|Molecular Vision 2003;
Received 19 March 2003 | Accepted 28 October 2003 | Published 3 November 2003
Epithelial activity of hexokinase and glucose-6-phosphate dehydrogenase in cultured bovine lenses recovering from pharmaceutical-induced optical damage
Andrew T. E. Hartwick,
Jacob G. Sivak
School of Optometry, University of Waterloo, Waterloo, Ontario, Canada
Correspondence to: Jacob G. Sivak, School of Optometry, University of Waterloo, 200 University Avenue West, Waterloo, Ontario, Canada N2L 3G1; Phone: (519) 888-4567, ext 2233; FAX: (519) 725-0784; email: email@example.com
Purpose: In a previous toxicological study, cultured bovine lenses exposed to three topical anesthetics displayed distinct patterns of optical damage and recovery. This work investigated the epithelial activity of the metabolic enzymes hexokinase (HK) and glucose-6-phosphate dehydrogenase (G6PD) in lenses recovering from anesthetic-induced damage.
Methods: Cultured bovine lenses were exposed to the anesthetics Alcaine®, Fluress® and Fluoracaine® for 2 h. An automated laser scanner was used to determine the focal length variability (FLV) of the lenses at time-points up to 24 h following their return to fresh culture medium. The epithelial enzyme activities for HK and G6PD were then assayed at the 24 h time-point.
Results: Lenses exposed to Alcaine® displayed an abrupt increase in FLV, while Fluoracaine® treated lenses exhibited optical damage at a slower rate. The FLV in these two groups recovered to near-control levels after 24 h. Fluress® treated lenses did not differ in FLV from controls at any time. The activities of both HK and G6PD were significantly reduced in epithelial samples from each of the three anesthetic treatment groups, relative to controls.
Conclusions: These results show that lens optical quality can recover despite a severe reduction in epithelial HK and G6PD activity, indicating that the optical function of the lens may not be directly related to epithelial metabolic activity. The ScanTox In Vitro Assay System provides an objective measure of lens optical quality, enabling a direct comparison of optical damage and recovery to lens biochemical changes.
The ocular lens, as an avascular tissue, can be cultured as a whole organ with its optical function intact. A method has been developed that is capable of quantifying the optical quality of the bovine lens in vitro using an automated laser scanner . This technique has been used as a toxicological tool to assess the relative toxicity of a number of chemical compounds by determining their effect on the ability of the cultured lens to maintain a well-defined focal point (for a review, see ). A toxicological study on three corneal anesthetics (Alcaine®, Fluress®, and Fluoracaine®) found that each of these anesthetics had a distinctly different effect on lens optical function . Both Alcaine® and Fluoracaine® disrupted lens optical function significantly, while lenses exposed to Fluress® were not significantly different from untreated control lenses. By the end of this 48 h study, all of the lenses had returned to control levels, indicating that the recovery mechanisms within the lens were intact and functioning .
The biochemical basis for the regulation of lens transparency is still not fully understood. A decrease in lens metabolic activity could result in the deterioration of the structural integrity of the lens and a degradation of its optical quality. The lens generates about 70% of its energy through anaerobic glycolysis, and the enzyme hexokinase (HK) catalyzes the phosphorylation of glucose to glucose-6-phosphate in the first rate-limiting step of this pathway [4,5]. The sulfhydryl (-SH) group present on hexokinase has been shown to be sensitive to oxidative modification, resulting in the inactivation of the enzyme [6,7]. Due to the importance of hexokinase in lens glycolysis, the activity of this enzyme could serve as a useful marker for lens metabolic function.
In addition to glycolysis, an alternative pathway, the hexose monophosphate shunt, accounts for another 10-20% of the glucose metabolized by the lens [4,8]. In the first step of this pathway, the enzyme glucose-6-phosphate dehydrogenase (G6PD) catalyzes the first conversion of glucose-6-phosphate by reducing NADP+ to NADPH. The reduced NADPH is then used to maintain glutathione in its reduced state. The hexose monophosphate shunt can increase under conditions of oxidative stress in order to meet the greater need of NADPH by the glutathione scavenger system , and conversely an inhibition of the hexose monophosphate shunt can make the lens more vulnerable to oxidative damage . Thus, in addition to its contribution to lens metabolism, G6PD plays an important role in protecting the lens from oxidation.
The epithelial layer contains the bulk of the metabolic enzymes in the lens, and it has been suggested that the epithelial layer may be the first site of oxidative damage in the lens that precedes the development of lens opacity . As interference with the lens metabolic pathways may be directly related to the disruption of lens optical quality, the recovery of key epithelial enzymes could be a necessary step in the lens repair processes. It has been shown that lens epithelial enzyme activity can increase in response to a toxic insult, suggestive that enzyme function may have an active role in lens recovery mechanisms [12-14].
The repair processes utilized by the lens in response to damage are not fully understood, and the formation of cataracts may be related to a failure in these mechanisms. It was hypothesized that the recovery of lens optical quality could be related to the activity of key metabolic enzymes. In this study, the epithelial activities of the enzymes hexokinase and glucose-6-phosphate dehydrogenase were measured in lenses recovering from pharmaceutical-induced optical damage.
Culture reagents were obtained from Gibco-BRL (Burlington, ON, Canada). All other chemicals were obtained from Sigma (St. Louis, MO), unless noted otherwise. The three commercial anesthetics used in this study have trademarked brand names; Alcaine® (Alcon Canada, Mississauga, ON, Canada), Fluress® (Allergan, Markham, ON, Canada), and Fluoracaine® (Akorn Pharmaceuticals, Markham, ON, Canada).
Bovine eyes were obtained from a local abattoir and the lenses were excised within 1-5 h post-mortem under sterile conditions, as in previous work [2,15,16]. Briefly, the lenses were placed in a two-chambered cell made from glass and silicon rubber and immersed in 25 ml of culture medium consisting of Medium 199 with Earle's salts, 100 mg/l glutamine, 3% fetal bovine serum, 2.2 g/l sodium bicarbonate, 5.96 g/l HEPES and 1% antibiotics (100 units/ml penicillin and 0.1 mg/ml streptomycin). The cultured lenses were incubated at 37 °C under an air atmosphere with 4% CO2 for 48 h prior to experimental use to ensure that lenses with visible opacities or damage were removed from the study.
As in a previous study , the treated lenses were exposed for 2 h to one of three commercial corneal anesthetic solutions; Alcaine® (0.5% proparacaine HCl, 0.01% benzalkonium chloride), Fluress® (0.25% fluorescein sodium, 0.4% benoxinate HCl, 1% chlorobutanol), or Fluoracaine® (0.25% fluorescein sodium, 0.5% proparacaine HCl, 0.01% thimerosal). The culture medium was drained from the lens containers, and the treated lenses were then completely submerged in the undiluted commercial anesthetic solutions. All lenses, including the untreated control lenses, were stored at room temperature during the 2 h exposure period. The anesthetic solutions were then aspirated off, the lenses were rinsed three times with culture medium, and the lens containers (including those in the control group) were replaced with fresh culture medium. The lenses were incubated at 37 °C (4% CO2-air) for the remainder of the experiment except for the short periods (5-10 min) the lenses were in the automated laser scanner system.
Lens optical quality was assessed with the ScanTox In Vitro Assay System (Harvard Apparatus, Holliston, MA), a redesigned automated laser scanner , consisting of a 4 mW 670 nm diode laser mounted on a computer driven X-Y table, two digital video cameras and a video frame digitizer. A series of laser beams, parallel to the lens optical axis, were directed towards the cultured lens along one meridian in 0.5 mm increments. The laser beams were refracted after passing through the lens, and the ScanTox system software calculated the back vertex focal length for every beam position (measured as the distance from the back vertex of the lens to the intercept of the refracted beam with the optical axis). After the 2 h exposure to the test solutions, the lenses were scanned at 1, 3, 5, 8 and 24 h-intervals following the return to fresh culture medium. As in previous work , the focal length variability (FLV) was used to assess optical damage. A large variance (measured as the standard error of the mean) in the focal lengths measured for the different beam positions is indicative of reduced lens optical quality.
The probability of differences in FLV between treated and control lenses at each time point was determined with a one-way ANOVA and the post hoc Dunnett's multiple comparison test (p<0.01). A significance level of p<0.01 (the Bonferroni correction to account for multiple time-points) was necessary in the overall ANOVA model before proceeding to the Dunnett's test.
To illustrate changes in lens transparency induced by the three anesthetics, photographs were taken of a representative lens in each test group at the 1 and 8 h time point. The lens was submersed completely in 0.9% saline and placed, with its anterior surface up, on top of a grid consisting of white lines and a black background. The lenses were then photographed using a Nikon light microscope equipped with a 35 mm camera using Ilford Pan F Plus ISO 50 black and white film.
Enzyme analysis was performed 24 h after the 2 h exposure to the corneal anesthetics. Dissection of the lens was performed on a tray of ice and the lens was thoroughly rinsed with 0.9% saline to rinse off any debris (i.e. pigment from the ciliary body). A cut was made along the entire equator of the lens with iris scissors and then the lens capsule and its adhering epithelium were peeled away from the rest of the lens using jewelers' forceps. Care was taken to ensure that no lens fiber material remained attached to the epithelium.
The capsule was then placed in an Eppendorf tube containing 350 μl of 50 mM potassium phosphate buffer (pH 7.0). The tissue was twice sonicated at 50 W for 10 s, in bursts lasting 60% of every s, while the tube was surrounded by an ethanol ice bath. The sample was centrifuged for 10 min at 14,000x g and then the enzyme activities of the supernatant were measured.
HK activity was assayed using the method of Dovrat and Weinreb  with a reaction mixture of 50 mM Tris HCl (pH 8.0), 13.3 mM MgCl2, 10 mM glucose, 1 unit G6PD, 0.5 mM ATP, and 0.2 mM NADP. G6PD activity was determined using the method of Dovrat and Gershon , and the assay mixture consisted of 55 mM Tris HCl (pH 7.8), 3.3 mM MgCl2, 3.5 mM glucose-6-phosphate, and 0.2 mM NADP. For both assays, the formation of the product NADPH was measured by monitoring the increase in absorbance at 340 nm using a spectrophotometer.
The ingredients for each assay were thoroughly mixed with a sample of the lens epithelial cell homogenate, poured into plastic cuvettes, and then immediately placed in the holder of an Ultrospec 2000 single-beam spectrophotometer (Amersham Biosciences, Piscataway, NJ). The cuvette compartment in the spectrophotometer was thermo-regulated with a circulating water bath to maintain a constant temperature of 30 °C. Every assay was performed in duplicate, and a blank was run for each enzyme assay by adding potassium phosphate buffer (pH 7.0), rather than a sample of cell homogenate, to the reaction mixture.
Analysis of the enzyme assays was performed with SWIFT-KIN software (Amersham Biosciences). The absorption at 340 nm for each assay was measured for 7.5 min and the slope of the resultant curve was determined between the 2 min and the 7.5 min time points. The first 2 min were omitted from the slope calculation to allow for the temperature to stabilize and the reaction to commence. The software then subtracted the absorbance curve of the appropriate blank from the cell sample absorbance curve, and the slope (dA/dt) of the resulting curve was used to calculate relative enzyme activity.
The remaining supernatant not used in the assays was stored at -70 °C for 2-3 weeks when protein analysis was performed. Protein determination was measured using a Modified Lowry Kit (P5656; Sigma) that is based on Peterson's modification  of the micro-Lowry method . The absorbance was measured at 750 nm, and the protein concentration of the sample was determined from a bovine serum albumin calibration curve using SWIFT-QUANT software (Amersham Biosciences).
The relative enzyme activities were compared using a nonparametric Kruskal-Wallis one-way ANOVA on ranks with the Student-Newman-Keuls (SNK) post hoc test (p<0.01 considered statistically significant)
The three anesthetics each had a distinctly different effect on optical function, as determined by measuring the change in focal length variability (Figure 1). The FLV of lenses exposed to Alcaine® was sharply increased at the first scan, one h following their return to fresh culture medium. The damage appeared to peak at the 3 h time-point, after which the lens began to recover to near-control levels at the 24 h mark. The increase in FLV induced by Fluoracaine® was more gradual and was still increasing after 8 h. Similar to the Alcaine® treated lenses, the Fluoracaine® treated lenses had nearly recovered by the 24 h time-point, although both groups were still significantly different than the control lenses (p<0.01, ANOVA, Dunnett's). Fluress® did not have a noticeable effect on lens optical function. Although the Fluress® treated lenses displayed a slightly increased FLV relative to control lenses, this was not significant at any time during the 24 h study.
Photographs of one representative lens from each treatment group, taken at the 1 and 8 h time-point with the anterior surface facing the camera, illustrate the change in both optical quality and transparency caused by the three anesthetics (Figure 2). One h following the initial 2 h exposure, the Alcaine® treated lens was quite hazy and its optical quality was obviously disrupted (grid appears distorted). The "Y" suture appeared to be especially affected by Alcaine®, as it developed an opaqueness that made it very prominent (Figure 2A). Both the Fluoracaine® and the Fluress® treated lenses looked similar to each other, appearing slightly hazy but the image of the underlying grid pattern was reasonably distinct (Figure 2B and Figure 2C). The control lens (Figure 2D) had excellent optical quality and transparency. It was noted that the opacities caused by exposure to the anesthetics seemed to be localized in the anterior and posterior sub-capsular regions, with some anterior cortical changes.
After 8 h in the fresh culture medium, the Alcaine® treated lens had improved from the 1 h time point, but the suture was still clearly visible and the lens was somewhat hazy (Figure 2E). The anterior capsule of the Fluoracaine® treated lens had become more opaque at 8 h post exposure (Figure 2F) than it was after 1 h. Visually, it appeared that the opacities were localized around the anterior and posterior sutures, but the suture itself was not as clearly affected as in the Alcaine® exposed lenses. The Fluress® treated lens exhibited a diffuse haze and there were vacuoles present in the anterior sub-capsular region, but the image of the grid pattern appeared fairly distinct (Figure 2G). The control lens was clear, and is shown as a reference (Figure 2H).
The protein concentration in each sample of epithelial cell sonicate was measured to express the results in terms of specific enzyme activity. However, it was noted that the protein content in the Alcaine® and Fluress® treated samples was consistently higher than that for Fluoracaine® treated or control samples (Figure 3A, p<0.01, ANOVA, SNK). Due to these differences in protein concentrations, relative enzyme activity was calculated rather than specific enzyme activity. For each of the three treatment groups, the rate of product formation (dA/dt) in the enzyme assays was expressed as a percentage of the dA/dt obtained in the assays for the untreated control samples (see Discussion below for further comment).
For both the HK and G6PD enzyme assays, the measured change in absorbance over time for the control samples was approximately linear (linearity coefficient p>0.985), indicating that the assay product formation was occurring at a constant rate. The relative enzyme activities for HK (Figure 3B) and G6PD (Figure 3C) were significantly reduced in the epithelial supernatant from the bovine lenses exposed to Alcaine® (19.1% HK, 6.1% G6PD), Fluoracaine® (18.5% HK, 11.9% G6PD), and Fluress® (16.2% HK, 6.8% G6PD), as compared to the controls (p<0.01, Kruskal-Wallis ANOVA, SNK).
The role of metabolic enzymes in cataract formation is not fully understood. While it has been shown that enzyme activities measured in whole lenses appear to decline with age [18,21], it has been suggested that this could be due to an accumulation of more metabolically inert cells in the lens nucleus rather than a loss in enzyme function . Nevertheless, it is thought that a disruption in lens enzyme function may contribute to cataract development [5,11,23]. It has been reported that a large proportion of epithelial samples from human cataractous lenses show HK inactivation , and that mature cataracts exhibit reduced G6PD activity relative to immature cataracts .
The enzyme activity measured in human lens tissue can display large individual variability [26-28], and the lack of a comparison to age-matched clear controls is problematic as samples are often only obtained from subjects that have undergone cataract surgery. Inducing lens damage experimentally allows for a controlled disruption of optical quality, and it permits a direct comparison of opaque lenses to clear control lenses. There have been a number of experimental models of cataract that have been used to study the biochemical changes associated with lens opacity formation [6,12,13,17,29].
In this study, the relative effect on focal length variability induced by the three anesthetics, as determined with the ScanTox system, paralleled the results from a previous study using an older-generation version of an automated laser scanner . The 2 h exposure to each solution had a unique effect on lens function. Alcaine® caused a severe, acute increase in focal length variability, from which the lens began to recover after about 5 h. Fluoracaine® resulted in less pronounced optical damage and a slower recovery than Alcaine®. Fluress® again did not alter optical quality at any time, relative to the controls. Although the Alcaine® and Fluoracaine® treated lenses differed significantly from the controls at the 24 h mark, the FLV had improved to near-control levels, indicating that the optical recovery mechanisms were intact. The photographs of the lenses offer visual evidence of the effect each anesthetic had on the lens optical quality, relative to the untreated controls.
It is not the assertion of this study that the disruption of lens optics induced by topical anesthetics serves as an exact model for human cataracts. The loss of lens transparency and optical quality, i.e. cataract formation, is likely a complex multi-factorial process . However, these results do indicate that these anesthetic solutions can cause distinct and repeatable patterns of optical damage and recovery. Cataractogenesis may be related to a failure in the lens repair processes and these mechanisms are not well characterized. The enzyme activity in the anesthetic treated lenses was determined at the 24 h point to investigate potential repair mechanisms within the lens. It has previously been demonstrated that an increase in the activity of a number of enzymes including G6PD can occur in lenses with naphthalene- and streptozotocin-induced cataracts . Similarly, it has been shown that cultured epithelial cells exposed to oxidative stress can display increased activity of a number of protective enzymes including G6PD . These studies suggest that an increase in epithelial enzyme activity may play a key role in the maintenance of lens optical quality.
In this study, lens exposure to each of the three anesthetics resulted in a large decrease in HK and G6PD activity, relative to controls, at the 24 h time-point following removal of the test solution. At this time-point, Alcaine® and Fluoracaine® treated lenses displayed substantial optical improvement from the damage evident earlier in the study. The optical quality of Fluress® treated lenses was not significantly different from controls at any time-point during the optical analysis, including the 24 h end point. These results suggest that it is possible for a lens to retain and/or recover its optical function despite a significant reduction in its epithelial activity of the metabolic enzymes HK and G6PD.
An unexpected finding during the course of these studies was the difference in the protein concentrations obtained for the epithelial samples. The concentration for Alcaine® treated and Fluress® treated samples, while similar to each other, was significantly increased relative to that obtained for the control and Fluoracaine® treated samples. It is possible that the exposure to these two anesthetic compounds induced an increased production of protein, or altered the ratio of water soluble to insoluble proteins in the lens epithelial cells. As the dissection of the whole epithelium was kept as consistent as possible for every lens, it is unlikely that the differences in protein content were due to a true difference in the number of epithelial cells. Therefore, using specific activity values (dividing the rate of product formation by the protein concentration) could artificially give lower values for the enzyme activity in the Alcaine® and Fluress® treated lenses. Instead, the rate of product formation in the enzyme assays for the 3 treatment groups was normalized to that of the control group, with the assumption that the number of epithelial cells extracted during the dissection was similar for each of the treatment groups. Using this more conservative approach, the enzyme activities for HK and G6PD in all 3 treatment groups were still dramatically reduced (by more than 80%) relative to the untreated control samples.
There have been very few studies that have tried to quantify changes in lens optical quality objectively, rather than using a subjective grading of lens transparency. Dovrat and Weinreb  used an automated laser scanner similar to the one used in this study to compare the optical quality to the enzyme activity of HK, G6PD, and catalase in bovine lenses damaged by UV-A radiation and found a decrease in enzyme activity before changes in optical quality were apparent. In another study, these researchers reported that lens optical quality could recover from UV-A damage, despite the observation that the enzyme activity of NaK-ATPase at the equator of the lens showed no recovery . This is in agreement with the findings of this study that recovery of lens optics can occur with apparent compromised epithelial enzyme function.
In summary, exposure of the bovine lens to three commercial anesthetics resulted in repeatable changes in lens optical quality, measured as a change in the focal length variability. These pharmaceutical-induced optical changes offer a controlled model to investigate the biochemical processes involved in both lens damage and recovery. The ScanTox In Vitro Assay System provides an objective measure of lens optical quality, enabling a more quantitative comparison to lens biochemical changes. The epithelial enzyme activities for HK and G6PD were significantly reduced in samples from each of the three treatment groups, relative to the controls, 24 h after exposure to the test solutions. These results indicate that it is possible for lenses to recover (in the case of the Alcaine® and Fluoracaine® treated lenses) or maintain (in the case of the Fluress® treated lenses) optical function despite having dramatically reduced activity levels for two key metabolic enzymes. These findings suggest that while decreased epithelial metabolic function may be associated with reduced lens optical quality, it is not necessarily a causative factor.
We would like to thank Kelley Moran for her assistance on this project. This work was supported by the an operating grant to JGS from the Natural Sciences and Engineering Research Council of Canada (NSERC) and an NSERC PGS-A Scholarship to ATEH.
1. Sivak JG, Gershon D, Dovrat A, Weerheim J. Ocular lens culture for in vitro toxicology. Alternative methodology in toxicology 1986; 5:499-505.
2. Sivak JG, Herbert KL. Optical damage and recovery of the in vitro bovine ocular lens for alcohols, surfactants, acetates, ketones, aromatics, and some consumer products: a review. J Toxicol Cutaneous Ocul Toxicol 1997; 16:173-87.
3. Hartwick ATE, Sivak JG, Herbert KL. Relative toxicity of three corneal anesthetics measured in vitro with the cultured bovine lens. J Toxicol Cutaneous Ocul Toxicol 1997; 16:253-66.
4. Cheng HM, Chylack LT. Lens metabolism. In: Maisel H, editor. The ocular lens: structure, function, and pathology. New York: Dekker; 1985. p. 223-64.
5. Paterson CA, Delamere NA. The lens. In: Hart WM Jr, editor. Adler's physiology of the eye: clinical application. 9th ed. St. Louis: Mosby; 1992. p. 348-90.
6. Kletzky DL, Tung WH, Chylack LT Jr. The protective effect of glucose on soluble rat lens hexokinase in the presence of oxidative stress. Curr Eye Res 1986; 5:433-9.
7. Qiao F, Xing K, Lou MF. Modulation of lens glycolytic pathway by thioltransferase. Exp Eye Res 2000; 70:745-53.
8. Kinoshita JH. Carbohydrate metabolism of lens. Ama Arch Ophthalmol 1955; 54:360-8.
9. Giblin FJ, Nies DE, Reddy VN. Stimulation of the hexose monophosphate shunt in rabbit lens in response to the oxidation of glutathione. Exp Eye Res 1981; 33:289-98.
10. Giblin FJ, McCready JP. The effect of inhibition of glutathione reductase on the detoxification of H2O2 by rabbit lens. Invest Ophthalmol Vis Sci 1983; 24:113-8.
11. Spector A. Oxidative stress-induced cataract: mechanism of action. FASEB J 1995; 9:1173-82.
12. Korte I, Hockwin O, Bours J, Wegener A. Alterations of lens metabolism with experimentally induced cataract in rats. Ophthalmic Res 1988; 20:174-8.
13. Huang LL, Zhang CY, Hess JL, Bunce GE. Biochemical changes and cataract formation in lenses from rats receiving multiple, low doses of sodium selenite. Exp Eye Res 1992; 55:671-8.
14. Spector A, Wang RR, Ma W, Kleiman NJ. Development and characterization of an H2O2-resistant immortal lens epithelial cell line. Invest Ophthalmol Vis Sci 2000; 41:832-43.
15. Sivak JG, Stuart DD, Herbert KL, Van Oostrom JA, Segal L. Optical properties of the cultured bovine ocular lens as an in vitro alternative to the Draize eye toxicity test: preliminary validation for alcohols. Toxicology Methods 1992; 2:280-94.
16. Sivak JG, Herbert KL, Segal L. Ocular lens organ culture as a measure of ocular irritancy: the effect of surfactants. Toxicology Methods 1994; 4:56-65.
17. Dovrat A, Weinreb O. Recovery of lens optics and epithelial enzymes after ultraviolet A radiation. Invest Ophthalmol Vis Sci 1995; 36:2417-24.
18. Dovrat A, Gershon D. Rat lens superoxide dismutase and glucose-6-phosphate dehydrogenase: studies on the catalytic activity and the fate of enzyme antigen as a function of age. Exp Eye Res 1981; 33:651-61.
19. Peterson GL. A simplification of the protein assay method of Lowry et al. which is more generally applicable. Anal Biochem 1977; 83:346-56.
20. Lowry OH, Rosenbrough NJ, Farr AL, Randall RJ. Protein measurement with the folin phenol reagent. J Biol Chem 1951; 193:265-75.
21. Cheng HM, Chylack LT Jr, von Saltza I. Supplementing glucose metabolism in human senile cataracts. Invest Ophthalmol Vis Sci 1981; 21:812-8.
22. Zheng WZ, Augusteyn RC. Ageing of glutathione reductase in the lens. Exp Eye Res 1994; 59:91-5.
23. Lou MF. Thiol regulation in the lens. J Ocul Pharmacol Ther 2000; 16:137-48.
24. Dovrat A, Horwitz J, Sivak JG, Weinreb O, Scharf J, Silbermann M. DL-propanolol inhibits lens hexokinase activity and affects lens optics. Exp Eye Res 1993; 57:747-51.
25. Dwivedi RS, Pratap VB. Alteration in glutathione metabolism during cataract progression. Ophthalmic Res 1987; 19:41-4.
26. Huang QL, Lou MF, Straatsma BR, Horwitz J. Distribution and activity of glutathione-S-transferase in normal human lenses and in cataractous human epithelia. Curr Eye Res 1993; 12:433-7.
27. Ohrloff C, Hockwin O, Olson R, Dickman S. Glutathione peroxidase, glutathione reductase and superoxide dismutase in the aging lens. Curr Eye Res 1984; 3:109-15.
28. Belpoliti M, Maraini G, Alberti G, Corona R, Crateri S. Enzyme activities in human lens epithelium of age-related cataract. Invest Ophthalmol Vis Sci 1993; 34:2843-7.
29. Spector A, Wang GM, Wang RR, Garner WH, Moll H. The prevention of cataract caused by oxidative stress in cultured rat lenses. I. H2O2 and photochemically induced cataract [published erratum in Curr Eye Res 1993; 12:379]. Curr Eye Res 1993; 12:163-79.
30. Dovrat A, Weinreb O. Effects of UV-A radiation on lens epithelial NaK-ATPase in organ culture. Invest Ophthalmol Vis Sci 1999; 40:1616-20.