Aqueous Outflow–Enhancing Effect of tert-Butylhydroquinone: Involvement of AP-1 Activation and MMP-3 Expression

Abstract

purpose. To test the effect of stimulators of activator protein (AP)-1, on expression of stromelysin (MMP-3) in human TM cells and on aqueous outflow in perfused human anterior segments.

methods. Change in MMP-3 expression was determined by immunoassay of proMMP-3 levels in the media of cultured human TM cells. Anterior segments of human donor eyes with or without glaucoma were perfused with vehicle or the AP-1 stimulator tert-butylhydroquinone (tBHQ). The outflow rates or intraocular pressure (IOP), and proMMP-3 levels in the perfusate were monitored.

results. AP-1 stimulators, such as β-naphthoflavone, 3-methylcholanthrene, and tBHQ, significantly upregulated (2–4-fold) TM cell expression of MMP-3. The stimulatory effect of tBHQ was concentration dependent, with an EC₅₀ of approximately 3 μM, and was blocked by concomitant treatment with 100 nM SR11302, which sequesters AP-1. When nonglaucomatous human eyes were perfused with tBHQ (10 μM), both outflow rates and perfusate proMMP-3 level increased significantly within the first 24 hours. The outflow effect of tBHQ was suppressed when SR11302 (100 nM) was added in the perfusate. tBHQ also lowered the IOP by more than 40% in perfused glaucomatous eyes.

conclusions. An AP-1 activator, tBHQ, upregulated expression of MMP-3 in cultured human TM cells and perfused human eyes and enhanced outflow ex vivo. These effects were blocked by sequestering AP-1, suggesting that activation of AP-1 can lead to increased MMP-3 production in the TM, which in turn improves outflow facility. This unique mechanism may provide a novel therapy for glaucoma.

Matrix metalloproteinases (MMPs) enhance aqueous outflow facility in perfused human eye organ culture. ¹ This effect is probably mediated by the enzymes’ catalytic activity in the degradation of extracellular matrix. Extracellular matrix, when excessively accumulated in the trabecular meshwork (TM), can physically reduce the extracellular space, hinder the aqueous outflow, increase intraocular pressure (IOP), and contribute to the development of primary open-angle glaucoma. ^{2 3 4 5 6 7 8} Among the various MMPs, stromelysin (MMP-3) by itself is sufficient to produce the outflow-enhancing effect when perfused ex vivo into eyes. ¹ It is thus expected that other treatments that upregulate MMP-3 or augment its activity in the relevant structures of the eye should also improve aqueous outflow. Indeed, interleukin (IL)-1α, a proinflammatory cytokine that stimulates the production of MMP-3 in human TM tissue ^{9 10} and cultured human TM cells, ^{9 10 11} was demonstrated to increase aqueous outflow facility in the perfused human eye ¹ and lower IOP when injected into the anterior chamber of rats. ¹² Correspondingly, MMP inhibitors reduce outflow of aqueous humor. ¹ Furthermore, the ocular hypotensive effects of laser trabeculoplasty and prostaglandin FP receptor agonists appear to be mediated at least partly by the activation of MMPs. ^{13 14 15 16 17}  

Although MMPs and their cytokine stimulators are efficacious outflow enhancers, they are not practical as therapeutic agents for the clinical management of glaucoma. IL-1α is a proinflammatory molecule that can cause various undesirable side effects when administered locally in the eye. IL-1α activates many cellular signaling pathways in tissues, such as the nuclear factor (NF)-κB, phospholipase A₂, and activator protein (AP)-1 pathways. ^{18 19 20} Some of these pathways may be responsible in the upregulation of MMP-3 expression, whereas others may contribute to its undesirable effects. Using selective enzyme inhibitors as pharmacological tools, we recently found that the IL-1α–stimulated production of MMP-3 in cultured human TM cells requires the functional presence of the AP-1 pathway and does not depend on the activation of NFκB or phospholipase A₂ pathway. ²¹  

Certain compounds have been suggested to stimulate the AP-1 pathway in various cells and tissues. Hence, we tested the effects of these compounds on the production of MMP-3 in cultured human TM cells. One of the compounds that stimulated MMP-3 production in the TM cells, tert-butylhydroquinone (tBHQ), was further tested for its effect on aqueous outflow in the perfusion human eye organ culture. 

Methods

Human TM Cell Culture

Human TM cells were isolated, characterized, and cultured as described previously. ^{22 23} In this study, five human TM cell lines were used: TM16A, TM35D, TM75C, TM79, and TM332/344. The TM cells were grown and allowed to reach confluence in 24-well plates at 37°C and 5% CO₂ in a solution containing Dulbecco’s modified Eagle’s medium with stabilized l-glutamine (Glutamax I; Invitrogen/Gibco, Grand Island, NY) supplemented with 10% fetal bovine serum (Hyclone, Logan, UT) and 50 μg/mL gentamicin (Invitrogen/Gibco). Cells were serum-deprived for 24 hours before treatment with test agents. On the day of study, cells were rinsed with serum-free medium and incubated with the appropriate test compound(s) in 0.3 mL of serum-free medium for the indicated amount of time. 

Enzyme-Linked Immunosorbent Assay

Production of MMP-3 by the cells was assessed by measuring the levels of MMP-3 proenzyme in the cell media, as described previously. ¹¹ Recommended procedures of a commercially available ELISA kit (Bindazyme; The Binding Site, Birmingham, UK) were followed. In this assay, the detectable limit for proMMP-3 was 0.3 ng/mL. 

Human Eye Perfusion Organ Culture

Human ocular perfusion organ culture was performed as described. ^{24 25 26} Results in 12 pairs of nonglaucomatous donor eyes and 7 donor eyes with primary open-angled glaucoma are reported in this study. The eyes were obtained and used according to the provisions of the Declaration of Helsinki for research involving human tissue. None of the patients was known to have other ocular diseases. The donor eyes, 16 to 20 hours after death, were dissected at the equator and the iris, lens, most of the ciliary body, and vitreous were removed. The anterior segment of the eye, including cornea and scleral ring containing the TM, was placed into a custom-made Plexiglas culture dish and sealed in place with a Plexiglas O-ring. Dulbecco’s modified Eagle’s medium (Invitrogen/Gibco) was perfused through a central cannula in the bottom of the dish. 

The anterior segment, cultured at 37°C and 5% CO₂, was allowed to equilibrate for 2 to 4 days before the study began. Tissues that did not reach a stable baseline IOP or flow rate were discarded (approximately 50% of the eyes). Acceptable tissues were perfused with media containing the indicated compound(s) for the indicated period. A fresh solution of each test agent was prepared daily. 

We used both constant perfusion pressure and constant flow rate methods to evaluate the outflow effect of tBHQ. In the constant pressure method, the media reservoir was raised to produce a hydrostatic pressure of 12 mm Hg, which was confirmed by constant monitoring with a sensitive pressure transducer (custom manufactured by the Department of Bioengineering, University of Texas Southwestern Medical Center, Dallas, TX). Flow rate of the perfusion medium was calculated by once daily weighing of the reservoir. In the constant flow method, the anterior segment was perfused at a flow rate of 2 μL/min using a perfusion pump (Harvard Apparatus, South Natick, MA). The IOP was monitored by a second cannula attached to a pressure transducer. The IOP was recorded every 5 minutes, and hourly averages were calculated. 

In specified experiments, perfusates were collected for the evaluation of proMMP-3 levels. 

Morphologic Evaluation of Perfused Eyes

At the end of each perfusion organ culture study, the tissues were perfusion fixed at 15 mm Hg constant pressure, dissected into four quadrants, and processed for light microscopy and transmission electron microscopy, as previously described. ^{25 26} The viability of the outflow pathway tissue, especially the TM, was evaluated in a masked fashion. Studies were regarded as invalid and data discarded if more than one quadrant per eye had unacceptable morphologic findings, such as excessive TM cell loss, denudation of trabecular beams, an excess in cellular debris in the TM region, loss of Schlemm’s canal endothelial cells, or breaks in the Schlemm’s canal inner wall lining. Based on these criteria, approximately 30% of the perfused tissues were rejected. Furthermore, the TM and the corneal endothelium of selected anterior segments were carefully examined to determine whether the stimulatory effect of tBHQ on expression of MMP-3 affects the ultrastructure of these tissues. 

Test Compounds

β-Naphthoflavone, 3-methylcholanthrene, and tBHQ were obtained from Sigma-Aldrich (St. Louis, MO). SR11302, a selective AP-1 inhibitor retinoid, ²⁷ was supplied by Mark Hellberg (Alcon, Fort Worth, TX). 

Statistical Analyses

Results were evaluated by two-tailed Student’s t-test or one-way ANOVA, followed by the Dunnett test versus the vehicle control, and were considered statistically significant if P < 0.05. 

Results

Effect of AP-1 Activators on ProMMP-3 Expression in Cultured TM Cells

Selected compounds reported to activate AP-1, such as β-naphthoflavone, 3-methylcholanthrene, and tBHQ, ²⁸ were evaluated for their effects on MMP-3 expression in one of the cultured human TM cell lines, TM35D. All three compounds (10 μM) were efficacious in stimulating MMP-3 expression (P < 0.05; Fig. 1 ). For example, β-naphthoflavone increased the proMMP-3 levels to 241% ± 28% (mean ± SEM, n = 14), 3-methylcholanthrene raised the level to 271% ± 21% (n = 13), and tBHQ increased the proMMP-3 levels to 211% ± 17% (n = 33). Because of the reported potential toxic effects of β-naphthoflavone ^{29 30} and 3-methylcholanthrene, ^{31 32} we chose to study further the biological significance of the stimulatory effect of tBHQ on expression of MMP-3 by the TM cells. 

In the TM35D cells, the effect of stimulation with tBHQ on expression of MMP-3 was concentration dependent (Fig. 2A) , with an EC₅₀ of approximately 3 μM. It did not have any detectable effect at 1 μM and produced maximum stimulation at 10 μM and 100 μM. tBHQ stimulated the expression of proMMP-3 to between 200% and 400% of the basal level. This stimulatory effect depended on the functional presence of the AP-1 signaling pathway, because SR11302 (100 nM), an AP-1 inhibitor, completely abolished the effect of tBHQ (Fig. 2B) . Under the tested conditions, both tBHQ and SR11302 were not toxic to the cells, as assayed by neutral red cell survival assay and morphologic observations (data not shown). 

tBHQ increased the expression of MMP-3 in all cultured human TM cell lines tested. As shown in Table 1 , it raised the proMMP-3 levels in all five cell lines, with the increase reaching statistical significance in four. Notably, the stimulatory effect of tBHQ was obviously lower than that of IL-1α, which generally produced 10-fold or higher increases in proMMP-3 levels. The effect of tBHQ ranged from 124% to 227% of that of control levels (Table 1) . 

Effects of tBHQ on Perfused Human Eye Organ Culture

MMP-3 has been shown to enhance aqueous humor outflow in perfusion organ culture, ¹ and compounds such as IL-1α that upregulate MMP-3 in the TM also increase outflow facility. ¹ We tested whether the increased expression of TM proMMP-3 induced by tBHQ correlates with an increase in outflow rate, by using the perfusion organ culture. Initially, we tested the effects of tBHQ in nonglaucomatous eyes, using the constant-pressure perfusion method. We subsequently determined that the tissue viability of glaucomatous eyes perfused using this method was invariably unacceptable. Presumably, the pathologically high resistance in aqueous outflow through the TM in these eyes limited the flow rate and sufficient supply of nutrients to sustain tissue health during the study period. Hence, for glaucomatous tissues, the constant-flow method was used. 

Figure 3A demonstrates that when the anterior segments of nonglaucomatous eyes were continuously perfused with 10 μM tBHQ, the aqueous outflow rate was significantly augmented in comparison with the vehicle-treated tissues. The flow rate at time 0 for the vehicle-treated group was 0.295 ± 0.083 μL/min · mm Hg (mean ± SEM, n = 6) and that of the tBHQ-treated group was 0.339 ± 0.052 μL/min · mm Hg (difference not statistically significant). The vehicle-treated eyes had a slight but statistically insignificant (P > 0.05) decline in outflow rate over the perfusion period. In contrast, tBHQ induced a trend in increase in outflow rate. The mean aqueous outflow rate in tBHQ-treated eyes was significantly increased (P < 0.05) from the vehicle-treated eyes at 24, 72, and 96 hours after treatment. At 96 hours, the difference in outflow rates between the tBHQ- and vehicle-treated group reached approximately 50%. 

When perfusates were assayed for production of proMMP-3, the increase in outflow induced by 10 μM tBHQ appeared to correlate with enhanced proMMP-3 levels in the perfusate. The daily production of proMMP-3 by the perfused anterior segments increased gradually from 212.5 ± 32.4 (mean ± SEM, n = 4) to 300.3 ± 27.7 ng/d within the first 48 hours of 10 μM tBHQ treatment (Fig. 3B) , followed by an elevated steady state lasting at least until 96 hours. In contrast, the production of proMMP-3 in vehicle-treated tissues was relatively unchanged. ProMMP-3 levels at the 24-, 48-, and 72-hour time points were significantly different between the two groups (P < 0.05). 

Based on our findings in cultured TM cells, ²¹ we hypothesized that the increase in MMP-3 production and outflow-enhancing effects of tBHQ in the perfusion cultured eyes required activation of the AP-1 signaling pathway. Indeed, concurrent perfusion of the anterior segments with tBHQ (10 μM) and the AP-1 inhibitor SR11302 (100 nM) prevented the tBHQ-induced increase in outflow facility (Fig. 4) . The preventive effect was apparent at 48 hours after the initiation of treatment and became statistically significant at 72 and 96 hours (P < 0.05). 

The effect of tBHQ on aqueous outflow was evident not only in nonglaucomatous donor eyes, but also in glaucomatous eyes. Data from seven eyes of four donors with glaucoma are reported herein. In each pair, one eye was perfused with tBHQ, and the contralateral eye was treated with vehicle. One of the vehicle-treated eyes did not pass our postperfusion viability criteria, and hence its data were not reported. Donors had died at 87 to 91 years of age, all of cardiovascular diseases. Two were white females and two were white males. There was no information regarding their medical treatment for glaucoma. Figure 5 demonstrates that perfusion of glaucomatous eyes with 10 μM tBHQ lowered IOP. The baseline IOP of the vehicle- and tBHQ-treated groups were 21.2 ± 0.7 (mean ± SEM, n = 3) and 19.8 ± 1.6 mm Hg (n = 4), respectively. Continuous perfusion with tBHQ lowered the IOP as early as 24 hours after the start of treatment and produced significant hypotension after 60 hours (P < 0.05). At 80 hours, the IOP in the tBHQ-treated group was lowered to 46% ± 17% of the baseline pressure. 

Perfusion with tBHQ did not seem to have any apparent effects on the ultrastructural morphology of the eyes. As shown in Figure 6 , the TM of tBHQ-treated anterior segments of nonglaucomatous eyes did not differ significantly from those of vehicle-treated ones. Cellularity of the TM, morphology of the trabecular beams, endothelium of Schlemm’s canal, and patency of the trabecular space all appeared to be normal. In some samples, a trend of less extracellular debris was observed in the tBHQ-perfused eyes. However, morphometric measurements with a larger sample size or longer perfusion time would be needed to demonstrate conclusively any effect of tBHQ on the extracellular matrix. 

Similarly, tBHQ perfusion did not affect the morphology of corneal endothelium and stroma, even though proMMP-3 levels in the perfusate were increased. The corneal endothelial cells of tBHQ-treated anterior segments appeared morphologically normal, with healthy looking plasma membrane, cytoplasm, and intracellular organelles. Histologically, the cell–cell and endothelium-stroma junctions of the tBHQ-treated eyes were similar to those of vehicle-treated tissue (Fig. 7) . This agent did not produce any observable changes in the corneal stroma (data not shown). 

Discussion

In this report, we demonstrated that tBHQ, β-naphthoflavone, and 3-methylcholanthrene, compounds that are structurally diverse but share a common biological effect in the activation of the AP-1 pathway, ²⁸ stimulated the expression of MMP-3 in the cultured human TM cells. We also showed that the stimulatory effect of tBHQ was concentration dependent and was observed in all cultured human TM cell lines tested. This effect is probably a result of AP-1 activation, because it could be completely blocked by pretreatment with the AP-1 inactivator SR11302. The effective inhibitory concentration of SR11302 was similar to its effective concentration for inhibition of the AP-1 activity. ²⁷ When used to perfuse the anterior segments of nonglaucomatous eyes, tBHQ significantly increased the aqueous outflow rate at a concentration (10 μM) shown to increase the production of proMMP-3 by the TM cell cultures. The change was detectable at 24 hours after initiation of tBHQ treatment, and the difference between the treated and control outflow rates continued to widen in subsequent time points. We hypothesize that an increase in the trabecular outflow is the most likely mechanism of the outflow effect of tBHQ. However, the current evidence cannot completely exclude other possible mechanisms, such as a change in transscleral flow. 

In addition to the activation of AP-1 and stimulation of MMP-3 expression, tBHQ has other biological effects. Most notably, it is a safe and popular antioxidant used as preservative in many food products. ³³ Related to its antioxidant action, tBHQ can activate expression of various genes that are regulated by the antioxidant-responsive element—for example, NADPH-quinone oxidoreductase. ³⁴ t-BHQ also interferes with intracellular calcium homeostasis. It inhibits the calcium pumps in the endoplasmic reticulum, depleting intracellular calcium stores ³⁵ and blocking calcium influx through the L-type calcium channel. ³⁶ Although we cannot completely exclude the contributions of these various cellular actions of tBHQ, its outflow effect is probably mediated by the AP-1 pathway and subsequent MMP-3 expression. The tBHQ-induced increase in aqueous outflow rate correlated with an upregulation of MMP-3 production in the perfused eyes. More important, this outflow effect was eliminated by the AP-1 inactivator SR11302, which does not affect oxidative functions or intracellular calcium levels. These results demonstrate that activation of AP-1 and expression of MMP-3 are essential for tBHQ to affect aqueous outflow. 

During the 4-day perfusion period, tBHQ was not toxic to the ocular tissues. The morphology of the TM region in the perfused anterior segments was normal and not appreciably different from that of the vehicle control. Although we noticed a trend of less extracellular debris in some of the tBHQ-treated samples, this observation could not be evaluated statistically because of small sample sizes and the inherent biological variability present in the older eyes used in this study. If this preliminary result is confirmed in future studies, it would be intriguing to speculate that the decrease in outflow resistance is a consequence of the reduction of trabecular debris due to the upregulation of MMP-3. 

Similarly, perfusion of human eyes with tBHQ did not affect the morphology of the corneal stroma and endothelium. We could not evaluate the corneal epithelium of these samples, because the epithelium generally was compromised by the inherent condition of the donor eyes. tBHQ did not appear to damage the corneal endothelium, even though this tissue was directly exposed to the drug solution. The activation of certain MMPs is implicated in corneal ulceration ³⁷ and is also involved in corneal wound healing and remodeling. ^{38 39} The apparent lack of corneal tBHQ toxicity suggests that the corneal cells are different from the TM cells in their responses to tBHQ. tBHQ may not activate AP-1, or AP-1 activation may not lead to upregulation of MMP-3 expression in the corneal endothelium or stroma. It also is possible that corneal ulceration, wound healing, and remodeling involve MMPs other than MMP-3. Alternatively, the 4-day treatment period may not be sufficient to produce detectable morphologic changes in the cornea. Clarification of these several hypotheses awaits future studies. 

The outflow effect of tBHQ was also observable in glaucomatous eyes. Similar to its effect in nonglaucomatous eyes, the aqueous outflow enhancement as reflected in a reduction in IOP was present as early as 24 hours after the initiation of treatment with 10 μM tBHQ and continued to produce a significant reduction IOP at later times. Even though the glaucomatous TM tissue was defective and contained fewer TM cells, ⁴⁰ the effectiveness of tBHQ in these glaucomatous eyes indicates that there were sufficient remaining TM cells capable of responding to tBHQ stimulation to produce significant effects on IOP. This implies that compounds with pharmacological action similar to tBHQ may be useful in lowering IOP in patients with glaucoma. 

The demonstration that small molecules, such as tBHQ, can upregulate MMP expression in the TM cells and increase aqueous outflow facility in glaucoma donor eyes has significant clinical implications. Human perfusion organ culture results demonstrated that selected compounds of this pharmacological class improved trabecular (conventional) outflow. This new mechanism of action distinguishes itself from most of the current glaucoma medications, which suppress aqueous production. These new compounds are likely to have an additive effect with current medications, including compounds that increase uveoscleral outflow, such as prostanoids. tBHQ probably will not become an ocular hypotensive medication because it is not a stable compound and has a very short shelf life. ⁴¹ Its aqueous solution was somewhat irritating to the rabbit cornea (unpublished observation) and therefore probably not a suitable therapeutic agent for topical ocular administration. However, other inducers of MMP-3 expression, especially small molecules that readily cross the cornea, may become interesting and novel pharmacological agents for the management of ocular hypertension and glaucoma. 

 Supported by Alcon Research, Ltd.

 Submitted for publication July 26, 2002; revised January 30, 2003; accepted February 13, 2003.

 Disclosure: I.-H. Pang, Alcon Research, Ltd. (E, P); D.L. Fleenor, Alcon Research, Ltd. (E, P); P.E. Hellberg, Alcon Research, Ltd. (E); K. Stropki, Alcon Research, Ltd. (E); M.D. McCartney, Alcon Research, Ltd. (E); A.F. Clark, Alcon Research, Ltd. (E, P)

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Effect of β-naphthoflavone, 3-methylcholanthrene, and tBHQ on proMMP-3 production by cultured human TM35D cells. Cells were incubated with the indicated compounds for 24 hours and the cell media assayed for proMMP-3 content. Mean level of vehicle control of each experiment defined 100%. Bars, mean ± SEM; n = 13–33. *P < 0.05 versus vehicle group.

Figure 2.

 

Effect of tBHQ on proMMP-3 production in cultured human TM (TM35D) cells. (A) Concentration-response curve of tBHQ on proMMP-3 production. (B) Blockade of the tBHQ effect by the AP-1 inactivator SR11302. Cells were incubated with the tBHQ with or without SR11302 for 24 hours and the cell media assayed for proMMP-3 content. Vehicle control defined 100%. Data are expressed as the mean ± SEM; n = 6. *P < 0.05 versus vehicle group.

Table 1.

 

Effect of tBHQ on ProMMP-3 Production in Various Cultured Human TM Cell Lines

Cell Line	ProMMP-3
		Vehicle	IL-1α (5 ng/mL)	tBHQ (10 μM)
TM35D (n = 30)	100 ± 3	1706 ± 122*	211 ± 17*
TM16A (n = 14)	100 ± 6	2944 ± 356*	227 ± 26*
TM75C (n = 12)	100 ± 5	2578 ± 379*	124 ± 11
TM79 (n = 12)	100 ± 1	983 ± 130*	169 ± 10*
TM332/344 (n = 12)	100 ± 4	1348 ± 110*	204 ± 7*

Figure 3.

 

Effect of tBHQ on aqueous outflow rate and proMMP-3 production in human ocular organ culture perfused at constant pressure. Tissues were from nonglaucomatous eyes. (A) Effect of tBHQ (10 μM) on outflow rate. Flow rate at time 0 of each eye defined 100%. (B) Effect of tBHQ (10 μM) on proMMP-3 content in perfusates. Perfusate samples were assayed for proMMP-3 contents. Data are expressed as the mean ± SEM; n = 4 to 6. *P < 0.05 versus the vehicle group.

Figure 4.

 

Blockade of the effect of tBHQ by the AP-1 inactivator SR11302. Human ocular anterior segments were perfused with tBHQ (10 μM) or tBHQ with SR11302 (100 nM). Data are expressed as the mean ± SEM; n = 6. *P < 0.05 versus the vehicle group at the same time point.

Figure 5.

 

Effect of tBHQ (10 μM) on IOP of perfused anterior segments from glaucomatous eyes. Tissues were perfused with a constant flow rate of 2 μL/min and the IOP recorded. IOP at time 0 defined 100%. Data are expressed as the mean ± SEM; n = 3 to 4. *IOPs between the two groups were significantly different at time points from 61 to 80 hours (P < 0.05).

Figure 6.

 

Effect of tBHQ perfusion on ultrastructure of TM. After 4 days of perfusion with 10 μM tBHQ or vehicle, all perfused tissues were fixed and observed with both light and electron microscopy. The TM appeared normal in the drug-treated tissue (A) and not different from that of the vehicle-treated samples (B). Six independent studies were performed showing similar results. Bar, 10 μm.

Figure 7.

 

Lack of effect of tBHQ perfusion on morphology of corneal endothelial cells in perfused anterior segments. After 4 days of perfusion with vehicle (A) or 10 μM tBHQ (B), the tissues were fixed and observed with electron microscopy. The corneal endothelium appeared normal in the drug-treated tissue and not different from that of the vehicle-treated samples. Four independent studies were performed showing similar results. Bar, 1.5 μm.

The authors thank the donors of ocular tissues used in this study, without whose selfless generosity advances in ophthalmological research would be significantly hindered; and Paula Billman and the Central Florida Lions Eye and Tissue Bank for the procurement of donor tissues.