Regulation of GSH in αA-Expressing Human Lens Epithelial Cell Lines and in αA Knockout Mouse Lenses

Abstract

purpose. To study the mechanism of regulation of GSH in HLE-B3 cells expressingα A-crystallin (αA) and in αA knockout mouse lenses.

methods. GSH levels and maximal rates of GSH synthesis were measured in immortalized, αA-transfected HLE-B3 cells containing varying amounts of αA. The mRNA and protein for the rate-limiting enzyme for GSH synthesis, γ-glutamylcysteine synthetase (GCS), were also determined in αA- and mock-transfected cells by Northern blot analysis and Western blot analysis of heavy (GCS-HS) and light (GCS-LS) subunits. The effect of absence of αA and αB on lens GSH concentrations was evaluated in whole lenses of αA knockout and αB knockout mice as a function of age. GCS-HS mRNA and protein were determined in young, precataractous and cataractous αA knockout lenses.

results. GSH levels were significantly higher in HLE-B3 cells expressing αA- compared with mock-transfected cells and were correlated positively with αA content. Mean rate of GSH synthesis was also higher inα A-expressing cells than in mock controls (0.84 vs. 0.61 nmol · min⁻¹ per mg protein, respectively). GCS-HS mRNA and GCS-LS mRNA were approximately twofold higher in αA-expressing cells, whereas the heavy and light GCS subunit proteins increased by 80% to 100%. In αA(−/−) mouse lenses, GSH level was not different from that of wild type up to 2 months from birth, after which it dropped to ∼50% of controls. On the other hand, GCS-HS and GCS-LS proteins showed a significant decrease before cataract formation as early as 15 days after birth. GSH level in cataract-free αB(−/−) lenses was similar to that of wild type for up to 14 months.

conclusions. Expression of αA caused an increase in cellular GSH, in part, because of an increase in mRNA and protein of both GCS subunits. GSH levels decreased with increasing age in cataractous αA(−/−) lenses but not in the noncataractous αB(−/−) lenses. It is suggested that neonatal precataractous lenses (with normal GSH and decreased GCS) may maintain their GSH level by other compensatory mechanisms such as increased GSH transport.

Expression of several members of different heat shock protein (HSP) families confers increased thermoresistance in various cell systems. ^{1 2} For the ubiquitous 15- to 30-kDa range, small HSPs (sHSPs), a similar protection with expression has been reported. ^{3 4} The αA- and αB-crystallins (αA andα B, respectively) fall into the category of sHSPs in that there is close similarity between the C terminus parts ofα -crystallins and HSPs. ^{5 6} Together the 20-kDaα A and αB subunits form soluble complexes of up to 800 kDa, constituting one of the most abundant protein components (>50%) in the vertebrate eye lens. ⁵ The two polypeptides are ∼60% identical in amino acid sequence and are encoded by separate, unlinked genes. ⁵ α-Crystallins have been shown to be associated with a variety of cytoskeletal proteins, including actin, vimentin, desmin, and lens beaded filament proteins. ^{7 8} Like other sHSPs, αA and αB can act as molecular chaperones in vitro, preventing aggregation induced by heat and other stresses. ⁹ Extralenticular expression of both forms ofα -crystallins has been reported. ^{10 11 12 13} αA and αB are expressed at low levels in lens epithelial cells, and their expression increases dramatically during differentiation to lens fibers. ⁵ Because of their extralenticular expression, autokinase activity, phosphorylation patterns, link with neurodegenerative diseases, and protective activity from heat shock and other stress, a generalized cellular function for α-crystallins other than their well-known role in refraction has been suggested. ⁵  

Members of the sHSP family are also important in cell growth and differentiation. ^{14 15 16 17 18} For example, HSP27 protects cells during stress by preserving actin microfilaments and preventing apoptotic cell death. ^{19 20} Recent work by Mehlen et al. ²¹ showed that expression of sHSPs including αB inhibited several downstream effects arising from TNFα-mediated reactive oxygen species increment, NF-κB activation, lipid peroxidation, and protein oxidation. The expression also was associated with increased intracellular GSH levels in L929 and NIH 3T3-ras cells. ²¹ The authors suggested that the sHSP-expression–mediated increase in GSH is essential for the protective activity of these proteins against oxidant-induced cell death. In recent studies, we have shown that expression ofα A-crystallin in HLE-B3 cells renders these cells resistant to cell death from UVA exposure. ¹⁴ Whether GSH plays a role in this protection is not known. 

Brady et al. ²² and Wawrousek and Brady ²³ have generated mice with targeted disruption of the mouse αA and αB genes, respectively. Interestingly, αA(−/−) lenses developed cataract several weeks after birth, whereas αB(−/−) lenses were devoid of cataract until their death due to unrelated causes in a year or more. We have found that primary lens epithelial cells isolated fromα A(−/−) lenses were more susceptible to UVA-induced oxidant stress and cell death compared with cells isolated from wild-type mice. ¹⁴  

One of the important determinants of cellular GSH is its biosynthesis from precursors. ²⁴ The synthesis of GSH from its constituent amino acids, l-glutamate, l-cysteine, and l-glycine, involves two ATP-requiring enzymatic steps. The first step of GSH biosynthesis is rate limiting and is catalyzed by γ-glutamylcysteine synthetase (GCS). GCS is composed of a heavy (GCS-HS, M _r ∼73,000) and a light (GCS-LS, M _r ∼30,000) subunit, which are encoded by different genes in both rat and in humans. ^{25 26} Although the heavy subunit is active catalytically, it has a high K _m for glutamate and a lower K _i for GSH compared with holoenzyme. ^{27 28} Thus, the light subunit plays an important regulatory role for the overall function of the enzyme and allows the holoenzyme to be catalytically more efficient and subject to lesser inhibition by GSH than the heavy subunit alone. The low affinity of the heavy subunit for glutamate and the high feedback inhibition exerted by GSH suggest that the heavy subunit alone is not likely to be active physiologically. Regulation of GCS, the rate-limiting enzyme of synthesis, has been a subject of intense study in several cell types, particularly in hepatocytes. Several studies suggest that the two subunits of GCS appear to be differentially regulated, depending on the experimental conditions. ^{29 30 31 32}  

Although GCS has been purified from the lens and a decrease in its activity is shown in aging and cataractogenesis, ^{33 34} we are not aware of any studies on GCS gene regulation at the molecular level in the lens. To examine the effect of αA expression on GSH levels, we studied the relationship of αA to GSH level and GCS expression in extended life span human lens epithelial cells (HLE-B3). As an additional model, we have used lenses from αA knockout mice to study GSH metabolism in relation to αA expression. 

The results show that αA-expressing HLE-B3 cells exhibit increased GSH associated with upregulation of the both heavy and light subunits of GCS and that GCS is downregulated in neonatal αA(−/−) lenses, with maintenance of normal GSH until the development of cataract. 

Methods

All procedures used in these studies adhered to the tenets of the Declaration of Helsinki and were in accordance with National Institutes of Health guidelines and the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. 

Cultured Cells

Human lens epithelial cells with extended life span (HLE-B3 cells) have been described previously. ³⁵ They were derived from an infant human lens epithelial culture by Ad12-SV40 hybrid virus infection and propagated through at least 11 passages. After 11 passages, HLE-B3 cells ceased to produce αA. At this stage, αA cDNA was reintroduced into these cells by cDNA transfection, and stably transfected cell lines with different amounts of αA were generated as described previously. ¹⁴ Cultures were examined by quantitative Western blot analysis for the expression of αA and compared with mock-transfected cells (vector without αA insert). The mock-transfected cells did not express any αA as shown by immunoblot analysis. ¹⁴ Cultures of αA- and mock-transfected cells were passaged in an identical manner in 20% FBS-MEM as described ¹⁴ and were used at the same passage for all experiments. 

αA and αB Knockout Mice

αA(−/−) knock out 129SvJ mice have recently been generated by a targeted disruption of the mouse αA gene as described previously. ²² The lenses from these mice showed progressive lens opacification that became apparent several weeks after birth. ²² They contained insoluble αB in their fiber cells. The αB knockout lenses, which were also generated by the same laboratory, were found to be cataract free. ²³ Theα B(−/−) knockout mice were generated by standard embryonic stem cell manipulations. In the second gene-targeting vector, most of the coding region of the HSPB2 ³⁶ gene, the common HSPB2/αB promoter region, and αB coding sequence through the middle of the last exon (exon 3) are replaced with a PGK/neo selectable marker. ²³ Both the αB gene and the muscle-specific HSPB2 gene are effectively inactivated in these mice (data not shown). 

Whole lenses (with encapsulated epithelium) from mice bred and maintained at the Washington University, St. Louis animal facility ¹⁴ were isolated under RNase-free conditions. Lenses from different ages (0.5–14 months) were isolated from αA andα B knockout mice and their age-matched wild-type animals. 

GSH Levels and Rates of GSH Synthesis

GSH levels in HLE-B3 cells and in whole lenses were measured either by recycling assay or by a fluorescent technique that we described previously. ^{37 38} Maximal rates of GSH synthesis in mock- and αA-transfected cells were determined in predialyzed cytosol by the rate of formation of monochlorobimane adduct in the presence of excess amino acid precursors of GSH as described. ³⁸ The molecular form of GSH and thiols in cells and in whole lenses was verified by HPLC according to the method of Fariss and Reed. ³⁹  

Northern and Western Blot Analysis of GCS Subunits

From the several clones used for GSH determinations shown in Figure 1 , we picked low (∼0.1–0.2 ng/μg protein) and high (∼1.5–2.0 ng/μg protein) αA-expressing clones and their mock-transfected controls for Northern and Western blot quantitation of GCS subunits. Total RNA was isolated from HLE-B3 cells according to Chomczynski and Sachhi. ⁴⁰ Poly(A)⁺ RNA was isolated using oligo(dT) cellulose columns according to protocol provided by Life Technologies (Grand Island, NY). The RNA concentration was determined spectrophotometrically before use. In the case of total RNA, the integrity was checked by electrophoresis, with subsequent ethidium bromide stain. 

Northern hybridization analysis was performed on total RNA (20–30μ g) and poly(A)⁺ RNA using standard procedures. ⁴¹ The GCS-HS cDNA probe is comprised of a 390-bp fragment corresponding to nucleotides 79 to 468 of the published rat kidney sequence, ²⁵ and the GCS-light subunit (GCS-LS) cDNA probe is comprised of a 1.1-kb fragment corresponding to nucleotide 122 to 1232 of the published rat kidney sequence. ²⁶ Both were labeled with[ ³²P]dCTP using a random-primer kit (Primer-It Kit; Stratagene, La Jolla, CA). To ensure equal loading of RNA samples, the same membrane was hybridized with ³²P-labeled human β-actin (Clontech, Palo Alto, CA). 

Autoradiography and densitometry (Gel Documentation System; Scientific Technologies, Carlsbad, CA, and NIH Image software program) was used to quantitate relative RNA content. Results of Northern blot analysis were normalized to β-actin. 

A rabbit polyclonal antibody against a synthetic peptide (TVEDNMRKRRKEA), which corresponds to amino acid residues 119 to 131 of rat kidney GCS-HS was used for Western blot analysis of GCS-HS. ^{42 43} Both peptide synthesis and antibody generation were carried out by a commercial source (Multiple Peptide Systems, San Diego, CA). Cell extracts from mock-transfected andα A-expressing cells as well as tissue homogenates from αA(−/−) and wild-type mouse lenses were used in analysis. Mouse liver homogenate was used for comparison. Cell extracts or tissue homogenates containing 20 to 30 μg protein were solubilized in equal volumes of sample buffer consisting of 285 mM Tris, pH 6.8, 30% glycerol, 6% SDS, 1.5% mercaptoethanol, and 0.01% bromphenol blue, subjected to SDS 10% PAGE, and electrotransferred to nitrocellulose membranes with the use of Semidry Transfer cell (BioRad). The nitrocellulose membranes were subsequently subjected to the Amplified Alkaline Phosphatase Immun-blot Assay according to procedures described in the kit. The first antibody was rabbit antikidney GCS-HS peptide preimmune or postimmune serum diluted to 1:250 in Tris-buffered saline-Tween 20 (TBST). Equal protein loading was ensured by Coomassie Blue staining of gels after transblotting. Quantitation was performed by densitometric analysis. 

Western blot analysis of GCS-LS was performed in a similar manner to that of GCS-HS above. The polyclonal antibodies for the rat GCS light chain were kindly provided by Terrence Cavanagh (University of Washington at Seattle). ⁴⁴ The secondary antibody was horseradish peroxidase–conjugated goat anti-rabbit IgG (Boehringer Mannheim). The antibodies were found to react with the human and the mouse protein. 

Results

GSH Levels and Rates of GSH Synthesis in Mock-Transfected andα A-Expressing Clones

The HLE-B3 cells used for transfection had no detectable level ofα A. GSH levels were determined in mock-transfected HLE-B3 cells and several clones expressing varying amounts of αA. The αA-expressing clonal cell lines had αA content varying from 0.1 to 2.0 ng/μg protein as determined by Western blot analysis ¹⁴ . The increase of GSH in αA clones, expressed as percent increase over the mock clones, showed a positive correlation to αA content (Fig. 1) . For example, GSH concentration in a low αA- (0.2 ng/μg protein) containing clone was 42.0 nmol/mg protein (a 17% increase) over that of a mock-transfected clone with 35.9 nmol GSH/mg protein. In a representative clone with high αA (1.5 ng/μg protein), GSH concentration was 67.2 nmol/mg protein (an 80% increase) over that of a mock control with a GSH level of 37.3 nmol/mg protein. HPLC analysis showed that GSH was predominantly (>99%) in the reduced form and the GSH/GSSG ratio was not different between the mock controls andα A-expressing cells. It should be noted that mock-transfected cells, like untransfected cells, had no detectable level of αA. 

Maximal rates of GSH synthesis in mock-transfected and αA-expressing clones are shown in Figure 2 . Figure 2A shows a representative tracing of the measurement of synthetic rate in a mock clone and αA-expressing clone that contained∼ 0.2 ng αA/μg protein. As shown in Figure 2B , GSH synthetic rates in αA-expressing cells (0.84 ± 0.05 nmol · min⁻¹/mg protein) were significantly higher than that in mock-transfected HLE-B3 cells (0.61 ± 0.02 nmol · min⁻¹/mg protein). GSH synthetic rates derived from measurements of cytosolic proteins may be underestimates because they represent rates per milligram of soluble protein, of which αA is only a fraction. 

GCS-HS and GCS-LS mRNA and Protein in αA-Expressing Cells

Effect of αA expression on the level of GCS mRNA was quantitated by Northern blot analysis of several mock and αA clones. Figure 3 shows a representative Northern blot analysis from a pair of mock andα A-expressing clones. αA expression caused a significant (approximately twofold) increase in GCS-HS mRNA and GCS-LS mRNA. The mRNA levels (means ± SEM, n = 3) for GCS-HS and GCS-LS in αA-expressing clones as estimated by image analysis were 219% ± 20% and 194% ± 23% of mock-transfected controls, respectively. 

Western blot analysis of αA-transfected HLE-B3 clones expressing 0.2 to 2.0 ng αA/μg protein compared with mock clones showed that the amount of both GCS subunits quantitated by image analysis increased by 83% ± 36% for GCS-HS and 97% ± 28% for GCS-LS (mean ± SEM, n = 3) in αA-expressing cells over that of the mock controls. Figure 4 shows a representative Western blot of an αA-expressing clone with 0.2 ng αA/μg protein and a mock-transfected control. The increase in GCS-HS and GCS-LS protein due to αA expression was 62% and 73%, respectively. 

GSH Concentrations in αA(−/−) 129SvJ Mouse Lenses

GSH concentrations of whole lenses from αA(−/−) andα B(−/−) mouse lenses and their age matched controls of a wide age range (0.5–14 months) were determined. In Figures 5A and 5B , levels of GSH in αA(−/−) and αB(−/−) lenses as a function of age are expressed as a percent of that of age-matched controls. GSH levels in αA(−/−) lenses were not significantly different from that of wild-type controls in very young lenses, that is, 0.5 month and 1 to 2 months groups. GSH levels decreased significantly (∼45%) in 3- to 4 month-old αA knockout lenses concomitant with cataract formation, and this level of (decreased) GSH was maintained for the entire age span studied (Fig. 5A) . In contrast, GSH levels in αB(−/−) lenses were not significantly different from that of age-matched, wild-type lenses for any of the age groups studied (0.5–14 months; Fig. 5B ). 

Figure 6 shows slit lamp pictures of eyes from 7- and 10-week-old αA(−/−) mice. Although the 7-week-old lenses showed minimal opacification, the 10-week-old αA(−/−) lenses clearly showed cataract formation. As reported earlier, ²² complete opacification (mature cataract) of the lens occurs within ∼18 to 20 weeks from birth (not shown). GSH levels in 7-week-old αA(−/−) and wild-type lenses were not significantly different from each other (see also 1–2 month group in Fig. 5A ). Levels of GSH in 10-week-old lenses were significantly different from that of wild-type controls (Fig. 5A) . Because eyes of 0.5-month-old mice are barely open, we could not perform slit lamp examination on these lenses. 

To determine whether knockout of αA gene alters the profile of thiols and disulfides, we performed HPLC of lens homogenate from 10- and 22-week-old αA(−/−) lenses and their age-matched, wild-type lenses. Cataract formation, although mild to moderate, was already evident in 10-week-old αA(−/−) lens, whereas the 22-week-old αA(−/−) lens had a fully developed, mature cataract. As seen in Figure 7 , we could confirm by HPLC that the molecular form of glutathione is predominantly GSH in αA(−/−) lens as in the wild type. GSH level in 10-week-old αA(−/−) lens was approximately 25% lower than that in wild type, whereas it decreased to approximately 45% that of wild type in 22-week-old αA(−/−) lens. GSSG levels were very low (<1%) in both 10- and 22-week-old wild-type and αA knockout groups. 

GCS Protein in αA Knockout Lenses

Quantitation of protein levels of GCS indicated that in the absence of αA, GCS protein also decreased. Figure 8 shows a Western blot for 0.5-month-old αA(−/−) lens along with an age-matched, wild-type lens. Equal loading of proteins was confirmed by Coomassie Blue staining (not shown). Quantitative densitometry showed that both GCS-HS and GCS-LS subunits in αA(−/−) lenses were significantly decreased. The levels were 41% ± 2% for GCS-HS and 45% ± 3%, for GCS-LS; mean ± SEM, n = 3) compared with wild-type lenses. 

Discussion

In the present study, we have shown that αA expression in human lens epithelial cells increases cellular GSH and the rate of GSH synthesis. Increased GSH was shown to be primarily due to upregulation of the gene and protein of the regulatory and catalytic subunits of the rate-limiting enzyme GCS in a coordinate fashion. On the other hand, absence of αA in knockout lenses was associated with a decreased expression of GCS-HS and GCS-LS, and the decrease was shown to occur before cataract formation. 

Mehlen and coworkers ²¹ have recently reported that a possible mechanism of protection of cytokine-induced cell death by sHSP hsp27, and αB is through elevated GSH. Their study was performed in NIH-T3 ras cells and L929 cells expressing sHSPs and used TNFα for induction of cell death. The mechanism of the elevation of GSH and whether this protective action also holds true for αA, another sHSP family member, was not studied. We have recently established a cell culture system of immortalized human lens epithelial cells useful in studies of epithelial cell metabolism of physiological substrates. ^{37 38} Passaging these extended life span cells several times results in loss of αA. ⁴⁵ αA could then be reintroduced in these cells by transfection techniques. ¹⁴ Levels of αA in αA-transfected cells approached that of primary cultured lens epithelial cells of early passages. In studies with αA-transfected human lens epithelial cells with varying amounts of αA, we could show recently that αA expression protected apoptotic cell death induced by UVA radiation. ¹⁴ In further support for this role for αA, it was found that primary lens epithelial cells isolated from αA knockout mice were more susceptible to cell death from the above apoptotic stimuli compared with those of wild-type lenses. ¹⁴  

We hypothesized that protection from UVA-induced cell death by αA may be mediated in part by GSH. Cellular GSH was increased in HLE-B3 cells expressing αA and was positively correlated with αA content. An increase in mRNA of both subunits of rate-limiting GCS with αA could also be demonstrated. Protein levels of GCS-HS and GCS-LS also showed a significant increase for αA-expressing cells over that of the mock-transfected cells, which lack αA. According to a number of published reports in several cell types, transcriptional regulation of GCS can occur by differential or coordinated increases in the mRNA of the heavy and light subunit of GCS. ^{46 47 48} In our model ofα A-transfected human lens epithelial cells, we found that the two genes were coordinately regulated with αA expression. We could not get data on the effect of αA knockout on GCS mRNA levels because of the limitation in the availability of tissue material for mRNA isolation for Northern blot analysis. However, similar to mock-transfected HLE-B3 cells compared with αA-transfected HLE-B3 cells, GCS-HS and GCS protein in the absence of αA (in αA knockout lenses) was significantly lower than that of wild-type control lenses. 

In knockout mice, Brady et al. ²² reported that αA(−/−) lenses develop mild cataract about 7 weeks after birth, and a mature cataract with dense opacity can be seen in 18 weeks. On the other hand, lenses from αB(−/−) mice remained cataract-free up to 14 months until their death. ²³ In our effort to delineate the relationship among GSH levels, α-crystallin content, and the degree of cataractogenesis, we determined GSH levels in αA(−/−) andα B(−/−) lenses as a function of age, which gave some interesting results. GSH levels were unchanged in αA(−/−) lenses compared with wild-type controls in very young lenses, namely 0.5 months and 1 to 2 months of age. Cataract formation is minimal in this age span in theα A knockout lenses (see Fig. 6 , also Ref. ²² ). The observation that GCS-HS protein in prenatal αA knockout lenses is significantly decreased while GSH is maintained suggests that there are alternate mechanisms to offset decreased biosynthesis. We believe that increased GSH uptake in very young lenses may be important in this regard. We have shown that GSH uptake is high in the lens and brain of very young animals and declines with age. ^{49 50} GSH levels in αA(−/−) lenses began to decline compared with wild-type levels after 2 months of age. The αA(−/−) lenses at a later age span (3–14 months) had more or less similar GSH concentrations, which is 50% to 60% of that of age-matched, wild-type lenses. On the other hand, as mentioned before, αB(−/−) lenses did not show opacification, and their GSH levels were unaffected throughout the study period. 

Among the possibilities for increased GCS mRNA in αA-expressing cells, the involvement of transcriptional factors is particularly important. This process may involve activation, stabilization (or reduced inactivation), or increased efficiency of transcription. αA may increase cellular GSH content by increasing transcription factors in nuclear extracts of αA-expressing cells. Recent studies have shown transcriptional regulation of GCS-HS through the AP-1 response element-like binding site in its promoter and increased transcription through the antioxidant elements ARE-3 and ARE-4. ^{29 51} The present view is that out of the positive and negative regulatory elements, AP-1 site appears to play a key role. ⁵² AP-1 and NF-κB were found to be the main factors that mediate GCS-HS transcription in other cells. ^{29 32 43} It will be of interest to study if the above or any other transcription factors are responsible for the observed increase in GCS due to αA. 

The current studies, along with our recent work on protection of lens epithelial cells from apoptosis under conditions of αA expression, suggest an additional, antioxidative role for αA. Given the pleiomorphic properties of α-crystallins, this may be expected. However, it is unclear at the present time whether or not the protective function of αA is an independent phenomenon that can be dissociated from its well-known chaperone activity. In this context, it would be of interest to examine the effect of expression of chaperone-defective αA mutants ^{53 54} in HLE-B3 cells with respect to GSH regulation. 

An important consideration to be taken into account with respect to mechanism of GCS regulation by transcriptional factors is the cellular localization of sHSPs including α-crystallins. The association ofα -crystallins with nuclear and cytoskeletal elements suggests that they may have pleiomorphic functions in cells. ^{16 17} Recently, Bhat et al. ¹⁶ have demonstrated the presence ofα B inside the nucleus under conditions of its ectopic expression in stably transfected, unstressed CHO cells. On the other hand, information on subcellular localization of αA in normal and stressed states is limited. Therefore, in addition to nuclear translocation, operation of other direct or indirect mechanisms such as regulation or stabilization and/or intracellular signal transduction by cytosolic transcription factors cannot be excluded to explain the phenomenon of increased GCS mRNA with αA. 

In summary, we have demonstrated that αA expression in HLE-B3 cells caused an elevation in cellular GSH, in part because of an increase in mRNA and protein of both GCS subunits. GSH levels decreased with increasing age in αA knockout cataractous mouse lenses but not in noncataractous αB knockout lenses. Another interesting finding was that steady state GSH was maintained in young, precataractous αA knockout lenses with diminished GSH biosynthesis, possibly by an upregulation of the GSH transport processes. Studies on the elucidation of molecular mechanisms of the interrelationship between αA and GSH are being actively pursued in our laboratories. 

 Supported by National Institutes of Health Grants EY 11135, EY 05681, and core grants for Vision of Washington University (EY 02687) and USC Center for Liver Diseases (DK 48522), and an unrestricted grant from Research to Prevent Blindness, Inc. UPA is the recipient of Research to Prevent Blindness Lew R. Wasserman Award.

 Submitted for publication June 6, 2000; revised October 13, 2000; accepted November 3, 2000.

 Commercial relationships policy: N.

 Presented in part at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, May 1999.

 

Figure 1.

 

Increase in GSH levels in HLE-B3 cells expressing αA over mock-transfected cells. Percent GSH increase in GSH levels inα A-expressing cell lines with low αA (0.1 ng/μg protein) to highα A (2.0 ng/μg protein) over mock-transfected cells is shown. At least two αA-expressing clones were used for each αA concentration, and the mean percent increase in GSH values over mock-transfected cells is presented. A trend for positive correlation between αA content and increase in cellular GSH levels was observed.

Figure 2.

 

Measurement of maximal GSH synthetic rate in HLE-B3 cells expressingα A. GSH-SR was measured by the monochlorobimane (mBCl) fluorescent technique as described in Methods. Predialyzed cytosol from a mock clone (lower curve, A) and αA-expressing clone (upper curve, A) was incubated in the presence of GSH precursors and cofactors and mBCl. GSH-SR was determined by the difference in the rate of synthesis in the presence (background) or absence of buthionine sulfoximine (BSO). For simplification, only the tracings from BSO-untreated cytosol are shown. (B) Bar graph showing GSH synthetic rate (mean ± SEM, n = 3) in mock-transfected HLE-B3 cells and inα A-expressing cells with ∼0.1 to 0.2 ng αA/μg protein. The GSH synthetic rates in αA-expressing cells were significantly (P < 0.05) higher than those in mock-transfected cells.

Figure 3.

 

Northern blot analysis of GCS-HS (A) and GCS-LS (B) in HLE-B3 cells transfected with αA and mock controls. A representative blot for a mock clone and an αA-expressing clone is shown. β-Actin was used as a standard for quantitation of both GCS subunits. Details for Northern blot analysis are given in Methods.α A-expressing cells exhibited a significantly higher gene expression of the two GCS subunits compared with controls.

Figure 4.

 

Western blot analysis of GCS-HS and GCS-LS in a αA-expressing clone and a mock-transfected clone. Mouse liver was used as a positive control. Details of antibody preparation and analysis are given in Methods. The 78- and 30-kDa protein bands for GCS-HS and GCS-LS were significantly higher in the αA-expressing HLE-B3 clone compared with the mock-transfected clone.

Figure 5.

 

Whole lens GSH in wild-type and αA(−/−) lenses (A) andα B(−/−) lenses (B). GSH levels in αA and αB knockout lenses for each age span are expressed as percent of GSH levels in wild-type lenses which is taken as 100. Data are means ± SEM from three to four lenses for each age group. Total GSH was unchanged for the first 2 months in αA(−/−) lenses compared with wild-type controls and decreased thereafter. No significant difference in GSH concentrations was found between wild-type and αB(−/−) lenses of all age groups.

Figure 6.

 

Examination of eyes from 7- and 10-week-old αA(−/−) and αA(+/+) mice by slit-lamp biomicroscopy. Eyes were dilated and examined by slit lamp. Top: wild-type 7- and 10-week-old mice; bottom: αA knockout lenses of the same age. Normal reflection of the slit lamp from the surface of the cornea, and the lens can be seen. Light scattering within the lens (white haze in photograph) was significantly higher for theα A(−/−) mice compared with wild-type mice for both ages. A mild cataract is seen in the lenses of 7-week-old αA(−/−) mice, and the cataract progresses to a moderate opacification in 10-week-old lenses. A fully mature cataract with dense opacity developed in 18 weeks in these mice (not shown; see Ref. ²² ).

Figure 7.

 

HPLC profile of lens homogenate from αA(−/−) lenses and wild-type lenses. Two lenses each from 10- (left) and 22-week-old (right) mice from both groups were used for HPLC. Soluble thiols (TCA supernatant) from control and αA(−/−) lenses were processed as described. ³⁹ Equal aliquots representing the same amount of lens wet tissue for the two age groups were derivatized for HPLC. Elution times of GSH and GSSG are marked.

Figure 8.

 

Western blot analysis of GCS-HS (left) and GCS-LS (right) in lens homogenates from a 0.5-month-old αA(−/−) lens (lane 2) and age-matched, wild-type lens (lane 3). A normal mouse liver was used as a positive control (lane 1). Generation of the two antibodies and the method for analysis are described in Methods. The 78- and 30-kDa protein bands for GCS-HS and GCS-LS were significantly lower in αA(−/−) lenses compared with the wild-type.

The authors thank Diana Tang and Zheng Song for excellent technical assistance. The authors also thank Shelly C. Lu (University of Southern California) and Terrence J. Cavanagh (University of Washington) for GCS antibodies and Steven Bassnett (Washington University) for helpful discussions.