzalkonium chloride

Table 1.

Summary of different BAC conditions used to establish dry eye animal model

Study	BAC (%) Animal model	Frequency and period of administration (days/weeks)	Effects on the ocular surface	Model stability after BAC stoppage
Xiong et al.[9]	0.1% BAC Rabbit model	Twice daily, 2 weeks	↓ Tear secretion ↑Fluorescein staining score ↓Goblet cell density ↓MUC5AC ↓No. and size of microvilli in corneal epithelium Thinner conjunctiva and corneal epithelium	No evidence
Tseng et al.[18]	0.1% BAC Rabbit model	Thrice daily, 4 weeks	↓Tear secretion ↑Fluorescein staining	Good
Li et al.[19]	0.1% BAC Rabbit model	Twice daily, 5 weeks	↓Tear volume ↑Fluorescein staining score ↓Goblet cell density ↓No. of microvilli in corneal surface Thinner corneal epithelium	Good (2 weeks)
Carpena- Torres et al.[20]	0.2% BAC Rabbit model	Twice daily, 5 days	↑Tear volume without anesthesia ↓Tear volume with anesthesia ↓TBUT ↑IL-6 mRNA levels ↓Goblet cell density and height of mucin	Good
Lin et al.[14]	0.2% BAC Mice model	Twice daily, 1 week	↓Tear volume ↑Inflammatory index in cornea ↓TBUT↓MUC5AC  Microvilli partially destroyed  Corneal epithelium apoptosis	Good

BAC=Benzalkonium chloride, IL-6=Interleukin-6, TBUT=Tear film breakup time

Another study determined the stability and validity of BAC-induced DED rabbit model by dividing 80 rabbits into four groups and instilling 0.1% BAC twice daily for 2, 3, 4, and 5 weeks in one eye from each rabbit.[19] A stable rabbit DED model was achieved by inducing BAC for 5 weeks. It showed a significant decrease in goblet cell density and MUC5AC along with corneal and conjunctival metaplasia till 3 weeks after BAC discontinuation. In addition, there was a marked difference in tear secretion and fluorescein staining till 2 weeks after BAC removal. Therefore, this model can be used to evaluate treatments for dry eye associated with mucin deficiency along with water-deficient dry eye. Subsequently, Carpena-Torres et al.[20] developed a DED rabbit model with a higher concentration of BAC (0.2%) administered twice daily for five consecutive days onto the ocular surface of both the eyes of five rabbits (n = 10) to reduce the number of BAC instillation days compared to the original model. Both the eyes of another five rabbits (n = 10) were treated as controls. In spite of the high BAC concentration, this model showed no sign of neovascularization or corneal ulcer. In addition, there was a significant increase in corneal staining, tear secretion without anesthesia, and IL-6 levels, along with a decrease in goblet cells, TBUT, and tear secretion with anesthesia compared to the control group, showing proper DED signs. However, there are certain limitations of this study, which could be improved by proper evaluation of the ocular surface of rabbits every day to further reduce the number of days of BAC instillation; in addition, administration of topical anesthesia is required to measure tear osmolarity.

Besides rabbit, mice dry eye model was created by Lin et al.,[14] who administered 0.2% BAC twice daily for 1 week onto the right ocular surface of 10 mice, while the other 10 mice were treated as phosphate buffered saline (PBS) control. It caused inflammation, squamous metaplasia, and epithelial cell death, which were in accordance with the human DED. Therefore, based on the results, these dry eye models can be further used in the study of inflammatory dry eye.

Therapeutic experimental studies using BAC model

Many experimental therapeutic studies have used these animal models to evaluate the efficacy of eye drops and other drug delivery vehicles, such as drug-eluting contact lenses and nanosystems containing various active ingredients, for DED treatment. These treatments alleviate the signs of DED on the ocular surface of DED animal models by restoring the tear film and corneal epithelial cell membrane and decreasing the inflammatory cytokines. Table 2 summarizes the data from the therapeutic studies using BAC animal model.

Table 2.

Summary of therapeutic studies using BAC-induced DED animal models

Study	Drug/molecule	BAC induction method in animal models	Treatment time	Outcomes
				Upregulated	Downregulated
Thacker et al.[21]	Artificial tears based on Bletilla striata polysaccharide	0.1% BAC, thrice daily, 4 weeks in rabbits (n=4 × 4 groups)	3 weeks, twice daily after BAC stoppage	Tear volume Corneal epithelium and thickness recovered	Fluorescein staining
Carpena- Torres et al.[22]	Artificial tears containing extracts of Artemia salina	0.2% BAC, twice daily, 5 days in rabbits (n=5 × 4 groups)	Treatment instilled thrice per day for 5 days along with BAC instillation	Tear volume TBUT Height of mucin cloud	Fluorescein staining IL-1β and MMP-9 levels
Tseng et al.[23]	Artificial tears containing EGCG and HA	0.1% BAC, thrice daily, 4 weeks in rabbits (n=4 × 5 groups)	3 weeks, twice daily after BAC stoppage	Tear secretion Corneal epithelium and stroma recovered	IL-8, IL-6, and TNF-α levels Fluorescein staining
Tseng et al.[24]	Virally inactivated SEDs	0.1% BAC, thrice daily, 4 weeks in rabbits (n=2 × 5 groups)	3 weeks, twice daily after BAC stoppage	Tear volume IL-6 expression Corneal epithelium layers recovered	Fluorescein staining IL-1β and TNF-α levels
Ehrenberg et al.[25]	Eye drops containing HA and PVP	0.1% BAC, twice on days 0, 3, 4, 5, 6, 7, 10, 11, 12, 13, 14 in rabbits (n=4 × 4 groups)	Twice, 5 min after BAC instillation on days 0, 3, 4, 5, 6, 7, 10, 11, 12, 13, 14	Corneal epithelium layers and goblet cells restored	Fluorescein staining
Choi et al.[26]	CsA eluting contact lenses	0.1% BAC, twice daily, 2 weeks in rabbits (n=6 × 6 groups)	Two weeks after BAC discontinuation	Tear volume TBUT Goblet cell density	Fluorescein staining IL-1β and IFN-γ expression
Li et al. [27]	Poly (catechin)- capped gold nanoparticles containing amfenac	0.15% BAC, twice daily, 2 weeks in rabbits (n=6 × 6 groups)	Twice daily for 4 days after BAC stoppage	MUC5AC expression Corneal epithelium layers restored	ROS and inflammation Fluorescein staining
Kim et al.[28]	EEDK drops	0.2% BAC, twice daily, 2 weeks in mice (n=8 × 5 groups)	Daily for 14 days after BAC discontinuation	Tear volume TBUT	Fluorescein staining IL-1β, IL-1α, TNF-α, IL-6 expression Inhibited squamous cell metaplasia and epithelial apoptosis
Kang et al.[29]	EERV drops	0.2% BAC, twice daily, 2 weeks in mice (n=6 × 5 groups)	3 days after BAC administration for 11 days	Tear volume TBUT Corneal smoothness	Cytochrome c and Bax expression Inhibited squamous cell metaplasia and inflammation
Beyazyildiz et al.[15]	Corticosteroids LE and PA	0.2% BAC, twice daily, 1 week in rats (n=10 × 3 groups)	Twice daily for 7 days after BAC discontinuation	Tear volume TBUT	Fluorescein staining Inflammation in cornea, conjunctiva, and meibomian gland

BAC=benzalkonium chloride, CsA=cyclosporine A, DED=dry eye disease, EEDK=ethanol extract of Diospyros kaki, HA=hyaluronic acid, IFN=interferon, IL=interleukin, LE=loteprednol etabonate, MMP-9=matrix metalloproteinase-9, PA=prednisolone acetate, PVP=polyvinylpyrrolidone, ROS=reactive oxygen species, SED=serum eye drop, TBUT=tear film breakup time, TNF=tumor necrosis factor

One such active ingredient is a natural plant-based polysaccharide, Bletilla striata polysaccharide, developed as an artificial tear for DED treatment in a study by Thacker et al.[21] Administration of this twice daily for 3 weeks in 0.1% BAC-induced DED rabbits significantly improved tear volume, corneal epithelium layers, and corneal thickness. Besides, upon instillation of this eye drops decreased fluorescein staining score as compared to the BAC-induced rabbits. Similarly, extract of Artemia salina[22] was evaluated in a 0.2% rabbit BAC model. There was a marked increase in tear volume, TBUT, and height of mucin cloud along with a decrease in corneal staining and inflammatory markers in conjunctival cells, such as matrix metalloproteinase (MMP)-9 and IL-1β, compared to the DED model. However, the major limitation of this study includes short-term treatment with A. salina as 0.2% BAC did not reproduce a stable DED condition as in previous studies. Polyphenols, such as epigallocatechin gallate (EGCG), are known to exhibit anti-inflammatory properties. Therefore, Tseng et al.[23] investigated the effects of artificial tears containing EGCG and hyaluronic acid (HA) on 0.1% BAC-induced DED rabbits. Administration of these drops twice daily for 3 weeks increased the tear volume and decreased proinflammatory marker levels, such as IL-8, IL-6, and TNF-α, along with apoptotic cells. In addition, EGCG/HA artificial tears revived normal corneal epithelium layers and stroma in the BAC model. Tseng et al.[24] yet again evaluated the effect of another molecule in a different study. Instillation of virally inactivated serum eye drops (SEDs) on the ocular surface of BAC-induced rabbit DED model improved tear secretion and the epithelial defects caused by BAC. In addition, normal corneal epithelium was revived in SED-treated corneas. However, it selectively reduced inflammatory markers in the eyes, for example, IL-1β and TNF-α levels were markedly reduced in the treated corneas, whereas IL-6 expression was higher. Ehrenberg et al.[25] evaluated the efficacy of a single eye drop containing sodium hyaluronate (HA) and polyvinylpyrrolidone (PVP) in rabbit DED model. DED was induced in rabbits by the procedure suggested by Xiong et al.[9] They observed significant reduction in corneal fluorescein staining and restoration of corneal epithelium and conjunctival goblet cells in DED rabbit models treated with the eye drops.

Various drug delivery vehicles such as drug-eluting contact lenses and nanosystems have also been tested for their therapeutic effects using BAC DED animal models. Choi et al.[26] developed cyclosporine A (CsA)-eluting contact lenses to treat DED using 0.1% BAC rabbit model. It demonstrated a significant increase in tear secretion, TBUT, and goblet cell density and, on the other hand, lowered fluorescein staining score and IL-1β and interferon-gamma (IFN-γ) levels in the treated group. Li et al.[27] evaluated the efficacy of amfenac-incorporated poly (catechin)-capped gold nanoparticles to treat DED in 0.15% BAC rabbit model. It significantly recovered damaged corneal epithelial layers and reduced DED-related inflammation and reactive oxygen species (ROS). It also increased the mucin content in DED models.

Other researchers have used BAC-induced mouse dry eye model to study the therapeutic effects of various anti-inflammatory agents. One such study investigated the efficacy of ethanol extract of Diospyros kaki (EEDK), which is an herbal medicine with anti-inflammatory properties.[28] It demonstrated a therapeutic effect on 0.2% BAC-induced mouse model of DED. It significantly increased TBUT and tear volume and decreased fluorescein staining score and the levels of inflammatory cytokines such as IL-1β, IL-1α, TNF-α, and IL-6. It also inhibited squamous cell metaplasia and corneal epithelial apoptosis. Similarly, ethanol extract of the leguminous plant Rhynchosia volubilis Loureiro (EERV)[29] was tested for its efficacy in the BAC mice model originally described by Lin et al.[14] However, the treatment was administered along with BAC instillation. It markedly improved tear secretion, TBUT, and corneal smoothness in the treated corneas. In addition, it also significantly downregulated cytochrome c and Bax levels, which inhibit corneal epithelial apoptosis. Another study compared the effects of two corticosteroids, loteprednol etabonate (LE) and prednisolone acetate (PA), for dry eye treatment in 0.2% BAC-induced rat models.[15] One week treatment with either of the corticosteroids successfully alleviated corneal and conjunctival epithelial damage and inflammation in the cornea. In addition, it significantly improved tear volume and TBUT. However, there was no significant difference between them.

Similarly, other active ingredients in the form of eye drops, such as amniotic membrane,[30] epidermal growth factor, HA–nimesulide conjugates,[31] capsanthin,[32] and CsA drops,[33] were also evaluated using the above-mentioned BAC models of DED.

Comparison of different BAC concentrations

In our laboratory, we conducted a dose-dependent study and compared the effects of different concentrations of BAC (catalog no. 12060; Sigma-Aldrich, St. Louis, MO, USA) over a period. BAC in concentrations of 0.1%, 0.15%, and 0.2% was instilled onto the ocular surface of rabbits, respectively, for two consecutive weeks. Various diagnostic tests such as Schirmer’s test, corneal fluorescein staining, and corneal smoothness were conducted.

Tear secretion was measured using Schirmer’s test, wherein the Schirmer’s paper strip was inserted into the lower conjunctival sac and the wetted length (mm) was measured after 5 min. This procedure was performed both before and after anesthesia. Interestingly, it was observed that there was an increase in tear volume before anesthesia in all the groups, which can be accounted for reflex lacrimation, while the tear volume decreased in all the groups when measured after anesthesia. Corneal fluorescein staining was performed by instilling 2 μl of 1% fluorescein sodium into the conjunctival sac of rabbits and examining the ocular surface under the blue filter of a slit-lamp microscope. It was observed [Fig. 3a and b] that with the increase in the concentration of BAC, the fluorescein staining score increased. Subsequently, corneal smoothness was performed by illuminating the ocular surface with ring light. Smoothness of the cornea was determined by the regularity of the reflected rings. The corneal irregularity was scored based on the number of distorted quarters of the ring light on the surface as follows: 0 = no distortion, 1 = distortion in one quarter, 2 = distortion in two quarters, 3 = distortion in three quarters, 4 = distortion in all the quarters, and 5 = severe distortion, wherein the ring is not recognized. In this experiment, corneal smoothness index score increased, deducing increased corneal irregularity with an increase in BAC concentration [Fig. 3]. However, there was no sign of corneal ulcer or neovascularization seen in any of the models. It was observed that rabbits instilled with 0.2% BAC continued to show signs of DED for 3 weeks after BAC stoppage, while it was around 1–2 weeks for rabbits administered with 0.1% and 0.15% BAC.

Figure 3.

Clinical imaging and scoring after 2 weeks of BAC administration. Columns 1–3 comprise corneas instilled with 0.1%, 0.15%, and 0.2% BAC, respectively. (a) First row shows the slit-lamp images of corneas after fluorescein staining, while the second row shows the corneal smoothness images. Graphs represent the (b) fluorescein staining score and (c) corneal smoothness index. BAC = benzalkonium chloride

Summary and Future Perspectives

The establishment of an appropriate DED animal model is of utmost importance for evaluating new therapeutic treatments and to study the pathology of DED better. Therefore, this review provides an insight into the existing BAC-induced DED models of rabbits, mice, and rats. Different models have been developed using varying conditions in terms of percentage, frequency, and period of BAC administration. While some may continue to show DED symptoms for weeks after BAC discontinuation, others may cease to show. Accordingly, by understanding each model’s behavior, therapeutic treatments should be administered. The main advantage of creating a BAC model is that it simulates human DED and represents all the hallmarks of DED perfectly. It is also a relatively simple model to create as it only requires instillation of eyedrops rather than surgical interventions or environmental modifications. However, the BAC animal model can at times lack reproducibility, for example, tear secretion was lowered in previous studies of the same model, but few studies showed increase in tears after BAC administration due to reflex lacrimation. An ideal animal model of DED should be reliable, so that it is possible to replicate the tests performed in previous studies for a uniform comparison. Creating this model in rabbits is also advantageous because clinical DED diagnostic platforms can be used to evaluate the outcomes and monitor the response to therapies.

Conclusion

In conclusion, in this review, we summarized the previously described BAC animal models of DED and shared original data on rabbit DED models created using various concentrations of BAC administration. The BAC-induced DED animal models are straightforward to create and, especially, rabbit models have the potential to be used for preclinical validation of newer emerging therapies for DED.

Financial support and sponsorship

This work was funded by the Hyderabad Eye Research Foundation (HERF), Hyderabad, India.

Conflicts of interest

There are no conflicts of interest.