Dynamics of ascorbate in the aqueous humor and tissues surrounding ocular chambers

The level of ascorbate in the aqueous humor and surrounding intraocular structures in enucleated arterially perfused rabbit eye was investigated. The enucleated eye preparation was shown to be capable of secreting ascorbate from the perfusate into the aqueous humor. Ascorbate in the iris, ciliary body and cornea was released into the aqueous humor when the eye was perfused with ascorbate-free solution. Failure to obtain aqueous flow rates from the decay of ascorbate in the anterior chamber was due to the contribution of ascorbate from these ocular tissues during the perfusion. Histochemically, ascorbate was localized in the pigmented epithelial layer in the valleys between the ciliary processes and the pars plana of the ciliary body and in the iris. In the cornea, distinct localization of ascorbate was observed in the endothelium and basal cell layer of the epithelium.     

Ocular molecules and cells that regulate immune responses in situ

Regulation of T cell-dependent immune responses is mediated in part by bone marrow-derived antigen presenting cells (APC) that (a) process and present antigens which engage the T cell receptor and (b) secrete cytokines that influence the threshold of T cell activation. The anterior chamber of the eye is lined by the corneal endothelium (which rests on a stroma and epithelium that is devoid of class II MHC + APC) and iris/ciliary body (which contain significant numbers of bone marrow-derived cells, one third of which are class II MHC +). When tested in vitro, these potential APCs fail to present antigens in a form that activates T cells. Moreover, iris/ciliary body cells actually suppress activation of T cells exposed to antigens on conventional APC. In addition, aqueous humor under normal circumstances contains factors (one of which is TGF_B) that are potent inhibitors of antigen-driven T cell activation, but spare other aspects of T cell function. Evidence suggests that the bone marrow-derived cells in iris/ciliary body are the source of this factor. Thus, the anterior chamber contains powerful forces that can prevent induction and can suppress expression of T cell mediated immunity. It is proposed that these forces are responsible for immunologic privilege and anterior chamber associated immune deviation, and for suppressing pathologic proliferation and inflammation in the anterior segment of the eye.     

Morphological study of corneal edema after anterior segment ischemia

This experiment was performed to investigate the correlation between morphological changes of the corneal endothelium and clinical corneal edematous opacities after a procedure causing obstruction of bilateral long posterior ciliary arteries of the albino-rabbits. Shortly after the procedure, corneal endothelium became irregular on the anterior-chamber side and there were numerous lysosomal granules in the cytoplasm of corneal endothelial cells. On the third day, endothelial cells became necrotic and collapsed and a part of the cells peeled into the anterior chamber. Fibroblast-like cells and inflammatory cells were piled up in severely damaged areas of the endothelium. On the 11th day, the corneal endothelium developed again as a single layered structure on Descemet's membrane. It's seemed that corneal edema was induced as a result of severe damage corneal endothelium exposed to biochemical mediators such as prostaglandins released into the aqueous humor because the severity of corneal edema indicated some relation to changes in the concentration of prostaglandins in aqueous humor after anterior segmental ischemia.     

The ocular pharmacokinetics of ketanserin and its metabolite, ketanserinol, in albino rabbits.

Ketanserin, a hypotensive drug with 5-HT2 receptor antagonism, when administered by topical infusion of a 0.25% w/v solution by corneal and scleral applications, was found to lower intraocular pressure with four times more activity than its metabolite, ketanserinol. Drug and metabolite were measured periodically in the corneal epithelium, corneal stroma/endothelium, aqueous humor, iris/ciliary body, conjunctiva, sclera and lens during the infusion period (0-120 min) and the postinfusion period (120-240 min) using a fluorometric reversed-phase HPLC assay developed and verified for the research. The infusion results showed that drug entered the eye by both the corneal and scleral routes. Lateral diffusion occurred between the conjunctiva and corneal epithelium. Drug and metabolite were also detected in the untreated fellow eyes, suggesting contralateral systemic redistribution. In vitro metabolism was studied and found to occur in the corneal epithelium, iris/ciliary body and bulbar and palpebral conjunctiva but not in the corneal stroma/endothelium, aqueous humor and sclera. From noncompartmental analysis, zero-order infusion rate constants, first-order absorption constants, mean residence time, volumes of distribution at steady state (Vss) and clearance (Cle) were obtained using equations specific to the topical infusion method. Vss and Cle of aqueous humor (0.972 ml and 13.55 microl/min) were greater than aqueous humor volume (0.311 ml) and turnover rate (4.7 microl/min).     

Collagen XVIII/endostatin shows a ubiquitous distribution in human ocular tissues and endostatin-containing fragments accumulate in ocular fluid samples

Background The endostatin domain of type XVIII collagen (ColXVIII) inhibits neovascularization and regulates cell migration and matrix turnover. This study was designed to demonstrate the protein and gene expression patterns of ColXVIII/endostatin in the human eye and to ascertain whether endostatin is detectable in ocular fluid samples.Methods Twenty human eyes enucleated on account of choroidal melanoma were used for immunohistochemical stainings with antibodies against ColXVIII and endostatin. In situ hybridization was used to localize cells responsible for the production of mRNA for ColXVIII. Tear fluid, aqueous humor, and vitreous gel samples were used for Western immunoblotting to detect endostatin fragments in these samples.Results ColXVIII was immunolocalized to almost all ocular structures, namely the basement membranes (BMs) of the corneal and conjunctival epithelia, Descement’s membrane, the anterior border layer and posterior pigmented epithelium of the iris, the BMs of the pigmented and non-pigmented ciliary epithelia, the internal wall of Schlemm’s canal and trabeculae, the ciliary and iris muscle cells, the BMs of the pigment epithelium of the retina, and the internal limiting membrane. Universal expression was seen in the BMs of vascular endothelial cells, and in fibroblasts located in the conjunctiva, the iris, and the ciliary body. Endostatin showed a corresponding pattern, but additional immunostaining was present in the corneal and conjunctival epithelial cells. Most epithelial and mesenchymal cells expressed the mRNA for ColXVIII. Endostatin-containing fragments varying in size were detected in tear fluid, aqueous humor and vitreous gel samples.Conclusions Practically all structures of the human eye contain ColXVIII/endostatin, emphasizing its possible important structural and functional role in the human eye. Furthermore, ocular fluid samples contain endostatin fragments, which may contribute to the antiangiogenic properties of the eye.     

Detection and identification of endogenous small molecules in ocular tissues by proton nuclear magnetic resonance spectroscopy

Proton nuclear magnetic resonance (nmr) spectra of rabbit ocular tissue homogenates (corneal epithelium, conjunctiva and iris ciliary body) and aqueous humor have been recorded for the first time by incorporation of a spin-spin relaxation reagent and use of the CPMG pulse sequence. A number of endogenous species such as lactate, glucose, alanine and valine have been observed in these ocular homogenates and significant differences have been noted in the distribution of these small molecules in the ocular tissues studied. This technique has been used to study the hydrolysis of acetylcholine by ocular esterases in the iris ciliary body.     

Lens epithelial cells in postoperative aqueous humor

The aqueous humor during or after cataract surgery often contains red blood cells (RBCs) from blood; inflammatory cells from white blood cells (WBCs); fibrin, pigment cells, and fibroblasts from either the iris or the ciliary body; lens epithelial cells (LECs); and corneal endothelial cells. Since LECs may be related to postoperative complications, including a fibrin reaction and a secondary cataract, confirmation of their presence postoperatively not only in the capsular bag, but also in the aqueous humor, is necessary for identifying these complications. Using aqueous humor obtained from pig eyes, I attempted to identify LECs immunohistochemically in the aqueous humor, following phacoemulsification (PEA) and aspiration of the cortex, and found the presence of cells that react to keratin antibody, the marker for epithelial cells. Although these cells could be lens, conjunctival, or corneal epithelial cells, they were found following PEA and aspiration of the cortex, and therefore are believed to be LECs.     

Corneal penetration of rifampin.

We tested the corneal penetration of rifampin in four vehicles: dimethylsulfoxide, polyethylene glycol, an ocular lubricant, and as rifampin ointment. We measured drug concentrations in the aqueous humor in rabbits after topical instillation of 1 and 2.5% rifampin according to two dosage schedules. Drug concentrations in the aqueous humor were bactericidal to Mycobacterium leprae. Since leprosy of the cornea, iris, and ciliary body may develop despite standard systemic bacteriostatic treatment, treatment of leprotic involvement of the anterior eye may be enhanced by intensive topical application of rifampin.     

Localization of myocilin/trabecular meshwork--inducible glucocorticoid response protein in the human eye

PURPOSE: To study distribution and cellular localization of myocilin/trabecular meshwork-inducible glucocorticoid response protein (TIGR) in the human eye. METHODS: A peptide antibody against a portion of the myosin-like domain of myocilin/TIGR was developed. Different ocular tissues from three human donors were investigated by one- and two-dimensional gel electrophoresis and Western blot analysis. Immunohistochemistry was performed on 25 human eyes enucleated because of posterior choroidal melanoma and on 7 normal human donor eyes. RESULTS: By Western blot analysis, a band at approximately 57 kDa was visualized in cornea, trabecular meshwork, lamina cribrosa, optic nerve, retina, iris, ciliary body, and vitreous humor. By immunohistochemistry, immunoreactivity for myocilin/TIGR was observed in cells of the corneal epi- and endothelium and extracellularly in the corneal stroma and sclera. In the trabecular meshwork, cells of the uveal and corneoscleral meshwork were stained, as was the cribriform area directly adjacent to Schlemm's canal. Positive staining was seen in cells of the ciliary epithelium, ciliary muscle, lens epithelium, and in stromal and smooth muscle cells of the iris. Throughout the entire vitreous body, fine filamentous material was positively labeled. In the retina, staining was seen along the outer surface of rods and cones, in neurons of the inner and outer nuclear layer, and in the axons of optic nerve ganglion cells. Optic nerve axons were stained in the prelaminar, laminar, and postlaminar parts of the nerve. In the region of the lamina cribrosa, astrocytes in the glial columns and cribriform plates were positively labeled. CONCLUSIONS: Myocilin TIGR is expressed in almost every ocular tissue. Depending on the respective tissue, it is observed extra- or intracellularly. The presence of myocilin/TIGR in optic nerve axons and lamina cribrosa astrocytes indicates that the trabecular meshwork might not be the only target of abnormal myocilin/TIGR in GLC1A-linked open-angle glaucoma.     

Tissue plasminogen activator in the trabecular endothelium

By quantitative analysis of tissue plasminogen activator (TPA) in the trabecular endothelium, corneal endothelium, and iris in the eyes of monkeys and dogs, we found significant levels of TPA activity. In a [125I]fibrin-coated well assay, the levels for the dog and monkey were, respectively: trabecular endothelium, 0.2 and 0.5; corneal endothelium, 0.8 and 0.5 IU per mg protein. The iris tissue showed high TPA activity, but its protein content could not be measured with the techniques employed. Activity in the aqueous humor was not detectable. By the ELISA technique, the values (in ng TPA/mg tissue protein) for the dog and monkey were, respectively: trabecular endothelium, 0.16 and 0.44; corneal endothelium, 0.48 and 0.92. Again, iris tissue showed high TPA activity, whereas the aqueous humor showed low activity (0.86 ng/ml). The data obtained with the two methods showed a reasonable consistency, although a direct comparison was not possible because two separate standards were used. The presence of TPA in the trabecular endothelium, corneal endothelium, and iris may be important in modulating the resistance to aqueous outflow under normal conditions as well as those of hyphema.     

Localization of nitric oxide synthase isoforms in porcine ocular tissues.

PURPOSE: To determine by immunohistochemistry the distribution of different nitric oxide synthases (NOS) isoforms in the porcine eye. METHODS: By light microscopy (immunofluorescence), porcine ocular tissues were investigated using monoclonal antibodies against neuronal NOS (nNOS; NOS I), endothelial NOS (ecNOS; NOS III), and macrophage NOS (macNOS; NOS II). RESULTS: Specific nNOS immunoreactivity was detected along the apical cytoplasmic portions of the non-pigmented and pigmented ciliary epithelial cells, in the endothelial lining of the corneoscleral meshwork and the uveal cords of the iridocorneal angle tissue, as well as in the photoreceptor layer of the sensory retina. Immunoreactivity for ecNOS was confined to the vascular endothelium of the vessels of the conjunctiva, iris, ciliary body, retina, choroid and optic nerve. A mild immunostaining for macNOS was present in the cytoplasm of some non-pigmented ciliary epithelial cells. CONCLUSIONS: The predominant localization of nNOS in ciliary epithelial and trabecular endothelial cells suggests a possible involvement of nNOS in both the production and outflow of aqueous humor in porcine eyes.     

3-amino-triazole effects on the eye of young and adult rabbits in the presence and absence of hydrogen peroxide

3-aminotriazole (3AT) is known to reduce catalase levels in ocular tissues when given intravenously or orally. Rabbits were given either 4 ml/kg of a 3M solution of 3AT intravenously or a 2% solution as drinking fluid. Intravenous 3AT administration was followed at 4 hrs by an intracameral injection of hydrogen peroxide (H2O2) to give an aqueous humor concentration of 3.2 mM in young (4-6 weeks of age) and a 3.3 mM in adult (6 months of age) rabbits. Tissues were taken for microscopy at either 6 or 24 hours after intracameral H2O2. Neither oral nor intravenous 3AT alone in adult rabbits, or intravenous 3AT in young rabbits, had any effect on either iris, ciliary process, or corneal endothelial morphology. After oral 3AT in adult rabbits, H2O2 caused highly edematous ciliary processes with dilated vessels; corneal endothelial cells were swollen. Previous studies in adult and young rabbits have shown that intracameral H2O2 alone caused few morphological changes in young, but marked changes in the adult that correlated with the 35 to 50% lower catalase levels found in iris and corneal endothelium, respectively, in adult ocular tissues. Young rabbits pre-treated with intravenous 3AT, when examined at 6 and 24 hours after intracameral H2O2, showed swollen ciliary processes, vessel dilation, alteration of the pigment epithelium and corneal endothelial damage. In non 3AT-treated young rabbits, H2O2 caused only minor morphological changes. In adult animals at 6 and 24 hours after intracameral H2O2 the ciliary processes were edematous in the absence of 3AT; after intravenous 3AT and intracameral H2O2 the changes were even more marked, with very severe swelling of ciliary processes and corneal endothelial damage. It is apparent that the decrease in catalase caused by 3AT allows H2O2 to induce damage even in young animals where it usually does not induce morphological changes. In adult animals, the effects of H2O2 are enhanced in the presence of 3AT.     

Herpes simplex virus entry receptor nectin-1 is widely expressed in the murine eye

PURPOSE: Nectin-1 belongs to the immunoglobulin superfamily, mediates cell-cell adhesion in cadherin-based adherens junctions, and acts as a receptor for herpes simplex virus (HSV). The goals of this study were (1) to determine whether nectin-1 is expressed in ocular tissue that is an important target of HSV infections and (2) to determine whether HSV type 1 (HSV-1) infection affects nectin-1 expression in the eye. METHODS: Expression of nectin-1 and HSV-1 protein was determined by immunohistochemical analysis of ocular tissues of untreated BALB/c mice and mice that were euthanized either 7 days or 7 months after corneal inoculation of HSV-1 or sterile tissue-culture medium (mock). RESULTS: In ocular tissues derived from untreated and mock-infected mice, widespread nectin-1 expression was detected among cells of the corneal epithelium and endothelium, conjunctiva, lens epithelium, ciliary body, iris, choroid, and retina. However, fibroblasts in the corneal stroma and the sclera did not express detectable levels of nectin-1. Ocular tissues from mice euthanized 7 days after corneal inoculation of HSV-1 frequently demonstrated corneal ulceration and inflammation and HSV-1 protein expression in the corneal epithelium, stroma, endothelium, conjunctiva, iris, and ciliary body but rarely in the retina. Ocular tissues from mice euthanized 7 months after HSV-1 inoculation demonstrated corneal epithelial and stromal inflammation, but HSV-1 protein expression was not detected. HSV-1 infection did not lead to a loss of nectin-1 expression in any of the tissues examined. In contrast to uninfected corneas, the inflamed and vascularized stroma of infected corneas contained mononuclear inflammatory cells, vascular cells, and fibroblasts that stained positive for nectin-1. CONCLUSIONS: Findings report that nectin-1 is widely expressed in murine ocular tissues. Only fibroblasts in the corneal stroma and sclera of uninfected tissues were devoid of nectin-1 expression. HSV-1-infected inflamed corneas contained some stromal fibroblasts with detectable nectin-1 expression, which potentially could be targeted by the virus. Widespread nectin-1 expression in the eye suggests that this receptor may play a role in the pathogenesis of ocular HSV infections.     

Platelets and polymorphonuclear leukocytes in experimental ocular inflammation in the rabbit eye

Accumulation of 111indium-oxine (111In)-labelled platelets and the kinetics of 111In-labelled polymorphonuclear leukocytes (PMNLs) were studied in the anterior eye during neurogenic inflammation (induced by topical neutral formaldehyde) or after paracentesis in the rabbit, after formaldehyde irritation, 40 times more platelets were found in the aqueous humor 60 min later than in the control eyes and 400 times more after paracentesis. Platelets were found to be increased in the ciliary body, but not in the iris or choroid. Under light microscopy, some of the blood vessels in the ciliary processes were occluded and filled with red blood cells and platelets. The amount of PMNLs increased in the aqueous humor 15-20 h following formaldehyde irritation. After paracentesis, increased amounts of PMNLs were found in the iris and ciliary body 3-6 h later, while in the aqueous humor PMNLs were observed already 2 h. In the present study, increased amounts of 111In-labelled platelets and PMNLs were demonstrated in the anterior eye during experimental ocular inflammation. This method provides a useful tool for evaluating the accumulation of cells or the cell kinetics in ocular tissues during experimental inflammation.     

Intraocular production and release of nerve growth factor after iridectomy

PURPOSE: To determine the presence of nerve growth factor (NGF), NGF mRNA, and NGF receptor (TrkA) in rabbit ocular tissues, and whether changes occur in NGF and NGF mRNA levels after experimental iridectomy. METHODS: Immunohistochemistry for NGF and TrkA and in situ hybridization for NGF mRNA were performed on rabbit cornea, iris, ciliary body, and lens in the basal state. Quantification of NGF mRNA and NGF protein levels in these tissues was performed by RT-PCR and immunoenzymatic assay, respectively. A time course of NGF concentration in the aqueous humor and the expression of NGF mRNA in iris and ciliary body were performed after the iridectomy and were compared with levels in a sham-treated group (paracentesis). RESULTS: Cornea, iris, ciliary body, and lens expressed NGF mRNA, NGF protein, and TrkA in the basal state. The highest levels of NGF were detected in the iris (8938.0 +/- 3968.1 pg/g), and the lowest were in the aqueous humor (22.8 +/- 9.7 pg/mL). Experimental iridectomy induced a transient increase of NGF concentration in the aqueous humor that reached its peak 4 hours after the experimental injury (464.4 +/- 29.9 pg/mL versus the control group 101.6 +/- 18.8 pg/mL; P < 0.001) and returned to baseline value after 7 days. A significant increase of NGF mRNA was also observed 1 hour and 4 hours after the iridectomy in the iris (1 hour, 788 +/- 85 OD; 4 hours, 760 +/- 81 OD versus baseline, 246 +/- 32 OD; P < 0.0001) and ciliary body (1 hour, 330 +/- 19 OD; 4 hours, 453 +/- 52 OD versus baseline, 219 +/- 37 OD; P < 0.05), but not in the cornea, lens, or any tissues from the control group. CONCLUSIONS: NGF is present and produced in the anterior segment of the eye and is released in the aqueous humor in the basal state. Experimental iridectomy induces increased production of NGF in the iris and in the ciliary body and an increased concentration of NGF in the aqueous humor.     

Ocular disposition of aminozolamide in the rabbit eye

We have previously determined that 6-amino-2-benzothiazolesulfonamide (aminozolamide) significantly lowers IOP in rabbits and, more importantly, in ocular hypertensive human subjects. Results from in vitro experiments established that both the inhibitory activity of aminozolamide against carbonic anhydrase B as well as the penetration rate across excised rabbit corneas were equivalent to ethoxzolamide. Consequently, we have investigated the ocular disposition of aminozolamide to explain its activity when instilled topically to the eye. A constant concentration of 67.4 micrograms/ml of drug was applied for 90 min to the left eye of anesthetized rabbits. Drug and metabolite were measured in both aqueous humor and iris/ciliary body over time. The metabolite was collected and purified. Identification using mass spectroscopy, high pressure liquid chromatography (HPLC) and fluorescence scans indicated that the metabolite was 6-acetamido-2-benzothiazolesulfonamide. Relatively high levels of metabolite were identified in the cornea and iris/ciliary body but were much lower in aqueous humor. Tissue concentrations over time for the metabolite in iris/ciliary body were approximately 2-fold higher than levels of metabolite measured in aqueous humor. When compared to drug levels measured in either aqueous humor or iris/ciliary body, metabolite levels in these respective tissues were much higher. It is hypothesized that topical activity is a consequence of both metabolite retention in the iris/ciliary body as well as inhibition of 99+% of carbonic anhydrase. Both of these events must occur over a sufficient time period to effect a significant lowering of IOP.     

Effects of calcium chelating agents on corneal permeability

Corneal penetration studies have been conducted in unanesthetized albino rabbits using various organic compounds representing both polar and nonpolar species. In the presence of calcium chelating agents, polar compounds generally demonstrate an increase in corneal penetration. Evidence that this corneal effect is reversible is presented. Concomitant with an increase in both corneal and aqueous humor drug levels was a decrease in drug concentration in both iris and conjunctival tissues tentatively attributed to chelation effects on vascular permeability of these tissues. EDTA, a known calcium chelator, was shown to penetrate the cornea, conjunctiva, and iris/ciliary body from a topically applied dose. The implications of this observation pertain to both toxicity effects, when EDTA is incorporated into ocular drug products for stability purposes, and novel strategems for improving ocular bioavailability of topically applied drugs.     

Improved ocular: systemic absorption ratio of timolol by viscous vehicle and phenylephrine

Increasing the ocular absorption of timolol relative to its systemic absorption is important clinically because ophthalmic timolol may cause serious respiratory, cardiac, and central nervous system side effects. The authors evaluated the effects of phenylephrine coadministration and solution viscosity on the aqueous humor:plasma and iris ciliary body:plasma ratios of peak timolol concentrations after ocular application. Timolol eye drops (5 mg/ml, 25 microliters) were administered to the eyes of pigmented rabbits. Coadministered phenylephrine (0.8-8.2 mg/ml) decreased the systemic peak concentrations of timolol significantly. Since ocular absorption of timolol was not affected by phenylephrine, the ocular:systemic concentration ratios were improved four- to fivefold. Phenylephrine slows down the systemic absorption of timolol by constricting the conjunctival and nasal capillaries. The ratios of the aqueous humor:plasma and iris ciliary body:plasma peak concentration of timolol were improved three- to ninefold in the presence of sodium carboxymethylcellulose compared with nonviscous eye drops. The improved ocular penetration is probably due to the longer corneal contact, and the decreased rate of systemic absorption may be caused by the slower spreading of the solution on the nasal mucosa. Compared with timolol eye drops, the ratio of the eye:plasma peak timolol concentrations was improved tenfold by using viscous eye drops with phenylephrine. Systemic concentrations of ophthalmic timolol and possibly related side effects can be decreased when timolol is instilled in a viscous vehicle with a low phenylephrine concentration.     

The disposition and bioavailability of 35S-GSH from 35S-GSSG in BSS PLUS in rabbit ocular tissues.

The purpose of this study was to evaluate the biodistribution and uptake of 35S-GSH into intraocular tissues following the administration of BSS PLUS containing 35S-GSSG by either an anterior chamber or intravitreal injection. This study evaluated the disposition and uptake of the 35S-radiolabel, the intracellular concentrations of 35S-GSH from extracellular 35S-GSSG, and the percentage of 35S-GSH to the total cellular GSH pool. Glutathione was analyzed by high-performance liquid chromatography (HPLC) using fluorescence detection after derivitizing the thiols in situ with monobromobimane. The effluent from the GSH peak was then collected for measurement of 35S-GSH. After an anterior chamber injection of 35S-BSS PLUS, 35S-radioactivity rapidly disappeared from the aqueous humor between 0.5 and 2 hours; corneal 35S-radioactivity remained constant over time. 35S-GSH was detected in the iris and ciliary body. However, in the cornea, 35S-GSH became the predominant radioactive thiol in the stroma, endothelium, and epithelium; the corneal stroma appeared to be a possible GSH reservoir for the adjacent corneal layers. After an intravitreal injection, 35S-radioactivity slowly decreased in the vitreous humor but was readily taken up by the tissues of the posterior segment, especially the retina and choroid, which showed the greatest concentrations of 35S-GSH of all tissues studied. The data from this study demonstrate that 35S-GSSG in BSS PLUS is metabolized and taken up by ocular cells and that 35S-GSH becomes incorporated into the intracellular GSH pool of ocular tissues.