Domesticated species commonly develop ulcerative keratitis, which can result in devastating consequences for the eye because ulceration can progress to corneal perforation. Loss of the corneal stroma following the initial corneal insult occurs owing to release of proteinases by etiologic agents (ie, bacteria or fungi), neutrophils, and corneal epithelial cells. The major proteinases that play a role in corneal degradation are MMPs and serine proteinases. Anticollagenase enzymes (ie, proteinase inhibitors) are present within the cornea and work to prevent corneal degradation by those proteinases. However, when proteinase inhibitors are overwhelmed by proteinases released following corneal insult, keratomalacia develops and the corneal ulcer deepens.1,2
Many compounds have anticollagenase properties and have been suggested for the treatment of corneal ulcers to prevent corneal degradation and keratomalacia. Some of those compounds are serum, plasma, EDTA, NAC, tetracyclines, and tetanus antitoxin.3–6,a Results of a study3 in which the in vitro effects of various concentrations of serum, NAC, and tetanus antitoxin on equine corneas were evaluated indicate that all 3 compounds provide corneas equal protection against collagenase digestion. In another in vitro study,4 serum, 0.2% EDTA, 0.1% doxycycline, 10% NAC, 0.2% ilomostat, and 0.1% and 0.5% α-1 proteinase inhibitor all inhibited MMP activity in the tear film of horses with ulcerative keratitis, with inhibition of MMP activity greatest in samples treated with EDTA and ilomostat.
Although the in vitro effects of single concentrations of many anticollagenase compounds have been assessed, to our knowledge, various concentrations of those compounds have not been compared to determine the most effective concentration. Nor have the effects of those compounds been evaluated when used in combination. Determination of the efficacy of combination treatment with anticollagenase compounds is important because most patients with ulcerative keratitis are concurrently treated with multiple compounds. Therefore, the purpose of the study reported here was to compare the effect of various anticollagenase compounds, alone and in combination, as well as at various concentrations, for the prevention of corneal degradation in an in vitro model.
Materials and Methods
Corneal specimens
Corneas were obtained from recently euthanized dogs and horses at the Purdue University Veterinary Teaching Hospital or local animal shelter. All animals were euthanized for reasons unrelated to the study. All corneas were examined by direct illumination prior to harvest and determined to have no evidence of disease. Each cornea was removed and sectioned into 4 (canine) or 6 (equine) pieces of approximately equal size. Corneal pieces were labeled by species and date and individually stored at −80°C until use. Corneal pieces were evenly distributed among control and treatment groups on the basis of storage time.
Serum samples
Blood samples (each approx 6 mL) were obtained from 5 healthy dogs and 5 healthy horses (donor animals). All dogs and some horses were owned by Purdue University employees, whereas the remaining horses were teaching animals owned by Purdue University. Consent was obtained from the owners of all employee-owned animals prior to blood collection. All live-animal procedures were approved by the Purdue University Animal Care and Use Committee. All donor animals were manually restrained, and blood was collected by jugular or cephalic venipuncture into serum separator tubes. Blood samples were allowed to sit undisturbed at room temperature (approx 22°C) for 1 hour after collection. Then, the serum from each sample was harvested and pooled by species. The pooled samples were stored at room temperature and used within 2 hours after collection.
In vitro corneal degradation
The methodology used for in vitro corneal degradation was as previously described.6 Briefly, corneal pieces were individually placed in plastic weigh boats and dried in an oven at 40°C for 3 hours, after which they were weighed and the pretreatment corneal weight recorded. Collagenase derived from Clostridium histolyticumb (800 U/mL) was added to 5mM calcium chloride-saline (0.9% NaCl) solution to prepare the incubation fluid. Saline solution was also used to create stock solutions of 2% EDTA,c 5% NAC,d 1% tetracycline,e doxycycline,f and minocycline.g Those stock solutions were then diluted to create test solutions of 1% and 0.3% EDTA; 0.5% and 0.1% tetracycline, doxycycline, and minocycline; and 1% and 0.5% NAC. For each species, 4 replicates were created for each of the 5 stock solutions and 13 test solutions as well as homologous serum and negative and positive controls. For each replicate, a 5-mL aliquot of incubation fluid was added to a 10-mL tube followed by 500 μL of homologous serum or one of the stock or test solutions and a homologous piece of cornea. For each combination test solution, 500 μL of each individual compound was added to the 10-mL tube. For each negative control replicate, 5 mL of 5mM calcium chloride in saline solution was added to a 10-mL tube followed by a piece of cornea. For each positive control replicate, 5 mL of incubation fluid was added to a 10-mL tube followed by a piece of cornea; no inhibitory compounds were added to the positive control replicates. Replicate tubes were incubated on a rocker at 40°C for 4 hours. Following incubation, the remaining corneal specimens were retrieved by pouring tube contents through a filter paper.h Each retrieved specimen was placed in its original plastic weigh boat and dried in an oven at 40°C for 3 hours. The corneal specimens were weighed, and the posttreatment corneal weight for each specimen was recorded. For each specimen, the percentage of corneal weight loss (degradation) was calculated as (pretreatment corneal weight - posttreatment corneal weight)/pretreatment corneal weight × 100.
Statistical analysis
For each species, the mean ± SD percentage corneal degradation was calculated for each of the 19 treatments (5 stock solutions, 13 test solutions, and serum) and negative and positive controls. A mixed linear regression model was used to compare the mean percentage corneal degradation among treatments for each species. Each model contained a fixed effect for treatment and a random effect for cornea nested within donor animal. The Scheffe post hoc method was used to control for type I error inflation owing to multiple pairwise comparisons. Values of P < 0.05 were considered significant. Results were graphically reported as box-and-whisker plots owing to the small number (n = 4) of replicates for each treatment and to demonstrate the distribution of results for each treatment relative to those for the negative and positive controls.
Results
The mean ± SD percentage corneal degradation was 91.0 ± 9.0% and 89.0 ± 9.6% for canine and equine positive control samples, respectively, and 11.2 ± 74% and 5.8 ± 4.6% for canine and equine negative control samples, respectively. For both canine and equine assays, the mean percentage corneal degradation was significantly lower (ie, treatment was more effective) than that for serum for treatments with tetracycline concentrations ≥ 0.5%, treatments with EDTA concentrations ≥ 0.3%, and treatments with NAC concentrations ≥ 0.5% (Figure 1). Additionally, for both species, the mean percentage corneal degradation did not differ significantly among the 3 concentrations (0.3%, 1.0%, and 2.0%) of EDTA (P = 0.99 for all 3 comparisons) or the 3 concentrations (0.5%, 1.0%, and 5.0%) of NAC (P = 0.99 for all comparisons of canine assays and P ≥ 0.89 for all comparisons of equine assays).
Box-and-whisker plots of the percentage corneal degradation (weight loss) for negative (NC) and positive (PC) controls and 19 anticollagenase treatments (serum [S]; 0.1% [0.1% D], 0.5% [0.5% D], and 1.0% [1% D] doxycycline; 0.1% [0.1% M], 0.5% [0.5% M], and 1.0% [1% M] minocycline; 0.1% [0.1% T], 0.5% [0.5% T], and 1.0% [1% T] tetracycline; 0.3% [0.3% E], 1.0% [1% E], and 2.0% [2% E] EDTA; 0.5% [0.5% N], 1.0% [1% N], and 5.0% [5% N] NAC; serum with 1.0% EDTA [S + 1% E]; serum with 0.5% tetracycline [S + 0.5% T]; and 1.0% EDTA with 0.5% tetracycline [1% E + 0.5% T]) following in vitro incubation with canine (A) and equine (B) corneal specimens. Each control and treatment was replicated 4 times for each species. For each treatment, replicates consisted of the incubation of a corneal specimen with 5 mL of incubation fluid (Clostridium histolyticum-derived collagenase [800 U/mL] in 5mM calcium chloride in saline [0.9% NaCl] solution) and 500 μL of the assigned treatment (or 500 μL of each constituent for combined treatments). For the negative control replicates, corneal specimens were incubated with 5 mL of 5mM calcium chloride in saline solution. For the positive control replicates, corneal specimens were incubated with 5 mL of incubation fluid without any anticollagenase compounds (ie, treatment). For each plot, the lower and upper limits of the box represent the 25th and 75th percentiles, the horizontal line within the box represents the median, and the whiskers delimit the range. Within each set of 3 concentrations for single-compound treatments, means for treatments denoted by different lowercase letters differ significantly (P < 0.05); the lack of lowercase letters indicates the means did not differ significantly among the 3 concentrations. *Mean differs significantly (P < 0.05) from that for the positive control. †Mean differs significantly (P < 0.05) from that for the serum treatment. ‡Mean differs significantly (P < 0.05) from that for the 0.5% tetracycline treatment. §Mean differs significantly (P < 0.05) from that for the 1.0% EDTA treatment.
Citation: American Journal of Veterinary Research 79, 5; 10.2460/ajvr.79.5.555
For canine assays, the mean percentage corneal degradation was significantly lower than that for the positive control for all treatments, except serum (P = 0.99), 0.1% minocycline (P = 0.99), and serum with 0.5% tetracycline (P = 0.20). The mean percentage corneal degradation for treatments with 0.1% tetracycline, 0.1% doxycycline, and 0.1% minocycline was significantly greater (ie, treatment was less effective), compared with that for their respective 0.5% and 1.0% counterparts. However, the mean percentage corneal degradation did not differ significantly (P = 0.99 for all comparisons) between the 0.5% and 1% treatments for any of the 3 tetracyclines.
For equine assays, the mean percentage corneal degradation was significantly lower than that for the positive control for all treatments, except 0.1% doxycycline (P = 0.11) and 0.1% tetracycline (P = 0.064). The 0.1% tetracycline and 0.1% doxycycline treatments were significantly (P < 0.001 for all comparisons) less effective than their 0.5% and 1% counterparts. However, the mean percentage corneal degradation did not differ significantly (P = 0.99 for all comparisons) between the 0.5% and 1.0% treatments for either tetracycline or doxycycline or among the 3 minocycline (0.1%, 0.5%, and 1.0%) treatments.
The efficacy for each of the 3 combination treatments was compared with the efficacy for each of their respective constituents. For the canine assays, the efficacy of the serum with 1.0% EDTA treatment was significantly (P < 0.001) greater than that of the serum treatment but did not differ significantly (P = 0.22) from that of the 1.0% EDTA treatment. For the equine assays, the efficacy of the serum with 1% EDTA treatment was significantly (P < 0.001) less than that of the 1.0% EDTA treatment but did not differ significantly (P = 0.22) from that of the serum treatment. For both canine and equine assays, the efficacy of the serum with 0.5% tetracycline treatment was significantly (P < 0.001 for both comparisons) less than that of the 0.5% tetracycline treatment but did not differ significantly (P = 0.99 for both comparisons) from that of the serum treatment. Additionally, for both canine and equine assays, the efficacy of the 1.0% EDTA with 0.5% tetracycline treatment did not differ significantly from that of the 1.0% EDTA (P = 0.99 for both comparisons) or 0.5% tetracycline (P ≥ 0.69 for both comparisons) treatment.
Discussion
Results of the present in vitro study indicated that, for both canine and equine corneal specimens, the mean percentage degradation following treatment with tetracyclines at concentrations ≥ 0.5%, EDTA at concentrations ≥ 0.3%, and NAC at concentrations ≥ 0.5% was significantly less than that for the positive control (corneal specimens incubated with collagenase but without an inhibitor), which suggested that those treatments might be efficacious for the prevention of corneal degradation in dogs and horses with ulcerative keratitis. Additionally, the mean percentage corneal degradation was significantly less than that of the positive control for equine specimens treated with serum but not for canine specimens treated with serum. Serum was also found to be less efficacious than the treatments with tetracycline (tetracycline, doxycycline, or minocycline) concentrations ≥ 0.5%, treatments with EDTA concentrations ≥ 0.3%, and treatments with NAC concentrations ≥ 0.5%. In both canine and equine assays, the 0.5% tetracycline and 1% EDTA treatments when used alone were as efficacious as or more efficacious than when used in combination with serum. In the present study, following incubation with serum, the mean ± SD percentage corneal degradation was 79 ± 15% for canine specimens and 57 ± 16% for equine specimens, which was similar to the mean ± SD percentage corneal degradation for canine (74 ± 15%) and equine (38 ± 12%) specimens incubated with serum in another study6 performed at our institution. However, in that study,6 serum significantly decreased corneal degradation relative to the positive controls for both species, whereas in the present study, serum decreased corneal degradation only in the equine assays. Although results of that other study6 suggest that the anticollagenase activity is similar between serum and plasma, we chose to use serum in the present study so that we could compare our results with those of other in vitro studies3,4 in which the extent of corneal protection for various anticollagenase compounds, including serum rather than plasma, was evaluated.
In veterinary patients, serum is often used as the first line of defense against keratomalacia, and the in vitro effects of serum on corneal degradation have been described for various veterinary species.6,a However, results of the present study indicated that other compounds are as efficacious as or more efficacious than serum for the prevention of corneal degradation. Although we do not discourage the use of serum in patients with keratomalacia, we propose that it may be beneficial to consider other anticollagenase compounds for use in conjunction with serum early in the treatment process. In addition to its anticollagenase properties, serum contains growth factors, such as epithelial growth factor, TGF-β1, and platelet-derived growth factor, that may be beneficial in the treatment of keratomalacia and corneal ulceration.7 Epithelial growth factor promotes corneal epithelial migration and proliferation that improves and accelerates wound healing.7 Expression of TGF-β1 in corneal epithelium increases during stromal repair.7 Even though TGF-β1 decreases keratocyte migration, it works synergistically with platelet-derived growth factor to promote myofibroblast differentiation.8
On the basis of the findings of the present study, tetracyclines, as a group, may be beneficial in reducing clinical keratomalacia. Tetracyclines chelate both calcium (a MMP cofactor) and zinc (a stabilizing ion of MMPs), which leads to MMP inhibition.4 In general, tetracyclines chelate zinc to a greater extent than calcium, and doxycycline is able to bind zinc more tightly than other members of the tetracycline group.9 For each of the 3 tetracyclines (tetracycline, doxycycline, and minocycline) evaluated in the present study, concentrations ≥ 0.5% significantly decreased corneal degradation (ie, inhibited collagenase) relative to that of the positive controls for both canine and equine specimens. Results of another study4 indicate that 0.1% doxycycline significantly inhibits proteases in the tear film of horses, but that concentration of doxycycline was not effective in inhibiting collagenase activity in the present study. However, the investigators of that study4 used gel zymography to measure protease activity rather than a corneal degradation model as was used in the present study, and differences between the experimental models may have contributed to the apparently conflicting results between the 2 studies.
The ability of doxycycline to penetrate the ocular tear film following oral administration has also been investigated. In 1 study,10 peak tear film doxycycline concentration ranged from 8.21 to 9.83 μg/mL (or 0.0008% to 0.0010%) following oral administration of the drug (20 mg/kg, q 24 h for 5 days) to horses, which was quite low relative to the doxycycline concentrations (0.1%, 0.5%, and 1.0%) evaluated in the present study. In another study,11 oral administration of minocycline (4 mg/kg, q 12 h for 5 days) to ponies resulted in a maximum tear film drug concentration of only 11.8 μg/mL (or 0.0012%), which was insufficient to inhibit MMP-2 and MMP-9 in vitro. In yet another study,12 oral administration of doxycycline (10 mg/kg, q 12 h for 5 days) to dogs resulted in a maximum tear film drug concentration of only 4.32 ng/mL (or 0.0000432%), and varying the doxycycline dose from 5 mg/kg to 10 mg/kg had no effect on the mean tear film drug concentration. Results of the present study suggested that the 0.1% doxycycline treatment was significantly less effective at inhibiting collagenase, compared with the 0.5% and 1.0% doxycycline treatments. Because reported tear film drug concentrations following oral administration of tetracyclines are much lower than 0.1%, it is questionable whether oral administration of tetracyclines is clinically beneficial for patients with keratomalacia. Therefore, if a tetracycline is being considered for a patient with keratomalacia solely for its anticollagenase effects, the topical route may be the preferred route of administration. Currently, oxytetracycline is the only tetracycline available in commercially available eye ointments. However, patients with keratomalacia often require frequent treatments, and solutions may be more beneficial than ointments for those patients. Also, in horses with keratomalacia, subpalpebral lavage is facilitated by the use of anticollagenase solutions.
For patients with corneal defects, topical administration of tetracyclines, particularly oxytetracycline, promotes corneal epithelial wound healing.13 That effect is suspected, but not yet proven, to be the result of tetracycline-induced upregulation of TGF-β1 and other specific growth and transcription factors that promote corneal reepithelialization. Therefore, tetracyclines may promote corneal healing in addition to their anticollagenase activities. Tetracyclines are also antimicrobial agents that can reduce the bacterial load in corneal ulcers, which decreases the collagenase load and aids in the prevention of further corneal degradation.
Ethylenediaminetetraacetic acid is a metal chelator and, similar to tetracyclines, chelates both calcium and zinc.14 It also decreases stimulation of neutrophil migration to corneal ulcers, which in turn decreases the release of proteinases.3 All 3 concentrations (0.3%, 1.0%, and 2.0%) of EDTA evaluated in the present study effectively decreased corneal degradation. The 0.3% EDTA treatment was chosen to mimic an EDTA solution that can be easily produced by the addition of 1.5 mL of sterile water to the contents of a 3-mL EDTA blood collection tube, whereas the 1.0% and 2.0% EDTA treatments represented commonly compounded EDTA solutions. It is important to note that, although all 3 concentrations of EDTA effectively prevented corneal degradation in vitro, EDTA can be potentially toxic to corneal epithelium in vivo. The effect of EDTA on corneal tissue has been investigated15 but only at concentrations (0.00001% to 0.01%) appropriate for use as a preservative, which are much lower than those required for anticollagenase activity. However, in 1 study,16 corneal epithelial toxicity of EDTA was evaluated at concentrations ranging from 0.0001% to 1.0%, and the mean ± SD corneal damage percentage was 5.03 ± 39.53% for the 0.0001% concentration and 57.86 ± 37.32% for the 1.0% concentration. That finding suggests that the EDTA concentrations evaluated in the present study have the potential to induce damage to the corneal epithelium in vivo, and further investigation into the potential corneal toxicity of EDTA in vivo is warranted.
All 3 concentrations (0.5%, 1.0%, and 5.0%) of NAC evaluated in the present study significantly reduced corneal damage relative to the positive controls. N-acetylcysteine is a derivative of the amino acid l-cysteine and inhibits collagenase irreversibly by reducing disulphide bonds and chelating calcium and zinc. It also inhibits MMP-9, although the exact mechanism by which it does so is unknown.17,18 In another in vitro study3 of the anticollagenase activity of NAC, a 2.0% concentration of NAC effectively decreased corneal degradation but concentrations of 0.2% and 0.02% did not. Because the results of that study3 suggest that the concentration breakpoint for NAC efficacy against corneal degradation is between 2% and 0.2%, we chose to evaluate NAC concentrations of 0.5% and 1.0% in the present study, both of which resulted in significant decreases in corneal degradation. The highest concentration of NAC chosen for evaluation in the present study was set at 5% because results of other studies3,19 indicate that NAC concentrations > 5% can be irritating to eyes. However, results of an in vivo study20 in which the effect of NAC concentrations of 3%, 10%, and 20% on corneal wound healing in dogs was investigated indicate that only the 3% solution significantly decreased wound healing time relative to that for control eyes that were treated with saline solution, although no adverse effects were reported for eyes treated with NAC at any of the 3 concentrations.
In both the canine and equine assays of the present study, the mean percentage corneal degradation for the serum with 0.5% tetracycline (combined) treatment did not differ from that for the serum only treatment but was significantly greater than that for the 0.5% tetracycline only treatment. Tetracycline binds to proteins in a variable manner; therefore, it is possible that a portion of the tetracycline in the combined treatment became bound to serum proteins, which decreased its anticollagenase activity relative to when it was used alone. Further research is necessary to assess the use of multiple anticollagenase agents simultaneously in a clinical setting. Specifically, additional investigation is required to determine whether the traditionally recommended 5-minute interval between administration of topical ophthalmic medications is sufficient to prevent interaction between the medications.
For the canine assays of the present study, the mean percentage corneal degradation for the serum with 1.0% EDTA treatment was significantly lower than that for the serum only treatment but did not differ significantly from that for the 1% EDTA only treatment. Therefore, the serum with 1% EDTA treatment did not appear to have a synergistic effect on the prevention of corneal degradation, but incubation with those 2 compounds in combination did not adversely affect the anticollagenase activity of EDTA, the more efficacious of the 2 compounds. Conversely, for the equine assays, the mean percentage corneal degradation for the serum with 1.0% EDTA treatment did not differ from that for the serum only treatment and was significantly greater than that for the 1.0% EDTA only treatment. Finally, the 1.0% EDTA with 0.5% tetracycline treatment did not have an additive or diminishing effect on the mean percentage corneal degradation, compared with the mean percentage corneal degradation for each of its constituent compounds.
In the present study, corneal degradation was measured on the basis of estimated collagen loss in specimens following incubation with a commercially available collagenase. The C histolyticum-derived collagenase used in this study is composed of 7 distinct proteases that fall into the categories of class I (γ) and class II (ζ) collagenases.21 These are zinc-dependent enzymes that are functionally related to MMPs; therefore, it was not surprising that collagenase inhibition was variable among the EDTA, NAC, and tetracycline treatments evaluated. Although that collagenase has been used to approximate keratomalacia in other in vitro studies,3,4,6 the collagenolysis it induces may not completely mimic keratomalacia in vivo because other sources of collagenases, including exogenous pathogens and endogenous cells, may be present in the diseased eyes of live patients. In serum, the most abundant proteinase inhibitor is α-2 macroglobulin, which is nonspecific and inhibits MMPs as well as other classes of proteinases such as serine proteinases, of which neutrophil elastase is the most abundant in vivo. The corneal degradation model used in this study did not account for different types of proteinases, continued release of proteinases, or repeated treatment with anticollagenase compounds. Therefore, in vivo studies of the anticollagenase effects of the compounds evaluated in this study are warranted.
To our knowledge, the present study was the first to compare the anticollagenase efficacy of multiple compounds at various concentrations alone and in combination in an in vitro corneal degradation model that involved corneal specimens from both dogs and horses. Results indicated that treatments with tetracyclines at concentrations ≥ 0.5%, treatments with EDTA at concentrations ≥ 0.3%, and treatments with NAC at concentrations ≥ 0.5% were all effective at preventing corneal degradation in both species. The efficacy for each of the 3 combination treatments (serum with 1% EDTA, serum with 0.5% tetracycline, and 1% EDTA with 0.5% tetracycline) evaluated was equivalent to or less than that for the individual compounds included in those treatments; thus, the combined treatments did not appear to have a synergistic effect on anticollagenase activity. Although in vivo studies are necessary to further evaluate the efficacy of anticollagenase compounds in patients with keratomalacia, results of the present study suggested that topical administration of EDTA, NAC, and tetracyclines may be useful for the prevention of corneal degradation.
Acknowledgments
Supported by a grant from the Willis E. and Mary J. Armstrong Memorial Fund for Research in Anomalies of the Eye.
ABBREVIATIONS
| MMP | Matrix metalloproteinase |
| NAC | N-acetylcysteine |
| TGF-β1 | Transforming growth factor-β1 |
Footnotes
Ben-Shlomo G, Roecker H. Efficacy evaluation of canine serum and plasma against protease activity for treatment of keratomalacia (abstr). Vet Ophthalmol 2014;17:E36.
Collagenase from Clostridium histolyticum, type XI, Sigma-Aldrich Corp, St Louis, Mo.
EDTA, Sigma-Aldrich Corp, St Louis, Mo.
N-Acetyl-L-cysteine, Sigma-Aldrich Corp, St Louis, Mo.
Tetracycline hydrochloride, Sigma-Aldrich Corp, St Louis, Mo.
Doxycycline hyclate, Sigma-Aldrich Corp, St Louis, Mo.
Minocycline hydrochloride, Sigma-Aldrich Corp, St Louis, Mo.
Whatman filter paper, Sigma-Aldrich Corp, St Louis, Mo.
References
1. Rehman AA, Ahsan H, Khan FH. α-2-macroglobulin: a physiological guardian. J Cell Physiol 2013;228:1665–1675.
2. Borth W. α2-macroglobulin, a multifunctional binding protein with targeting characteristics. FASEB J 1992;6:3345–3353.
3. Haffner JC, Fecteau KA, Eiler E. Inhibition of collagenase breakdown of equine corneas by tetanus antitoxin, equine serum and acetylcysteine. Vet Ophthalmol 2003;6:67–72.
4. Ollivier FJ, Brooks DE, Kallberg ME, et al. Evaluation of various compounds to inhibit activity of matrix metalloproteinases in the tear film of horses with ulcerative keratitis. Am J Vet Res 2003;64:1081–1087.
5. Burns FR, Stack MS, Gray RD, et al. Inhibition of purified collagenase from alkali-burned rabbit corneas. Invest Ophthalmol Vis Sci 1989;30:1569–1575.
6. Conway ED, Stiles J, Townsend WM, et al. Comparison of the in vitro anti-collagenase efficacy of homologous serum and plasma on degradation of corneas of cats, dogs, and horses. Am J Vet Res 2016;77:627–633.
7. Anitua E, Muruzabal F, Tayebba A, et al. Autologous serum and plasma rich in growth factors in ophthalmology: preclinical and clinical studies. Acta Ophthalmol 2015;93:e605–e614.
8. Lopez-Garcia JS, Murube del Castillo J, Garcia Lozano I, et al. Autologous serum and blood derivatives in ophthalmology. Arch Soc Esp Oftalmol 2012;87:376–377.
9. Brion M, Lambs L, Berthon G. Metal ion-tetracycline interactions in biological fluids. Part 5. Formation of zinc complexes with tetracycline and some of its derivatives and assessment of their biological significance. Agents Actions 1985;17:229–242.
10. Baker A, Plummer CE, Szabo NJ, et al. Doxycycline levels in preocular tear film of horses following oral administration. Vet Ophthalmol 2008;11:381–385.
11. Monk CS, Jeong SY, Gibson DJ, et al. The presence of minocycline in the tear film of normal horses following oral administration and its anticollagenase activity. Vet Ophthalmol 2018;21:58–65.
12. Collins SP, Labelle AL, Dirikolu L, et al. Tear film concentrations of doxycycline following oral administration in ophthalmologically normal dogs. J Am Vet Med Assoc 2016;249:508–514.
13. Chandler HL, Gemensky-Metzler AJ, Bras ID, et al. In vivo effects of adjunctive tetracycline treatment on refractory corneal ulcers in dogs. J Am Vet Med Assoc 2010;237:378–386.
14. Berman MB. Collagenase inhibitors: rationale for their use in treating corneal ulceration. Int Ophthalmol Clin 1975;15:49–66.
15. Epstein SP, Ahdoot M, Marcus E, et al. Comparative toxicity of preservatives on immortalized corneal and conjunctival epithelial cells. J Ocul Pharmacol Ther 2009;25:113–119.
16. Furrer P, Mayer JM, Plazonnet B, et al. Ocular tolerance of preservatives on the murine cornea. Eur J Pharm Biopharm 1999;47:105–112.
17. Hook CW, Brown SI, Iwanij W, et al. Characterization and inhibition of corneal collagenase. Invest Ophthalmol 1971;10:496–503.
18. Ramaesh T, Ramaesh K, Riley SC, et al. Effects of N-acetylcysteine on matrix metalloproteinase-9 secretion and cell migration of human corneal epithelial cells. Eye (Lond) 2012;26:1138–1144.
19. Thermes F, Molon-Noblot S, Grove J. Effects of acetylcysteine on rabbit conjunctival and corneal surfaces. A scanning electron microscopy study. Invest Ophthalmol Vis Sci 1991;32:2958–2963.
20. Aldavood SJ, Behyar R, Sarchahi AA, et al. Effect of acetylcysteine on experimental corneal wounds in dogs. Ophthalmic Res 2003;35:319–323.
21. Angleton EL, Van Wart HE. Preparation and reconstitution with divalent metal ions of class I and class II Clostridium histolyticum apocollagenases. Biochemistry 1988;27:7406–7412.
