Effect of UVA Light on the Activity of Several Aged Human Lens Enzymes

Abstract

purpose. To determine the effect of UVA irradiation on the specific activities of several protective and metabolic enzymes in aged human lenses.

methods. Intact human lenses (ages 55–75) in artificial aqueous humor were irradiated in a quartz cuvette with UVA light (925 J/cm² per hour) at +20°C. The lenses were homogenized and the activities of enzymes in the water-soluble (WS) fraction were measured in irradiated and nonirradiated lenses.

results. One hour of UVA photolysis of human lens resulted in a 70% loss in glutathione reductase (GR)–specific activity and a 24% loss in glyceraldehyde-3-phosphate dehydrogenase (G3PD)–specific activity. At the same time, glutathione peroxidase (GPx) and superoxide dismutase (SOD) showed little or no loss in specific activity. GR and G3PD showed similar losses when human lenses were photolyzed with the same dose of UVA light delivered to the lens over 8 hours, using a 12.5% neutral-density filter (ndf), or over 24 hours with a 4.25% ndf. One hour photolysis of the human lens WS fraction under anaerobic conditions yielded an almost complete inactivation of GR, but only an 18% loss of G3PD activity. Under aerobic conditions, however, both enzyme activities were almost completely lost. Clear fetal bovine lenses, photolyzed under the identical conditions, displayed essentially the same loss of GR activity.

conclusions. UVA light causes inactivation of GR in human and fetal calf lenses under both anaerobic and aerobic conditions. This suggests that flavine adenine dinucleotide (FAD), the prosthetic group of GR, may be responsible for the enzyme’s self-sensitizing properties. WS proteins from aged human lens generate reactive oxygen species (ROS) during UVA irradiation, which may be responsible for the inactivation of G3PD.

To maintain visual acuity, the human eye lens must remain transparent throughout the life of an individual. In addition, the human lens absorbs all the impinging light between 320 and 400 nm (UVA) due to the presence of intrinsic UV-filters, ^{1 2} thus protecting the retina from UV-mediated photochemical damage. ^{3 4 5} With advancing age, as a result of the accumulation of posttranslational modifications of the constituent proteins, the lens exhibits a nontryptophan fluorescence and extensive protein aggregation. ^{6 7 8 9} These processes, which continue throughout the life of an individual, may ultimately lead to the formation of high-molecular-weight aggregates that can scatter the light entering the lens. ^{10 11} It is thought that similar events occur at a greatly accelerated rate during the development of age-onset and brunescent cataract in humans and may cause the opacification of the lens. ^{7 9 10 11} Although the exact molecular mechanisms responsible for such events are not fully established, UVB-dependent formation of tryptophan oxidation products, ¹² glycation-mediated formation of UVA-responsive sensitizers, ¹³ UVA-induced formation of reactive oxygen species, ¹⁴ and UV-induced oxidation of lens proteins, ¹⁵ DNA, ¹⁶ and membranes ¹⁷ are commonly considered to be the contributing factors in cataractogenesis in humans. 

In this regard, epidemiologic studies conducted by Zigman et al. ¹⁸ and Brilliant et al. ¹⁹ showed that the exposure to high levels of the UV component of sunlight correlates directly with the incidence of cataract in humans. According to Sliney, ²⁰ of all the UV light that reaches the lens, the UVB component accounts for only 50 to 200 nW/cm² of absorbed light. These amounts of light are usually absorbed in the first 0.4 to 0.5 mm of the lens by tryptophan residues present in the lens proteins. ²¹ In contrast, the amount of UVA light that is absorbed by the lens is 1700 times higher, ¹⁸ and it is fully absorbed by compounds in the primate lens. ²²  

The recent review by Balasubramanian ²³ points out that lenticular damage can indeed occur due to the photodynamic action of several UVA-absorbing compounds that accumulate in the lens with age or in cataract. Reports from our laboratory have already shown that the lens chromophores that accumulate with age and during cataractogenesis can act as UVA-responsive sensitizers capable of producing significant amounts of reactive oxygen species (ROS), mostly singlet oxygen. ¹⁴ We have showed that the levels of ROS produced in vitro could be sufficient to oxidize Trp, His, and Cys residues within the lens proteins. ¹⁵ More important, our data have shown that the lens proteins become oxidized, even in the presence of high concentrations of glutathione (GSH) and ascorbic acid. ¹⁵ Because diminution of lens antioxidants and metabolic enzymes with age and during the development of cataract in humans is a well-characterized phenomenon, ²⁴ we hypothesized that UVA irradiation may have a synergistic effect with age on the loss of activity of lenticular enzymes. This effect was seen in the inactivation of lens detoxifying and metabolic enzymes, either by direct photolysis of the lenticular enzymes or through an oxidation of the enzymes by the ROS, which are formed as a direct outcome of UVA photolysis of lens sensitizers in the presence of oxygen. ^{14 15} Although little is known about the interactions of UVA light within lens enzymes, this idea has been studied by Torriglia and Zigman ²⁵ and Dovrat and Weinreb, ²⁶ who have shown that UVA light inhibits Na,K-adenosine triphosphatase (ATP) in rat and bovine lenses, respectively. These two groups have also shown that, depending on the dose of UVA irradiation, the photolysis of bovine lenses in cell culture leads to the inactivation of catalase, ^{26 27} hexokinase, and glucose-6-dehydrogenase. ²⁶  

The present study was conducted to elucidate further the impact of UVA irradiation on the activities of the lens oxygen-detoxifying enzymes, superoxide dismutase (SOD), glutathione reductase (GR), and glutathione peroxidase (GPx), and the activity of the metabolic enzyme, glyceraldehyde-3-phosphate dehydrogenase (G3PD), during photolysis of intact aged human and fetal calf lenses, and to study whether ROS generated by UVA-light–excited lenticular chromophores cause the inactivation of the above enzymes. 

Materials and Methods

Cytochrome c, xanthine oxidase, hypoxanthine, EDTA Na₂ salt, xanthine oxidase (EC 1.1.3.32 from buttermilk), SOD (EC 1.15.1.1 from bovine liver), GR (EC 1.6.4.2 type III from baker’s yeast), GPx (EC 1.11.1.9 from bovine erythrocytes), G3PD (EC 1.2.1.12 from rabbit muscle), glyceraldehyde-3-phosphate diethylacetal, reduced glutathione (GSH), oxidized glutathione (GSSG), nicotinamide adenine dinucleotide (NAD), and nicotinamide adenine dinucleotide phosphate (NADP) were of the highest quality available from Sigma Chemical Co. (St. Louis, MO). Reduced nicotinamide adenine dinucleotide (NADH) and reduced nicotinamide adenine dinucleotide phosphate (NADPH) were purchased from Alexis (San Diego, CA). 

All the salts, perchloric acid, iodoacetic acid, bathophenantrolinedisulfonic acid (BPDSA), 2,4-dinitrofluorobenzene, l-lysine dihydrochloride, cysteine, cystine, and 3-methyl phenol (m-Cresol Purple), used in this project were the highest analytical degree available from Sigma Chemical Co. (St. Louis, MO), Aldrich (Milwaukee, WI), or Fischer Scientific (Pittsburgh, PA). 

dl-Glyceraldehyde-3-phosphate was prepared from dl-glyceraldehyde-3-phosphate diethylacetal monobarium salt (Fluka, Milwaukee, WI) by hydrolyzing the salt in the presence of an ion-exchange resin (Dowex 50 ×4 H⁺; Dow Chemical Co., Midland, MI) in H⁺-form in boiling water for 3 to 5 minutes, according to the method published by Worthington Biochemical Co. (Freehold, NJ). ²⁸  

Normal Human Lenses

Human lenses (55–75 years old) were obtained after death from donor’s eye globes at the Heartland Lions Eye Tissue Bank, Columbia, MO. The tissue was obtained and used in accordance with the provisions of the Declaration of Helsinki for research involving human tissue. The lenses were extracted from the each pair of eyes (kept at +4°C) by dissection within 24 to 48 hours after removal of the cornea. Lens pairs were placed immediately into 5 mL sterile artificial aqueous humor (AAH) of the following composition: NaCl, 130 mM; KCl, 5 mM; CaCl₂, 1 mM; MgCl₂, 1 mM; d-glucose, 2.5 mM; NaHCO₃, 10 mM; HEPES, 10 mM (pH 7.3) in 10-mL screw-cap vials. ²⁹ The lenses were kept on ice and were usually used for the irradiation experiments within 1 hour after dissection from the globes. 

Fetal Calf Lenses

Enucleated fetal calf eyes were shipped in ice from PelFreeze (Rogers, AK) to our location within 12 hours after death. The lenses were excised immediately on arrival, rinsed twice with 5 mL sterile artificial aqueous humor and placed into 10 mL ice-cold AAH in 20-mL screw-cap vials. The lenses were stored on ice and were usually used for the irradiation experiments within 1 hour after dissection. 

Whole-Lens Irradiation

Irradiation experiments were performed in a 1.0 × 4.0-cm quartz cuvette (5-mm light path) containing 1 mL AAH at +17°C under aerobic conditions. The lens was positioned in the cuvette with the anterior portion of the lens toward the light beam. A 1000-W Hg-Xe high-pressure lamp (Oriel Corp., Stratford, CT) delivered the UV light. The light was filtered through a 10-cm water-jacketed filter containing 5% CuSO₄ and then through a 338-nm cutoff filter. Both the water-jacketed cuvette holder and the copper sulfate filter were cooled with a circulating water bath set at +17°C. According to our measurements, temperatures, no higher than 20°C were detected during 1 to 24 hours of irradiation. The intensity of UVA light was monitored in each experiment by placing a black block probe connected to a radiometer (Model 65A; YSI-Kettering, Yellow Springs, OH) behind the cuvette. Illumination was continued for the specified time at a total irradiance of 0.925 kJ/cm² per hour for light between 338 and 475 nm. A lens from the matching pair was kept in the dark at room temperature and served as a control. After irradiation, each lens was homogenized in a 2-mL tissue grinder (Dounce; Bellco Glass Co, Vineland, NJ) with 0.3 mL of ice-cold 10 mM phosphate buffer (pH 7.0), containing 0.5 mM EDTA. The homogenizer was rinsed with 0.2 mL of the same buffer. The homogenate and the rinse solution were combined in 1.5-mL snap-cap tube (Eppendorf, Fremont, CA) and centrifuged at 15,000 rpm for 30 minutes in a centrifuge (RC 2B; Sorvall, Newtown, CT) at +4°C. The lens supernatant was collected carefully with a pipette, and the pellet was resuspended again in 0.5 mL of the buffer followed by another cycle of centrifugation. The combined lens water-soluble (WS) fraction was kept on ice at all times and used directly for the enzymatic assays. In some experiments, whole lenses were separated into cortex and nucleus exactly as described by Rogers and Augusteyn ³⁰ and processed separately. 

The enzymatic activities of GR, GPx, G3PD, and SOD in the WS fractions from human and calf lenses were assayed on the day of irradiation under the same conditions for both the UV-irradiated samples and the dark control. Unless otherwise stated in the figure legends, all the enzymatic assays were performed in triplicate. Student’s t-test was applied for statistical analysis of the data. A change was considered significant if the difference between the specific enzymatic activity in UV-irradiated lens and the dark control was P < 0.05 or less. 

Enzymatic Assays

G3PD activity in the lens homogenate was estimated according to a modification of the procedure described by Velick, ³¹ recording the increase in absorbance at 340 nm (formation of NADH) at +25 ± 0.1°C. The reaction mixture contained 3.0 mM sodium arsenate, 0.3 mM NAD, 3.5 mM dithiothreitol (DTT), and an aliquot of the lens WS fraction in 15 mM sodium pyrophosphate buffer at pH 8.5. The reaction was initiated by the addition of 0.1 mL of freshly prepared 15 mM dl-G3PD with a cuvette mixer. One unit of enzyme was defined ²⁸ as 1 μmol/min NADH formed at 25°C. 

GR activity was measured according to the procedure of Carlberg and Mannervik. ³² The 1.55-mL reaction mixture consisted of degassed 100 mM KPO₄ buffer (pH 8.0), 1.0 mM EDTA, 1.0 mM oxidized glutathione (GSSG), and 0.2 mM NADPH. The reaction was initiated by addition of 50 μL of lens homogenate and the decrease in the optical density at 340 nm was recorded at 25°C. One unit of enzyme was defined ³⁰ as 1 μmol/min NADPH oxidized at 25°C. 

GPx activity was estimated according to the procedure of Paglia and Valentine. ³³ Each 1.55-mL reaction mixture consisted of 150 mM KPO₄ buffer (pH 7.0) containing 1.5 mM EDTA, 0.16 mM Tris-HCl, 62.5 μM NADPH, 1.25 mM GSH, 1.25 U/mL GR, and 50 μL lens homogenate. The reaction was initiated by an addition of 50 μL of 0.455 mM t-butyl peroxide. The decrease in optical density at 340 nm due to the oxidation of NADPH was monitored for 5 minutes in this coupled assay at +25°C. The units of enzymatic activity were calculated using an extinction coefficient of 6.22 mM/cm for NADPH. One unit was equivalent to the oxidation of 1 μmol/min of NADPH. 

SOD activity was measured by the cytochrome c spectrophotometric method of McCord and Fridovich ³⁴ with the modifications described by Elstner et al. ³⁵ SOD activity in the lens homogenate was expressed as the amount of enzyme that caused a 50% inhibition of cytochrome c reduction under the conditions of the assay. 

Chemical Analyses

Oxygen Measurements.O₂ measurements on the human lenses were performed with a fiber optic oxygen sensor system (FOXY; Ocean Optics, Inc., Dunedin, FL) equipped with a probe (Foxy-AL300; a 300-μm aluminum jacketed fiber optic probe embedded in a 22-gauge needle; Ocean Optics, Inc.). The signal generated from the probe was continuously transferred to a minispectrophotometer (Ocean Optics, Inc.) connected to the computer. The acquired data were processed by the software (OOI Sensors; Ocean Optics, Inc.). The probe and software were calibrated by using two standards of known oxygen concentration, as suggested by the manufacturer. Determination of O₂ level was performed under anaerobic conditions (Ar atmosphere) in a glove box at room temperature. 

Lenses from human poles (2 days after death, kept at +4°C) were dissected under sterile conditions. The left lens was placed into a 2-mL quartz cuvette (5-mm light path) filled with 1.3 mL sterile AAH and subjected to UVA irradiation for the times specified in Figure 4 (light intensity 925 J/cm² per hour). The right lens was placed in a 2-mL matching cuvette filled with AAH and kept in the dark at room temperature. 

At the end of irradiation, each lens was quickly and carefully dried on a filter paper and placed in a preweighed 4-mL screw-cap quartz cuvette with 2 mL of 50 mM KPO₄ buffer (degassed, pH 7.0) in an Ar atmosphere and sealed with a rubber septum. A fiber optic oxygen sensor was introduced into the solution by puncturing the septum. After a 1-minute equilibration, vigorous stirring with a sharpened magnetic stirrer bar disrupted the human lens cortex. Oxygen measurements were conducted for 10 minutes or stopped once the oxygen level reached a plateau. At the end of the oxygen measurement, the 4-mL cuvette containing the disrupted lens was weighed again. Oxygen concentration in the lens was calculated based on the difference between O₂ concentration at the plateau and at the threshold level multiplied by the dilution factor (weight of the lens plus buffer minus weight of the buffer) and expressed in moles per lens. 

Determination of sulfur-containing Nonprotein Compounds.The levels of reduced and GSSG and cysteine in the irradiated and nonirradiated lenses were determined by the method of Fariss and Reed. ³⁶ In brief, human lenses, both control or UVA irradiated, were weighed, and the nucleus of each lens was separated from the cortex by a no. 3 cork borer. Both lens cortex and nucleus were homogenized in a 2-mL tissue grinder (Dounce; Bellco Glass Co.) with 0.5 mL of degassed ice-cold 1 mM solution of BPDSA and internal standard γ-Glu-Glu. The mixtures were transferred to 1.5-mL snap-cap tubes, blanketed with Ar atmosphere, and centrifuged at 15,000 rpm (RB 2C centrifuge; Sorvall) at 4°C for 10 minutes. The supernatants were carefully withdrawn and mixed with equal volumes of ice-cold 20% perchloric acid and 2 mM BPDSA and centrifuged again. The supernatants were adjusted to pH 8.2–8.4 with degassed 2 M KOH and 2.4 M KHCO₃ and reacted with iodoacetic acid to convert reduced thiols to their corresponding S-carboxymethyl derivatives. The amino groups of lens sulfur-containing compounds were further modified with 2,4-dinitro-fluorobenzene. At the end of the derivatization procedure, samples were chromatographed on an HPLC system (Waters; Milford, MA) equipped with a Zorbax NH2 column (250 × 4.6 mm; Phenomenex, Torrance, CA) and a UV detector set at 365 nm. HPLC separations were conducted at a flow rate of 1.5 mL/min flow rate under gradient conditions using 80% methanol (vol/vol) as buffer A and 0.3 M CH₃COONa in 64% methanol as buffer B. The following gradient was used: 0 to 6.5 minutes, 20% (buffer) B; 6.5 to 20 minutes, 99% B; 20 to 25 minutes, 99% B; and 25 to 28 minutes, 20% B. Levels of sulfur-containing compounds were determined by comparison of the peak area of a particular compound with the peak area of the internal standard, γ-Glu-Glu. There was only 2% to 3% oxidation of GSH or Cys due to the specific sample manipulation described in this method. Derivatized samples were stable for at least 3 months at −70°C or at least for 1 week if kept at +4°C. 

Protein determinations were performed for each lens WS fraction by using the bicinchoninic acid (BCA) method ³⁷ according to the procedure described by the manufacturer (Pierce, Rockford, IL). Hydrogen peroxide was measured by the xylenol orange method. ³⁸  

Superoxide radical determinations were performed by the SOD-inhibitable cytochrome c reduction method of McCord and Fridovich. ³⁴ The generation of O₂ ⁻ was calculated using Δε_{ferro-ferricytochrome C} = 2.1 × 10⁴ M/cm. ³⁹  

Singlet oxygen production was monitored by the decrease in absorbance at λ = 440 nm, as described by Kraljic and El Moshni. ⁴⁰ A standard curve for this assay was derived from performing the RNO bleaching assay in the presence of known amounts of a temperature-sensitive singlet oxygen generator, 3-(4-methyl-1-naphtyl)-propionic acid endoperoxide, as described elsewhere. ¹⁵ All the assays were performed in triplicate for both irradiated and dark control samples. 

Results

In our preliminary experiments, the whole aged human lenses were separated into cortex and nucleus fractions, and the specific enzymatic activities of the GR, GPx, G3PD, and SOD in these fractions were measured. Table 1 shows that the specific activity of GR and G3PD in the nucleus of the human lens were significantly lower than in the cortex. There was almost a twofold decrease of GR specific activity, when the enzymes’ activity in the nucleus was compared with that in the lens cortex. This situation was even more dramatic for G3PD, with which an almost eightfold decrease in specific activity was observed between the nucleus and cortex in the same lenses (Table 1) . 

At the same time, SOD and GPx in the aged human lenses did not show a significant decrease in specific activity between the cortex and nucleus. The significant differences found in the specific activities of GR and G3PD along the visual axis suggested that with time the UVA-induced generation of ROS by lenticular sensitizers should be considered as a possible cause of their inactivation. Also, it is known that these two particular enzymes are susceptible to inactivation by singlet oxygen species. ^{41 42}  

To see the effect of irradiation, aged human lenses were placed in a quartz cuvette filled with artificial aqueous humor and irradiated with broad-band light from 320 to 475 nm (total irradiance of 0.925 kJ/cm² per hour) under aerobic conditions. On average, light absorption by a masked lens, which allowed light to pass through the lens only, gave 400 J/cm² per hour UVA light absorbed per lens. Intact human lenses (ages, 55–75 years) irradiated under these conditions for 12 hours showed (Fig. 1) a time-dependent loss in the specific activity of GR from 9.74 ± 1.6 to 0.21 ± 0.11 mU/mg lens protein (95% loss). Similarly, G3PD decreased from 79.85 ± 7.83 to 47.8 ± 0.34 mU/mg protein (40% loss). SOD was not affected during the entire duration of photolysis, and only a slight loss of GPx-specific activity was observed (Figs. 1B 1D) after 12 hours of irradiation. As can be seen, the susceptibility of individual enzymes toward UVA photolysis was markedly different. GR was the enzyme most inactivated by UVA light. Its specific activity was reduced to 30% of the original level after 1 hour of irradiation (7.15 ± 1.15 vs. 2.19 ± 0.45 mU/mg protein, P < 0.001) and the specific enzymatic activity was almost completely lost in the lens during the second hour of irradiation (8.3 ± 0.39 vs. 0.39 ± 0.05 mU/mg protein; P < 0.0001). Within the first hour of photolysis, the activity of G3PD within the lens decreased 24% (78.14 ± 1.93 vs. 59.32 ± 3.99 mU/mg protein; P < 0.001). During the next 11 hours of irradiation, however, the loss in specific activity leveled out, accounting for only a 16% additional loss in the enzymatic activity. GPx and SOD did not show any statistically significant loss of their specific activities during the entire duration of irradiation. No lens opacification was observed at any time during the 12-hour irradiation experiment of the aged human lenses. 

A series of experiments were conducted to determine whether enzyme inactivation could also be observed under lower light fluence. Whole human lenses were irradiated with an equal dose of UVA for either 1 hour or for 8 hours with a 12.5% ndf or for 23.5 hours with a 4.25% ndf, delivering 925 J of light/cm² in each case. Paired lenses were individually homogenized and the specific activities of four enzymes were measured (Fig. 2) . 

Only GR of all the enzymes assayed was significantly affected by UVA irradiation under diminished light conditions. According to our results GR activity was equally affected under lower intensity light conditions, provided the same dose of light was used. We observed that its specific activity decreased by 67.5% during 1 hour without neutral density filter, 67% after 8 hours of irradiation with the 12.5% ndf and 63% after 23.5 hours of photolysis with the 4.25% ndf. These data argue that UVA deactivation of the enzyme is light-dose dependent. Inactivation of the specific G3PD activity after the original 1-hour irradiation led to only a 10.3% loss of specific activity, whereas 8- and 23.5-hour photolysis with 12.5% and 4.25% ndf caused 20% and 23% loss in the enzyme specific activity, respectively (Fig. 2) . At the same time, prolonged periods of irradiation with diminished light intensity and the same dose of light had almost no effect on loss of the specific activities of SOD (no loss in 1 hour of irradiation versus 10% loss with 4.25 ndf and 24 hours of irradiation) and GPx (5.5% loss with no filter versus 4.5% loss with a 4.25% ndf) in human lens. 

To determine whether lens UVA-responsive sensitizers were involved in the lens enzyme inactivation by UVA light, fetal calf lenses, which should contain no age-related chromophores, were irradiated under the same conditions that had been used for human lenses for either 1 hour or 12 hours. The results (Fig. 3) show that calf lens enzymes, with the exception of G3PD, responded similarly to the human lens enzymes during the first hour of irradiation. During 1 hour of irradiation, only GR showed a decrease in its specific activity (0.32 ± 0.01 vs. 0.19 ± 0.03 mU/mg; 40% loss; P < 0.001). There was no appreciable inactivation of SOD, G3PD, and GPx during 1 hour of photolysis of fetal calf lenses. Unlike the human lenses, a 12-hour irradiation of fetal calf lenses caused a statistically significant decrease in the corresponding specific activities of GPx (30.8%) and G3PD (31.3%). Similar to the enzyme from the human lenses, GR lost 70.5% of its specific activity and SOD from fetal calf lenses lost no activity after 12 hours of irradiation (Figs. 3A 3D) . If it is assumed that similar photolytic mechanisms are involved in the inactivation of lenticular enzymes in human and fetal calf lenses, these data argue that lenticular chromophores do not participate in the inactivation of GR by UVA light. In contrast, inactivation of G3PD in human lenses by UVA light is probably affected by the presence of age-related chromophores. 

We have already shown that the UVA-absorbing chromophores covalently bound to the insoluble lens proteins behave as UVA-responsive sensitizers, ^{14 15} capable of producing significant amounts of ROS during UVA photolysis under aerobic conditions. Jedziniak ⁴¹ has already shown that G3PD is sensitive to the actions of ROS, particularly of singlet oxygen. Our next series of experiments was designed to establish whether oxidative stress occurs within the lens once it is submitted to UVA irradiation. 

Under the aerobic conditions of our experiments, we observed that the human lens took up oxygen from AAH. Our measurement of lenticular oxygen tension in lenses freshly excised from globes 1 to 2 days after death showed a level of approximately 10 μM (data not shown). Incubation of human lens for 1 hour or longer led to an elevation of the lenticular O₂ level from 0.1 to 0.16 mM (in Fig. 4 , dark control). 

These oxygen concentrations are only 1.5- to 2.5-fold lower than oxygen levels in air-saturated solutions. ⁴³ According to our data, 1 or 2 hours of UVA photolysis of the human lens led to a complete consumption of any oxygen absorbed by the lens, as can be clearly seen in Figure 4 (lower traces). These data point out that a human lens submitted to UVA photolysis consumes oxygen with the likely generation of ROS within the lens. This claim was substantiated by the findings that the level of reduced GSH in the cortex of the UVA-photolyzed lenses decreased to 58% of its original value (1362.3 ± 340.7 vs. 786.7 ± 95.1 nmol/lens, P < 0.02; Fig. 5 ). 

Such a decrease in the level of GSH was consistent with an almost 50% increase in the levels of oxidized GSSG in the same part of the lens (562.5 ± 187.7 vs. 1139.0 ± 117.4 nmol/lens, P < 0.002; Fig. 5 ). Consistent with these results were data on ascorbic acid levels, which showed an almost 30% decrease of this antioxidant during 1 hour of irradiation (data not shown). 

Because the data in Figures 4 and 5 clearly indicate that UVA irradiation promoted oxidative stress within the lens, we attempted to establish the major ROS generated by UVA irradiation of soluble proteins from aged human lenses. Table 2 shows that 1 hour of UVA photolysis of a dialyzed WS fraction from aged human lenses in the presence of oxygen produced mostly singlet oxygen (94.5% of all ROS), whereas the WS fraction from calf fetal lenses generated only insignificant amounts of the superoxide radical, hydrogen peroxide and no singlet oxygen under the same experimental conditions. These data point out that the WS fraction contains protein-bound chromophores that behave as UVA-responsive sensitizers and can mediate UVA-dependent inactivation of the specific lenticular enzymes by producing, ROS under the aerobic conditions of our experiment. 

The differential susceptibility of all four enzymes under investigation to ROS and UVA light was investigated. A dialyzed WS fraction from 55- to 75-year-old lenses was irradiated for 1 hour, either in the presence or absence of oxygen (Fig. 6) . These data reveal that GR photolysis by UVA light was essentially unaffected by the presence of molecular oxygen in solution. An almost 90% to 95% loss of GR-specific activity occurred within 1 hour of irradiation in both anaerobic and aerobic conditions (Fig 6) . One hour of UVA photolysis of the WS fraction under aerobic conditions, however, caused a 74% inactivation of G3PD-specific activity, compared with a 24% loss under anaerobic conditions. These data argue that UVA-mediated G3PD inactivation was largely oxygen-mediated, most probably due to the generation of ¹O₂ (Table 2) . 

No statistically significant losses of the specific enzymatic activities of SOD and GPx were observed during 1 hour of UVA photolysis of the WS fraction in anaerobic conditions (Fig. 6) , which rules out the possibility that direct UVA photolysis plays any role in direct inactivation of these two enzymes with age. Under aerobic conditions, although SOD specific enzymatic activity remained almost constant during the course of 1 hour of UVA irradiation, GPx in the WS fraction showed a small statistically significant decrease (7.3% loss) in specific activity after 1 hour of irradiation (Fig. 6) . 

Because singlet oxygen within the lens can be detoxified by the presence of high concentrations of both GSH ⁴⁴ (Fig. 5) and ascorbic acid, ⁴⁵ these antioxidants should protect the lens from enzyme inactivation by UVA light-mediated ¹O₂. On the contrary, the inactivation of the lenticular enzymes (GR and G3PD) in the whole lenses by UVA light (Figs. 1A 1C) occurred in the presence of relatively high residual concentrations of both GSH and ascorbic acid (Fig. 5) . To mimic conditions close to physiological, we conducted experiments with UVA light of diminished intensity (38 J/cm² per hour for 24 hours, totaling a dose of 0.925 kJ/cm² per hour). The irradiation of undialyzed WS fraction from human aged lens, which contained all antioxidants usually present in the lens, should have provided a uniform access for these molecules to the sites of ROS generation. Either nondialyzed or exhaustively dialyzed homogenates from old human lenses (both containing 20 mg/mL of lens protein) were irradiated under aerobic conditions for 24 hours with 4.25% ndf. All four enzymes were studied in this system (Fig. 7) . Our results revealed that under aerobic conditions, lenticular GR almost completely inactivated (96% −LMW and 86% +LMW) after 24 hours of irradiation either in the presence or absence of several low-molecular-weight (LMW) compounds. These results are in a good agreement with the results of the 24-hour irradiation of whole human lenses shown in Figures 1 and 2 . G3PD, in contrast, was greatly influenced by the presence of LMW substances. Only 28% of the specific enzymatic activity was lost during the first 16 hours of irradiation in the presence of LMW substances, whereas the absence of LMW compounds led to complete abolishment of the activity over the same duration. These data correlate with the results of 1-hour irradiation of aged human lens WS fraction under aerobic conditions (Fig. 6) , which suggests that inactivation of lenticular G3PD in the lens by UVA is mediated primarily by the generation of ROS. Similar to the results presented in Figure 6 , SOD-specific activity was not influenced by the presence of LMW compounds during the course of irradiation (Fig. 7) . Specific enzymatic activity of GPx in the absence of LMW compounds (Fig. 7A) showed a slight, statistically significant decrease in its activity (∼10% loss), which correlated well with the result in Figure 6 (aerobic conditions of irradiation). This effect was almost completely prevented by the presence of LMW compounds in the irradiated protein mixture (∼ 4% loss). 

To fully establish that human lens UVA sensitizers play no role in inactivation of GR against ROS generated during UVA photolysis, experiments similar to the ones described in Figure 6 were performed with a pure preparation of GR from baker’s yeast. Also, GSH and/or ascorbic acid were added to this system to see whether these reducing, ROS-scavenging agents had any protective ability during UVA photolysis. Purified GR was rapidly inactivated by UVA light, indicating that lens sensitizers were not required for its inactivation (compare Figs. 6 and 8 ). GSH under both anaerobic and aerobic conditions was slightly protective toward GR during photolysis (10%–15% protection), whereas ascorbic acid had no effect on the rate of GR inactivation by UVA light. These data again argue that ROS are not a significant factor in the inactivation of GR during human lens photolysis. 

Discussion

The human lens is a closed system with limited capability to repair or regenerate itself. ⁴⁶ Once damage has occurred to its constituent proteins (e.g., oxidation, modification, denaturation, or aggregation), there is little if any possibility that these effects can be reversed. ⁴⁴ One clear example of such deterioration is a diminution in specific activities of most of the protective ROS-detoxifying enzymes and the major glycolytic enzymes in the nucleus of mammals lens during both aging and cataractogenesis. ^{24 30 47 48 49 50}  

The UVR portion of sunlight has been shown to inactivate lens enzymes. ^{50 51} Direct UVR photolysis ^{50 51} and UVA-generated ROS by UVR-photolysis of lenticular sensitizers are among the major factors that may cause inactivation of enzymes. ^{14 15 50} Which part of UVR portion of light, UVB or UVA, is the most harmful to a human lens is still a matter of debate. ^{1 2 5 52 53} However, a recent report by Dillon et al. ⁵² has already established that due to the spectral characteristics of the cornea and aqueous humor of the human eye, the total UVB radiation that reaches the lens represents at most 2% of UVR. Only UVB wavelengths from 300 to 320 nm can reach the surface of the lens. ⁵² The energy of UVB photons in this range is not much different from the energy of UVA photons (e.g., E₃₀₀ = 95.3 kcal/mol vs. E₃₅₀ = 81.7 kcal/mol), ^{53 54} and, according to Sliney, ²⁰ a total dose of only 50 to 200 nW/cm² of UVB light is absorbed by the human lens. According to Lofgren and Soderberg ²¹ all UVB irradiation that impinges on the rat lens is absorbed within the first 0.4 to 0.5 mm of the lens and does not reach the lens nucleus. ²¹ On the contrary, in primate lenses the total UVA light dose could range from 1.1 to 9.7 mW/cm², ¹⁶ all of which is absorbed by lenticular chromophores and can penetrate to the lens nucleus. ⁵³ These data, combined with results from our laboratory that showed that aged lenses contain UVA responsive, ROS-producing sensitizers ^{14 15} (Table 2) , imply that the UVA portion of the sunlight may be involved in age-dependent lens protein modifications and particularly in the loss of the specific human lens enzymes. This hypothesis was studied on four lenticular enzymes—GR, G3PD, SOD, and GPx—the specific activity of which was shown to decrease with age and during cataractogenesis. ^{24 30 48 49} In addition, according to the literature, these four enzymes have already been shown to be inactivated, by direct UVA photolysis, ^{42 55} oxidation by ROS, ^{41 56 57 58} or both UVA photolysis and ROS oxidation. ⁴²  

Our experiments with UVA irradiation of intact human lenses revealed that two lenticular enzymes (GR and G3PD) of the four under investigation were inactivated by UVA photolysis (Figs. 1A 1C) . GR in the intact human lens was inactivated 70% by 1 hour of irradiation, and this inactivation was dose dependent (Fig. 2) . That a 40% loss of GR activity was also seen in fetal calf lenses, which lack any age-related chromophores, argues that GR is vulnerable to a direct UVA photolysis. To verify this fact, we conducted experiments in which the irradiation of a pure preparation of GR from baker’s yeast was performed (Fig. 8) . 

We have determined that purified GR was rapidly inactivated by UVA light, and at practically the same rate under both aerobic and anaerobic conditions (Fig. 8) . Therefore, although this enzyme can be inactivated by singlet oxygen ⁴² and hydroxyl radical ⁵⁶ in vitro, these mechanisms are apparently not significant in GR loss in the mammalian lens (compare Figs. 6 and 7 and Table 2 ). Also, ascorbate and GSH were ineffective in UVA-mediated inactivation (Figs. 7 8) . 

Our observations on inactivation of GR by UVA light are in a good agreement with the findings reported by Fuchs et al. ⁵⁵ on the inactivation of GR in mouse skin after UVA irradiation of the animals. They hypothesized that GR inactivation was due to the self-sensitizing properties of flavine adenine dinucleotide (FAD), which is the prosthetic group in GR. ⁵⁹ This suggestion is well justified, because FAD is the only UVA-absorbing molecule in GR and is known to be a UVA-light–responsive sensitizer. ⁶⁰ The absorption of UVA light causes FAD to undergo splitting, which produces mostly lumichrome and FMN (unpublished results). Light-excited lumichrome or flavin mononucleotide can photo-oxidize numerous compounds, including amino acids ⁶⁰ or cause the binding of FAD (and/or its degradation products) to the specific amino acid residues at or near the reaction center of GR. In both cases, the enzyme is irreparably inactivated by UVA light. 

GR is one of the major ROS-detoxification enzymes in the human lens and along with ascorbic acid is responsible for maintaining high levels of reduced GSH in the lens. The level of this enzyme falls almost 2- to 4.5-fold along the human lens visual axis in aged lenses according to our data (Table 1) and the results from the other laboratories. ^{30 49} This decline in enzymatic activity with age in the human correlates directly with a loss of free GSH ^{44 61} and the elevation in the levels of protein-bound GSH. ⁴⁴ Because the lens is chronically exposed to UVA light, UVA photolysis of GR may explain the existence of a GR activity gradient along the human lens visual axis, not only in the human, but also in other nonprimate species. ⁶² Oxidation may therefore occur because of GR loss, a decrease of de novo GSH-synthesizing ability, ⁶¹ and the development of a barrier around the nucleus that prevents free GSH from entering. ⁶³  

Our results on the loss of G3PD-specific activity during UVA photolysis of the intact human lens (Figs. 1 2 3 6 7) along with the data published by Dovrat et al., ⁴⁸ support the possibility of oxidation of an essential SH-group in this enzyme in the lens nucleus. ⁶⁴ Similar to GR, G3PD-specific activity was almost eight times lower in the nucleus than in the cortex in aged human lenses (Table 1) . Under the aerobic conditions of our experiments, G3PD from intact human aged lenses was inactivated by 24% within 1 hour of irradiation (Fig. 1C) . Because we did not observe this effect on G3PD in fetal calf lenses photolyzed for the same period (Fig. 3C) , it could be argued that this difference is the result of the action of the UVA-excited human lens sensitizers. Furthermore, the data in Figure 4 argue that traces of molecular oxygen present in the human lens were consumed because of UVA photolysis of the lens, leading to the oxidation of GSH and the subsequent formation of GSSG (Fig. 5) . We have already shown that UVA photolysis of aged lens water-insoluble (WI) lens proteins under conditions described in this report leads to generation of ROS. ^{14 15} The major species generated is singlet oxygen, ¹⁴ which accounts for more than 95% ROS generated, and it was shown that 1.0 mg of WI protein from aged human lenses produced approximately 1.2 mM singlet oxygen during 1 hour of UVA irradiation. In addition the same amount of WS proteins from human lenses from elderly donors (65–75 years of age) was able to produce on average a total of 0.55 mM singlet oxygen under aerobic conditions within 1 hour of irradiation (Table 2) . Jedzinak et al. ⁵⁷ and Jedziniak ⁴¹ have already shown that among the mechanisms that could account for inactivation of G3PD, ⁴¹ the oxidation of SH- and histidyl groups in the reaction center of G3PD by ¹O₂ is the most probable one. Pure bovine G3PD, after interaction with singlet oxygen was shown to produce a pattern of bands on an SDS-polyacrylamide gel similar to the one produced by native human G3PD extracted from aged human lens. ⁴¹ These results are reinforced even more by the results of Armstrong and Buchanan, ⁵⁸ who showed that active site acyl-binding groups of the enzyme are more susceptible than any other thiols to oxidation by ROS and by the findings of Cartier et al. ⁶⁵ that His₁₇₆ in the reaction center of the molecule is essential to significantly increase the reactivity of Cys₁₄₉ at neutral pH. 

Although the concentration of molecular oxygen within the freshly extracted lenses from human eyes 1 day after death (55–75 years of age) was low and lay within the range 10 to 20 μM according to our studies (Linetsky MD et al., unpublished results, 2002) and to the studies of Kwan et al., ⁶⁶ the data in Table 2 and Figure 4 imply that it is feasible that inactivation of G3PD by UVA photolysis may have been caused by the action of singlet oxygen. This idea is supported by the fact that irradiation of a dialyzed homogenate from human lens under aerobic conditions led to almost complete enzyme inactivation (Fig. 6 ; aerobic conditions). The ability of LMW substances (presumably antioxidants) to protect the enzyme during the first 16 hours of the experiment seems to provide further support for a role for singlet oxygen–mediated deactivation of enzyme (Fig. 7A) . 

Unlike GR and G3PD, GPx within the aged human lenses was only marginally affected by 1 hour of UVA photolysis (Fig. 1B) . The data in Figure 1B point out that the enzyme was tolerant to the levels of ROS generated by UVA-responsive sensitizers (Fig. 4 and Table 2 ). In addition, it was not photolyzed by direct UVA light, according to the data in Figure 3B . This enzyme, however, has been shown to be susceptible to inactivation by singlet oxygen, hydrogen peroxide, and hydroxyl radical in vitro in the presence of GSH. ⁵⁴ Our experiments with the dialyzed and UVA-irradiated WS fraction from human lens show that the levels of ROS (mostly ¹O₂) generated in 1 hour of photolysis marginally inactivated the enzyme (∼10%; see Fig. 6 , aerobic conditions). Once the lenticular enzyme was irradiated in the presence of LMW compounds (that presumably contained GSH and ascorbic acid) this effect was diminished (Fig. 7B) . 

This lack of UVA inactivation was substantiated by our data on the activity of GPx in the human aged lens that show absence of a statistically significant difference between the GPx-specific activity in the cortex and in the nucleus (Table 1) . The data from Rao et al. ⁶⁷ show that lens aging does not bring about a decrease in the level of this enzyme in the lens. This situation changes in cataractous lenses, where GPx activity decreases dramatically along the visual axis, with practically no activity in the nucleus of type IV cataractous lenses. ⁶⁸ A comparison of the data of Fecondo and Augusteyn ⁶⁸ with the data on singlet oxygen–generating ability of the WI lenticular proteins from this laboratory show a direct relationship between the decrease in the specific activity of GPx and the ability of these proteins to produce ¹O₂. ⁶⁹ Such action of ROS on mammalian GPx can make it thermolabile ⁷⁰ and susceptible to degradation by lenticular proteases. ^{71 72} As a direct result of chronic exposure of old and cataractous lenses to UVA light, all these factors may lead to an inactivation of this enzyme. 

Similar to data for GPx, our data on SOD activity in the both irradiated human and calf lenses showed that there was no UVA-dependent loss of this activity during 12 hours of irradiation (Figs. 1 3) . Even in dialyzed solution of the WS proteins from human aged lens under aerobic condition and acute UVA photolysis with high UVA light fluence, this enzyme was resistant toward an ¹O₂ oxidation during a 1-hour experiment. The lack of UVA inactivation of the enzyme under anaerobic conditions argues that it is not immediately susceptible to direct UVA photolysis (Fig. 6 , anaerobic conditions; Fig. 7 , aerobic conditions). Because it is known that the enzyme can be inactivated by the action of singlet oxygen, ⁴² it implies that ROS levels within the photolyzed human aged lens were probably too low to cause any appreciable enzyme inactivation. 

In contrast, levels of this enzyme in the aged and cataractous lenses decreased and according to the immunologic studies conducted by Scharf and Dovrat ⁷³ and Scharf et al. ⁷⁴ the catalytically inactive molecules of SOD are found in aged and cataractous lenses. It has been shown that SOD becomes heat sensitive with age and during cataractogenesis, which indicates that some conformational changes or specific amino group modifications occur within the enzyme. ⁷⁵ One of the known SOD modifications that lead to the enzyme inactivation and thermal instability is glycation. ⁷⁶ The glycation leads to modification of specific Lys in the reaction center of enzyme, which renders the enzyme inoperable. ⁷⁶  

In the current work, aged human lenses were irradiated with an amount of light that is at least 200 to 400 times higher than the human lens is exposed to during 1 hour on a sunny day. ⁷⁷ Under the conditions of our experiment, such high light fluence may be enough to facilitate high ROS output from human lens UVA sensitizers with low ROS quantum yield. Our experiments showed that even a 2-hour irradiation of WS lens proteins with much lower intensity UVA light (38 J/cm² per hour; 11 mW/cm²) inactivated GR by almost 20% and G3PD by 10% (Fig. 7) . According to Zigman ⁷⁷ this amount of light is roughly equivalent to 20 to 40 hours of lens irradiation with ambient light. Provided that UVA light GR and G3PD inactivation in the lens is dose-dependent (Fig. 2) and that there is almost no protein turnover in the human lens, our data may explain the age-dependent loss in the activity of these two enzymes as a UVA-light–dependent event. ^{30 48 49 57}  

 Supported by National Eye Institute Grant EY02035 and by a departmental grant from Research to Prevent Blindness, Inc.

 Submitted for publication June 18, 2002; revised August 15, 2002; accepted August 20, 2002.

 Disclosure: M. Linetsky, None; V.G. Chemoganskiy, None; F. Hu, None; B.J. Ortwerth, None

 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 4.

 

Oxygen levels in the irradiated (A, B, lower traces) and nonirradiated (A, B, upper traces) human lenses after 1 (A) and 2 (B) hours of UVA photolysis or incubation under aerobic conditions (dark control). Matched pairs of freshly enucleated lenses were used in the experiments.

Table 1.

 

Specific Activities of GR, GPx, G3PD and SOD in the Cortex and in the Nucleus of Aged Human Lenses (Ages 55–75)

Enzyme*	Specific Enzymatic Activity (mU), †
		Cortex	Nucleus
GR	10.62 ± 1.16	4.91 ± 0.98
GPx	2.24 ± 0.41	1.73 ± 0.43
G3PD	89.05 ± 12.08	9.46 ± 4.02
SOD	487 ± 140	516 ± 77

Figure 1.

 

Specific activities of GR (A), GPx (B), G3PD (C), and SOD (D) during 12 hours of UVA photolysis of the aged human lenses (ages 55–75 years). The data represent an average of the measurements performed on two pairs of lenses used for each time point. The specific activity of each enzyme in the lens homogenates was measured in triplicate.

Figure 2.

 

Specific activities of GR, GPx, G3PD, and SOD of intact human lenses irradiated with an equal dose of UVA light (925 J/cm² per hour) for 1, 8 (12.5% ndf), and 23.5 (4.25% ndf) hours. The data represent an average of the measurements performed on four pair of lenses. The specific activity of each enzyme in the lens homogenates was measured in triplicate.

Figure 3.

 

Specific activities of GR (A), GPx (B), G3PD (C), and SOD (D) after 1 and 12 hours of UVA photolysis of fetal calf lenses. The data represent an average of the measurements performed on two pairs of lenses used for each time point. The specific activity of each enzyme in the lens homogenates was measured in triplicate.

Figure 5.

 

Levels of lenticular sulfur-containing non-protein compounds in human aged lenses measured after 1 hour of UVA photolysis or incubation under aerobic conditions. The data represent an average of the measurements performed on four freshly enucleated and matched pairs of lenses. Every supernatant obtained from the nucleus (n) and cortex (c) of either irradiated or nonirradiated lenses was measured in duplicate to determine the levels of unbound GSH (oxidized and reduced), cystiene, and cystine.

Table 2.

 

Generation of Reactive Oxygen Species by WS Fractions from Human and Calf Lenses during 1 Hour of UVA Photolysis

Preparation	1O2 (μmol/mg protein/h)			O2 − (nmol/mg protein/h)			H2O2 (nmol/mg protein/h)
		UVA	Dark Control	UVA	Dark Control	UVA	Dark Control
Human lenses (ages: 64–74); (n = 3)	0.55 ± 0.1	0.0	19.8 ± 0.4	1.15	7.7 ± 1.4	0.2 ± 0.1
Fetal calf lenses (n = 3)	0.0	0.0	2.0 ± 0.2	0	0.8 ± 0.1	0.2 ± 0.1

Figure 6.

 

Specific enzymatic activities of SOD, GPx, G3PD, and SOD in exhaustively dialyzed aged human lenses WS fractions (lenses aged 55–75 years) under anaerobic (left) and aerobic (right) conditions during 1 hour of UVA photolysis (925 J/cm² per hour). Each time point represents the average of two independent experiments. Each enzymatic activity was assayed in triplicate.

Figure 7.

 

Specific enzymatic activities of GR, G3PD, GPx, and SOD in dialyzed (A) and nondialyzed (B) WS fractions from aged human lenses (ages 55–75 years) that were UVA light irradiated for 24 hours with a 4.25% ndf under aerobic conditions (38.5 J/cm² per hour). Each time point represents the average of two independent experiments, and each enzymatic activity was assayed in triplicate.

Figure 8.

 

Time-course of inactivation of baker’s yeast GR by UVA light in the presence and/or absence of GSH and ascorbic acid in anaerobic (A) or aerobic conditions (B). Each time point represents an average of three independent measurements.