Effect of Chitosan-N-Acetylcysteine Conjugate in a Mouse Model of Botulinum Toxin B–Induced Dry Eye

Objective  To evaluate the effect of a thiolated polymer lubricant, chitosan-N-acetylcysteine conjugate (C-NAC), in a mouse model of dry eye.

Methods  Eye drops containing 0.5% C-NAC, 0.3% C-NAC, a vehicle (control group), artificial tears, or fluorometholone were applied in a masked fashion in a mouse model of induced dry eye from 3 days to 4 weeks after botulinum toxin B injection. Corneal fluorescein staining was periodically recorded. Real-time reverse transcriptase–polymerase chain reaction and immunofluorescence staining were performed at the end of the study to evaluate inflammatory cytokine expressions.

Results  Mice treated with C-NAC, 0.5%, and fluorometholone showed a downward trend that was not statistically significant in corneal staining compared with the other groups. Chitosan-NAC formulations, fluorometholone, and artificial tears significantly decreased IL-1β (interleukin 1β), IL-10, IL-12α, and tumor necrosis factor α expression in ocular surface tissues.

Conclusions  The botulinum toxin B–induced dry eye mouse model is potentially useful in evaluating new dry eye treatment. Evaluation of important molecular biomarkers suggests that C-NAC may impart some protective ocular surface properties. However, clinical data did not indicate statistically significant improvement of tear production and corneal staining in any of the groups tested.

Clinical Relevance  Topically applied C-NAC might protect the ocular surface in dry eye syndrome, as evidenced by decreased inflammatory cytokine expression.

Dry eye syndrome (DES) is a highly prevalent ocular surface disease. Approximately 5 million Americans aged 50 years and older were estimated to have moderate to severe dry eye.^1,2 The management of DES encompasses both pharmacologic and nonpharmacologic treatments, including environmental management, avoidance of exacerbating factors, eyelid hygiene, tear supplementation, secretagogues, punctal plugs, anti-inflammatory agents, moisture chamber, and even salivary gland autotransplantation.^3,4 Despite many available options, there is no ideal or curative therapeutic treatment for DES at present.

Chitosan is a polysaccharide derived from chitin that shows structural characteristics similar to glycosaminoglycans, major components of the corneal stromal extracellular matrix. Chitosan exhibits many favorable biological behaviors, such as bioadhesion, low toxicity, and excellent biocompatibility, and also has interesting physical chemical characteristics, which make it a unique material for the design of ocular drug-delivery vehicles.^4-6 Precorneal retention time of medication significantly increases in chitosan-based ophthalmic formulations. Chitosan also has an antibacterial effect in topical applications of low concentration.⁷

N-acetylcysteine (NAC) is a derivative of amino acid L-cysteine and is a reducing agent with antioxidative activity. It is well known for its mucolytic and anticollagenolytic properties and is still widely used to reduce viscosity in many bronchopulmonary disorders.⁸N-acetylcysteine has been used in ophthalmology in topical form to treat many corneal and external eye diseases, such as corneal ulcers,⁹ alkaline-burned corneas,¹⁰ chronic blepharitis,¹¹ filamentous keratitis,¹² and keratoconjunctivitis sicca.^12,13 It has also been shown to promote corneal wound healing in dogs.¹⁴

Patients with DES have poor ocular surface wettability and tear clearance. These patients may develop filamentous keratitis, and their condition can progress to corneal ulceration and secondary infection. Therefore, the strategy of compounding both chitosan and NAC may provide added benefits for the treatment of DES. It is postulated that the modification of the cationic biopolymer chitosan with thiol groups will increase its residence time on the ocular surface via interaction with cysteine-rich subdomains of mucus glycoproteins.^15-17Especially in the case of mucus-deficient dry eye, the polymer will possibly form a protective barrier on the ocular surface and stabilize the tear film.^18,19 The purpose of this study was to investigate the effect of a new topically applied thiolated polymer, chitosan-NAC (C-NAC), in the mouse model of dry eye induced with botulinum toxin B (BTX-B).

Fifty 6- to 8-week-old CBA/J female mice (Harlan Laboratories, Indianapolis, Indiana) were used in accordance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research. The Institutional Animal Care and Use Committee at Johns Hopkins University approved the experimental protocols. Dry eye was induced in all mice with BTX-B injection into the right lacrimal glands (only 1 eye from each mouse was used for study). The BTX-B–induced dry eye murine model has been shown to significantly decrease tear production for up to 4 weeks and increase corneal staining for up to 8 to 10 weeks after an intralacrimal gland injection in CBA/J mice.²⁰ The mice were randomly assigned to 5 groups. The topical drops were applied twice daily from day 3 to 4 weeks after BTX-B injection. The outcome assessor and data analyzer were masked to treatment.

One group received topical C-NAC, 0.3%, and another received C-NAC, 0.5%, in isotonic borate-buffered solution (pH 6.5) eye drops. The control group received a topical vehicle alone (isotonic borate-buffered solution). The fourth group received topical artificial tears (methylhydroxypropylcellulose), and the final group received fluorometholone eye drops. Chitosan-NAC was synthesized by formation of amide bonds between the amino groups of chitosan glucosamine subunits and the carboxylic acid function of N-acetylcysteine under good manufacturing practice conditions with batch sizes of 20 g. We used 170-μM/g covalently attached NAC, which corresponds to the modification of every 30th glucosamine subunit with NAC. All medications were stored at 4°C until used.

Outcome measures included tear production using phenolsulfonphthalein threads, ocular surface damage grading using a corneal fluorescein staining photographic grading system, determination of blink rates, and gross determination of overall animal health. These parameters were recorded at baseline, 3 days before treatments were given, and 1, 2, and 4 weeks after BTX-B injection.

At 4 weeks after injection, mice were euthanized in a carbon dioxide chamber according to American Veterinary Medical Association guidelines. Ocular surfaces, including the cornea and conjunctiva, and lacrimal glands were harvested for real-time reverse transcriptase–polymerase chain reaction (RT-PCR) and immunofluorescence staining for various inflammatory cytokines, including macrophage migration-inhibitory factor (MIF), Toll-interacting protein, interleukins IL-1β, IL-10, and IL-12α, and tumor necrosis factor α (TNF-α), which has been shown to be upregulated in BTX-B–induced dry eye in the murine model.²¹ Tumor necrosis factor α, IL-1β, and MIF were chosen for immunofluorescence study because they appeared to be highly expressed.

Two researchers observed the blink rate for 30 seconds under constant temperature and humidity. If the blink rate differed by more than 2 blinks between observers, a recount was performed. Then, measurements of aqueous tear production and corneal fluorescein staining were performed, as previously reported.^21-24

In brief, phenolsulfonphthalein-saturated cotton threads were applied to the ocular surfaces at the lateral canthus using jeweler's forceps for 15 seconds in the unanaesthetized mouse. The wet threads were measured under a microscope using micron-scale digital calipers.

Corneal fluorescein staining with 1 μL of sodium fluorescein, 0.25%, was photographed with a digital camera equipped with 2 ×10-magnification macrolenses 1 minute after administration under cobalt blue light. All photographs were graded by 2 masked observers using a photograph-grading system that was modified from Nakamura et al.²⁵ The total area of staining was designated as grade 0 when there was no punctate staining, grade 0.5 when less than of the total corneal area was stained, grade 1 when 1/16 to was stained, grade 2 when more than ¼ to was stained, grade 3 when greater than ½ was stained, and grade 4 when the entire area was stained. Interobserver variability was estimated and controlled at the beginning of the study using κ statistics.

Eight eyes from each group were used for real-time RT-PCR. Total RNA was extracted from the lacrimal glands and ocular surfaces (cornea and conjunctiva) separately using the RNeasy Kit (Qiagen, Valencia, California) according to the manufacturer's instructions. Concentrations of RNA were determined by ultraviolet spectrophotometry. The first-strand complementary DNA was synthesized from 0.4 μg of lacrimal gland total RNA and 0.05 μg of ocular surface total RNA with oligonucleotide primers using a commercially available kit (SuperScript III Reverse Transcriptase; Invitrogen, Carlsbad, California). Samples of complementary DNA were aliquoted and stored at –20°C until use.

Real-time quantitative RT-PCR was performed and analyzed in a thermocycler. Reactions were performed in a 20-μL volume using the RT² SYBR Green qPCR Master Mix with RT² qPCR Primer Assay for Mouse (SABiosciences, Frederick, Maryland). The RT-PCR was done using the recommended parameters (95°C for 10 minutes then 40 cycles at 95°C for 15 seconds and at 60°C for 60 seconds). Fluorescence from SYBR green that attached to the double-stranded DNA was measured at the end of each cycle. The β-actin gene, a housekeeping gene, was used as a standard for normalization. Assays were performed in duplicate (Table 1). Melting curve analysis was performed to ensure good-quality specific PCR product. The real-time RT-PCR results were analyzed using the relative standard curve method and compared and calibrated against the control group.

Two eyes from each group were used for immunofluorescence study. Briefly, the lacrimal gland and ocular surface (cornea and conjunctiva) were harvested and preserved in paraformaldehyde, 4%, at 4°C. The tissues were cryoprotected with 30% sucrose incubation overnight before they were embedded and frozen in optimal cutting temperature embedding compound. The tissue blocks were then cryosectioned at 8 μm, placed on specially charged slides, and preserved at −20° C. After thawing at room temperature, slides were fixed with acetone for 10 minutes. The slides were washed with phosphate-buffered solution between each step. The fixed slides were incubated with donkey-blocking serum, 10%, for 20 minutes and then with primary antibodies for 1 hour. The primary antibodies used were 8-μg/mL goat polyclonal IgG anti–TNF-α, anti–IL-1β, and anti-MIF. Fluorescein isothiocyanate-conjugated secondary antibodies (0.4-μg/mL donkey anti–goat IgG with fluorescein isothiocyanate) were applied and incubated in a dark chamber for 45 minutes. After washing with phosphate-buffered solution, slides were mounted with mounting medium containing 4′6-diamidino-2-phenylindole·2HCl, which counter-stained the nucleus. Sections were visualized with a fluorescence digital microscope. Negative control experiments were performed by omitting the primary antibody incubation. Isotype control was also performed. All images were photographed with fixed exposure time for standard comparison.

The program SPSS, version 13.0 (SPSS Inc, Chicago, Illinois), was used for statistical analysis. The Mann-Whitney U test was used to compare mean corneal fluorescein staining score, tear production, and relative cytokine expression between the treated groups and the control group. The Wilcoxon signed rank test was used to compare change in mean corneal fluorescein staining score, tear production, and relative cytokine expression from baseline to after treatment. Two-sided P <. 05 was considered statistically significant. Intraobserver variability was calculated using κ statistics. A κ value of 0.4 or more indicated a good rate of interobserver agreement and reproducibility.

Only the group treated with artificial tears had a significantly higher blink rate than the control group at 1 week after BTX-B injection (P = .02). Other groups had no statistically significant differences in blink rate compared with the control group at any time (Figure 1A).

At baseline, there were no statistically significant differences in aqueous tear production between groups (Table 2). No significant decrease in tear production was observed between baseline and 3 days after injection. However, overall there were nonsignificant trends toward decreased tear production at 2 and 4 weeks after injection. Mice treated with C-NAC, 0.5%, displayed significant decrease in tear production at 2 weeks (mean [standard deviation], 1.96 [1.4] mm/15 seconds, P = .02) and 4 weeks (2.29 [1.3] mm/15 seconds, P = .04) compared with 3 days after injections (3.83 [1.7] mm/15 seconds). The control group also displayed a significant decrease in the tear production at 2 weeks (2.10 [1.06] mm/15 seconds) compared with 3 days after injections (3.44 [1.9] mm/15 seconds, P = .02). The artificial tear treatment group had significantly greater tear production (3.46 [1.25] mm/15 seconds) than the vehicle alone control group (2.04 [0.94] mm/15 seconds) at 4 weeks (P = .01). Other study groups showed no significant differences in aqueous tear production compared with the control group at any time.

A κ statistic of 0.53 reflected good agreement in obtaining the corneal fluorescein staining scores between 2 masked observers. At baseline, minimal punctate corneal fluorescein staining was observed in every group (Figure 1B). No statistically significant differences in staining score between groups were recorded at baseline. After combining data from each group, corneal fluorescein staining scores were significantly increased from baseline to 3 days after injection through 4 weeks after injection (all P < .002) and more prominent at the center of the cornea. Every group had a peak corneal staining score at 2 weeks after injection except for mice treated with C-NAC, 0.5%; this group had a peak staining score at 1 week, which then declined. The fluorometholone-treated group tended to have the lowest staining score compared with other groups but there was no statistically significant difference. Among the treatment groups, the group treated with C-NAC, 0.5%, tended to have a lower staining score at 2 weeks compared with others. However, no statistically significant differences were demonstrated in corneal staining score between any of the treatment groups and the control group.

One mouse died of anesthetic overdose on day 2 of the study. All other mice were active and generally well. No abnormal eye redness or discharge was noted during follow-up. Neither ptosis nor abnormal eyelid aperture were detected after BTX-B injection.

Chitosan-NAC, 0.5%, significantly suppressed ocular surface MIF expression compared with the control, artificial tears, and even the fluorometholone groups. Toll-interacting protein expression levels in ocular surfaces also significantly increased in the artificial tear and fluorometholone groups (Figure 2). The IL-1β, IL-10, IL-12α, and TNF-α expression levels significantly decreased in ocular surfaces of groups treated with C-NAC, 0.3%, artificial tears, and fluorometholone. Chitosan-NAC, 0.5%, also significantly suppressed IL-1β and IL-12α expression compared with vehicle (control group) (Figure 2). The messenger RNA expression levels of all inflammatory cytokines in lacrimal glands demonstrated no significant differences among groups.

Immunofluorescence study of corneas and lacrimal glands demonstrated no apparent differences in cytokine expression between groups (data not shown). Immunofluorescence staining detected the presence of MIF, IL-1β, and TNF-α in conjunctival epithelium in all groups except the negative control group. The expression levels of MIF in the C-NAC–treated groups were less prominent than the control group (Figure 3). In contrast, MIF expression levels were significantly higher in the groups treated with artificial tears and fluorometholone. The expression levels of IL-1β were less prominent in the treatment groups, especially in the C-NAC–treated groups, compared with the control group (Figure 4). The expression levels of TNF-α in the groups treated with C-NAC, 0.5%, artificial tears, and fluorometholone were significantly less than in the control group (Figure 5).

The pathogenesis of DES is complex with multiple etiologies. Inflammation of the tear-secreting apparatus that results in compositional changes of the tear film and loss of its integrity is thought to be 1 of the important mechanisms that leads to ocular surface injury in DES.²⁶ Cell damage in the cornea and conjunctiva, by means of apoptosis and direct mechanical and/or osmotic stress, will stimulate the reflex neurosensory arc, in turn stimulating lacrimal gland and neurogenic inflammation and inflammatory involvement of the conjunctival epithelium.²⁷ Lacrimal gland dysfunction that causes tear hyperosmolarity can lead to the release of IL-1, IL-6, IL-8, TNF-α, and matrix metalloproteinase 9, 1, 13, and 3 from the ocular surface epithelium.^28-30 In our previous study using gene microarray analysis in the BTX-B–induced dry eye murine model, we found that MIF, Toll-interacting protein, IL-1β, IL-10, IL-12α, and TNF-α were highly expressed.²¹ Various anti-inflammatory treatments have been demonstrated to be useful for DES treatment.^31-34 At present, topical cyclosporine has become 1 of the standard treatments for patients with moderate to severe DES.^35-38 However, no single treatment successfully cures DES. Other medications have been evaluated for improving quality of life in patients with DES. Absolon and Brown¹³ conducted a double-masked crossover trial in patients with dry eye, comparing NAC alone with artificial tears; the objective findings in the NAC-treated group were significantly better than those in the artificial tear group, though there were no significant differences in subjective findings.

Although corneal fluorescein staining scores were lower in the C-NAC, 0.5%, and fluorometholone groups compared with the control group, the effect was not statistically significant. In our study, we demonstrated that C-NAC, 0.3%, can decrease ocular surface messenger RNA expression of IL-1β, IL-10, IL-12α, and TNF-α. A higher concentration of C-NAC (0.5%) was shown to suppress MIF expression significantly better than a vehicle, artificial tears, and (surprisingly) fluorometholone in addition to significantly decreasing the expression of IL-1β and IL-12α. These inflammatory cytokines are thought to be important in the pathogenesis of disease in the BTX-B–induced dry eye murine model.²¹ We postulate that the mucoadhesive properties of thiolated chitosan in conjunction with the antioxidative and mucolytic property of N-acetylcysteine played a role in protecting the ocular surface and suppressing the inflammatory response in the mice. Furthermore, our data support an anti-inflammatory role for C-NAC, which is a novel concept and begs further investigation.

Unexpectedly, overall tear production did not significantly decrease after the BTX-B injection. Because application of topical anesthesia results in some surface staining, we only measured tear production without anesthesia, which can only be accomplished during a short period in an awake animal. Thus, variability may arise, at least in part, from variability of the reflex portion of the tear response as well as from an error in the short 15-second test period. In addition, there may also be some variability in the injection and injection response.

In conclusion, this study has demonstrated, for the first time, the potential usefulness of the BTX-B–induced dry eye mouse model to evaluate new investigational medication for the treatment of dry eye in a preclinical laboratory study. Topical C-NAC can cause decreased expression of many important molecular biomarkers of dry eye disease and tended to improve the ocular surface. Chitosan-NAC, 0.3%, can significantly decrease expression of IL-1β, IL-10, IL-12α, and TNF-α. A higher concentration of C-NAC (0.5%) can also suppress MIF expression in the ocular surface of the BTX-B–induced dry eye mouse model. However, clinical data did not indicate statistically significant improvement of tear production or ocular surface staining compared with the control group, artificial tear group, or fluorometholone group.

Correspondence: Roy S. Chuck, MD, PhD, Wilmer Ophthalmological Institute, Johns Hopkins University, 255 Woods Bldg, 600 N Wolfe St, Baltimore, MD 21287 (rchuck1@jhmi.edu).

Submitted for Publication: December 16, 2008; final revision received January 30, 2009; accepted February 3, 2009.

Financial Disclosure: This study was supported by an unrestricted grant from Croma-Pharma.

Funding/Support: This study was supported by Research to Prevent Blindness through a core grant to the Wilmer Eye Institute.