Regulation of Human Thioltransferase (hTTase) Gene by AP-1 Transcription Factor under Oxidative Stress

Abstract

purpose. To study the mechanism of activation of the human thioltransferase (hTTase) gene under oxidative stress.

methods. Human lens epithelial cells (HLE-B3) were exposed to 0.1 mM H₂O₂ for 0, 5, 10, 15, 30, or 60 minutes; lysed; and used for gel mobility shift assay (GMSA) and supershift assay for activating protein (AP)-1 transcription factor. The search for transcriptional coactivators was performed with Western blot analysis. The ability of different parts and a mutated fragment of the hTTase gene promoter region to activate the gene expression under the oxidative stress conditions was examined by reporter gene assay.

results. The AP-1–binding element was identified in the 5′ region of the hTTase gene, and evidence was obtained that binding of AP-1 with this element in vivo was redox sensitive. In addition, the pattern of AP-1 binding under the oxidative stress was similar to the pattern of TTase activity and mRNA synthesis modulation. In contrast, direct exposure of the cell lysate to oxidants, reductants, or redox-regulating enzymes in vitro had no influence on AP-1 binding. AP-1 transcriptional coactivator redox factor (Ref)-1 was present in the lens epithelium and was association with the AP-1–binding complex during oxidative stress. In the reporter gene assay, only the fragments of the hTTase 5′ region, which contained the AP-1–binding site, could activate the CAT reporter gene’s expression in an oxidative stress–dependent manner. The mutant with a replaced AP-1–binding site failed to stimulate CAT expression in an oxidation-sensitive manner. The results showed that the c-Jun component in the AP-1–binding complex was transiently phosphorylated during H₂O₂ treatment. The c-Jun N-terminal kinase or SAPK/JNK, which responds to stress signaling and is the upstream protein kinase of c-Jun, was activated and translocated from cytosol to nucleus under the same conditions.

conclusions. The data demonstrate that the activation of the hTTase gene under oxidative stress depends on the AP-1 transcription factor. The event was initiated only through an intact cell, possibly mediated through signal transduction by a phosphorylation–dephosphorylation mechanism. As far as the authors know, this is the first evidence of the association of AP-1 with the regulation of hTTase gene expression.

The endogenous generation of reactive oxygen species (ROS) and extracellular oxidative stress are generally under strict control by the cells. ROS in excess can cause serious damage to cellular structures and vital macromolecules, ^{1 2 3} but, at the same time, ROS in moderate concentration can participate in mediation of gene activity, ⁴ influence the major signal-transduction pathways, ⁵ and regulate the activity of preexisting proteins. ^{6 7} Therefore, cells must have sensitive and efficient mechanisms to control their redox status by using oxidation-defense enzymes, including superoxide dismutase, catalase, and glutathione peroxidase, and also by using free radical scavengers such as glutathione (GSH), ascorbic acid, and vitamin E. The cell keeps its intracellular environment in a reduced state by maintaining the ratio of GSH to its oxidized form (GSSG) at 100 to 1 or less. ³ Thus, oxidation of GSH to GSSG can change this ratio and dramatically affect cellular redox status. If the primary protection systems fail, the cells use thiol-regulating and repair enzymes, such as thioltransferase (TTase, also referred to as glutaredoxin) or thioredoxin (TRx), to inhibit the oxidation-induced damage to proteins and enzymes, to recover their functional activities, and to restore the thiol homeostasis in the cells. ^{7 8}  

The main oxidative damage to proteins is caused by the oxidation of protein sulfhydryl groups (S-thiolation) and generation of protein-GSH, protein–cysteine mixed disulfides, or protein–protein disulfides and intermediate products. ^{3 7} S-thiolation of proteins can lead to conformational changes with eventual partial or complete loss of function. Structural proteins, such as lens crystallins, form high-molecular-weight aggregates from these partially or completely denatured proteins, leading to the loss of lens transparency and eventual formation of cataract. ^{9 10} S-thiolated proteins in normal physiological conditions can be dethiolated by TTase. This enzyme catalyzes dethiolation of protein–thiol mixed disulfides, repairs the proteins that are modified by glutathionylation, regulates the activity of glutathione-dependent proteins, and participates in the control of cellular redox status. ^{11 12 13}  

It is well known that the expression of cellular redox-regulating enzymes is controlled by an extremely sensitive mechanism that allows the genes to be rapidly upregulated at the onset of oxidative stress and downregulated when the oxidant is dissipated. For instance, the gene coding for thioredoxin, which mostly disrupts protein–protein disulfides, can be upregulated two- to three-fold after treating the cells with H₂O₂ for 20 minutes. Once the oxidant is removed from the media, gene expression immediately returns to a normal level. ¹⁴ Engelberg et al. ¹⁵ observed upregulation of the xis ⁴ gene in yeast up to 10-fold over the control, within 15 minutes of UV irradiation, followed by gradual downregulation. A similar phenomenon was observed in the expression of the human (h)TTase gene in the lens epithelial cells (B3) exposed to a low level (0.1 mM) of H₂O₂. ¹⁶ Both TTase mRNA level and enzyme activity were doubled under such conditions. Furthermore, key metabolic enzymes, such as glyceraldehyde-3-phosphate dehydrogenase (G-3PD), inactivated by oxidative stress, are dethiolated and reactivated by TTase. ^{17 18 19}  

The expression of many genes is regulated by activating protein (AP)-1 transcription factor. The eukaryotic AP-1 is a product of homo- or heterodimerization of proteins that includes members of the Jun, Fos, or activating transcription factor (ATF) families. Most cellular AP-1 is formed by c-Jun and c-Fos proteins, and the complex is believed to be the most stable and active form in stimulating transcription of target genes. ^{20 21} The reason for such a variety of AP-1 composition is not quite clear, but apparently the difference in the activity of DNA binding and the ability to activate transcription between different forms of AP-1 plays a certain role in regulation of transcription. ²² The DNA-binding activity of AP-1 is controlled by association with other proteins, such as oxidation-dependent redox factor (Ref)-1, which is present in different cell types. ²³ Ref-1, which interchanges between oxidized and reduced forms, is able to act as a redox sensor and thus influences the DNA-binding activity of AP-1 and other transcription factors under oxidative stress conditions. ^{24 25} This process is mediated by thioredoxin (TRx). In response to oxidative stress signals, TRx, which is a thiol-regulating enzyme, is translocated from cytosol to the nucleus, where it interacts and reduces Ref-1. The reduced Ref-1 gains the ability to associate with the transcription factor and thus stimulates DNA-binding activity. ²⁶  

It has been shown that expression of multiple genes of the oxidative stress defense system is controlled by the AP-1 transcription factor. ^{27 28} This has been demonstrated for glutathione-S-transferase, ²⁹ glutathione reductase, ³⁰ thioredoxin, ³¹ thioredoxin reductase, ³² and γ-glutamyl-cysteine synthetase. ³³ However, the association of the AP-1 transcription factor with regulation of the hTTase gene has not been explored. 

In the present report, we studied whether AP-1 transcription factor might be involved in the regulation of expression of expression of the hTTase gene under oxidative stress conditions. Our results provide evidence that the hTTase gene has a binding site for AP-1 and that phosphorylation and activation of c-Jun N-terminal kinase (SAPK/JNK) in the signal-transduction pathways and its downstream target c-Jun in the AP-1 structure complex may be responsible for such regulation. 

Materials and Methods

Materials

Hydrogen peroxide (H₂O₂), EDTA, dithiothreitol (DTT), minimum essential medium (MEM), fetal bovine serum (FBS), gentamicin, trypsin-EDTA (1×), and thioredoxin (TRx) were purchased from Sigma Chemical Co. (St. Louis, MO); protease inhibitors, aprotinin, leupeptin, and pepstatin from Kirkegaard & Perry Laboratories, Inc. (Gaithersburg, MD); silica gel plates LK6 from Whatman (Maidstone, UK). ¹⁴C-labeled chloramphenicol from Amersham Pharmacia Biotech (Piscataway, NJ); glyceraldehyde 3-phosphate dehydrogenase (G-3PD) monoclonal antibody from Advanced Immunochemical, Inc. (Long Beach, CA); and protein assay reagent (bicinchoninic acid [BCA]) from Pierce Chemical Co. (Rockford, IL). The recombinant human lens thioltransferase was purified according to Qiao et al. ¹² All other chemicals and reagents were of analytical grade. 

Human Lens Epithelial Cell Culture

Human lens epithelial cell line HLE-B3 was kindly provided by Usha Andley (Washington University, St. Louis, MO). The cells were grown in 10 mL MEM supplemented with 50 μg/mL gentamicin plus 20% FBS (pH 7.2) in 100 × 20-mm tissue culture plates in a humidified atmosphere with 5% CO₂ at 37°C. Cells reached confluence (∼6 × 10⁶ cells/plate) within 4 days. 

Treatment of Cells with a Bolus of H₂O₂

The cells were grown to confluence and trypsinized, and 1.6 million cells were subcultured in 50-mm plates in 5 mL MEM with 2% FBS at 37°C overnight and then in serum-free medium for 30 minutes before the cells were used for H₂O₂ treatment. This gradual serum-starved culturing condition is necessary to minimize any serum interaction with the H₂O₂ used in the studies. The serum-starved cells were treated with a bolus of 0.1 mM H₂O₂ in 5 mL MEM. At different intervals (0, 5, 10, 15, 30, and 60 minutes), cells were harvested and used for various studies. We adopted this protocol, because the B3 cells tolerate the treatment of a bolus of 0.1 mM H₂O₂ well without DNA damage or morphologic change within 3 hours of the experiment. This concentration of H₂O₂ in the medium was detoxified by the cells within 60 minutes. ¹⁷  

Studies of AP-1 Binding to the hTTase Gene by Gel Mobility Shift Assay

HLE B3 cells (4 × 10⁶) were treated with H₂O₂ in 80-mm plates, as described earlier, and nuclear lysates were prepared from normal and H₂O₂-treated cells according to the method of Ausubel et al. ³⁴ Two complementary single-stranded oligonucleotides, containing a 22-bp region of hTTase gene 5′ region with putative AP-1–binding site were synthesized and annealed to each other (AP1BSd: 5′-TTCCCCAGAGGTGACTAAACTCTG-3′, AP1BSr: 5′-TCAGAGTTTAGTCACCTCTGGGG-3′). Both of the single-stranded oligos contained additional noncomplementary thymidine residues on their 5′ ends. After annealing, those thymidines created the 5′ T overhangs that were used for labeling of the oligo with ³²P-dATP and Klenow DNA polymerase (Gibco Life Sciences, Rockville, MD). The labeled probe was precipitated twice with ammonium acetate and glycogen ³⁵ to separate the nonincorporated ³²P-dATP. A gel mobility shift assay (GMSA) was performed as described by Ausubel et al. ³⁴ Aliquotes (5 μg) of nuclear lysates were preincubated in assay buffer with nonspecific competitor, poly(dI-dC) to poly (dI-dC), for 10 minutes at room temperature. An equal amount (3 × 10⁴ disintegrations per minute [dpm]) of the labeled oligo was added to each tube, and the reaction mixtures were incubated for an additional 30 minutes at room temperature and separated on a polyacrylamide gel (4.5%). 

Identification of c-Jun and c-Fos in the AP-1 Binding Complex by Supershift Assay

For supershift assay, anti-c-Jun (Santa Cruz Biotechnology, Santa Cruz, CA) or anti-cFos (Calbiochem, La Jolla, CA) antibody was added to the reaction mixture described earlier at a 1:2000 dilution. 

Effect of Oxidation and Reduction on AP-1 Binding In Vitro

Nuclear lysate prepared from human lens epithelial cells (1.6 million) was incubated with 0.1 mM H₂O₂, or 100 mM DTT, or disulfide-reducing enzyme (8.4 mU TRx), or a dethiolating enzyme (50 μU TTase; pure recombinant human lens TTase). The reaction was performed at 30°C for 30 minutes. Each of the reaction mixtures was used in a GMSA, as described earlier. 

Western Blot and Immunoprecipitation Analysis for the AP-1 Coactivator Ref-1, in the Lens Epithelial Cells

Normal and H₂O₂-treated cells were washed with PBS buffer and lysed on plates with modified RIPA buffer (50 mM Tris-HCl [pH 7.4], 1% NP-40, 150 mM NaCl, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride [PMSF], 1 mM NaF, 0.25% Na-deoxycholate, 1 mM Na₃VO₄, and 1 μg/mL each of aprotinin, leupeptin, and pepstatin). For Western blot analysis, 20 μg total protein was separated in 10% SDS-PAGE and transferred to a membrane (TransBlot; Bio-Rad, Hercules, CA). These membranes were blocked in 5% blocking reagent (Blotto; Santa Cruz Biotechnology) in TBS containing 0.1% Tween 20 and incubated with anti Ref-1 primary antibody (Santa Cruz Biotechnology) diluted 1:2000 in blocking solution. The membranes were then treated with secondary antibodies (horseradish peroxidase; [HRP]–conjugated; Santa Cruz Biotechnology) diluted 1:2000 in blocking solution, washed, and developed using a chemiluminescent reagent (Luminol; Santa Cruz Biotechnology). 

For immunoprecipitation of Ref-1/AP-1 complexes, the nuclear lysates from H₂O₂-treated HLE B3 cells were prepared as described earlier. Lysates were precleared by shaking with protein A–affinity medium (Protein A Sepharose; Sigma-Aldrich) for 15 minutes in a cold room, centrifuged, and incubated with anti-c-Jun antibody overnight at 4°C. The immunocomplexes were captured by adding the protein A–affinity medium and incubated for 2 hours at room temperature. Protein-bound beads were collected by brief centrifugation, washed three times with cold PBS, resuspended in 2× sample buffer, and boiled for 10 minutes to dissociate the immunocomplexes. Proteins were separated in 8% SDS-PAGE and analyzed by Western blot analysis, using anti-Ref-1 or anti-c-Jun antibodies, as described earlier. 

Analysis of hTTase Gene Promoter Activity by Reporter Gene Assay

Preparation of Plasmids.

Chromosomal DNA was isolated from HLE B3 cells by using a kit (Purigene; Gentra Systems, Minneapolis, MN). Fragments of the 5′ region of the hTTase gene was amplified using Taq DNA-polymerase (Takara; PanVera Corp., Madison, WI) and cloned before the promoterless bacterial chloramphenicol-N-acetyltransferase (CAT) gene into a plasmid, containing both pUC-18 and SV-40 origins of replication. As well as replication origins, this plasmid also contained ampicillin- and neomycin-resistant genes and a BHG polyadenylation site after the CAT gene. The downstream primer for PCR amplification was always +81, based on the hTTase sequence of Park and Levine ³⁶ : 5′-ACTCTTGAGC[TAT]GCCGATG-3′, with the brackets indicating a mutated ATG codon of an hTTase gene (TAT instead of CAT). Upstream primer locations were: –1190 (5′-ATTCCTCTTACGTGGCAGTC-3′), −776 (5′-TTGCCTCCATATTAAACTGC-3′), −437 (5′-GAGGAGTGCAAAGTCACAGT-3′), −183 (5′-CCAGAGTGTCTTTCATAGCT-3′). All the primers were synthesized by ID Technologies (Coralville, IA). The conditions of PCR were 92°C for 1 minute, 10 seconds; 52°C for 1 minute; and 72°C for 1 minute 20 seconds. PCR products (1271, 857, 518, and 264 bp respective to the upstream primer location) were purified with a gel-extraction kit (Qiagen, Valencia, CA) and cloned as described in Sambrook et al. ³⁵  

Transfection and CAT Assay.

Plasmids for CAT assay were purified with a kit (QIAfilter Plasmid Midi; Qiagen) and dissolved in sterile water. HLE B3 cells (1.6 million) were transfected with 2 μg of each plasmid, using transfection reagent (FuGene 6; Roche Molecular Biochemicals, Indianapolis, IN) in conditions recommended by the manufacturer. After the transfection, cells were cultured for 24 hours in MEM with 20% FBS and treated with H₂O₂, as described earlier. The cells were lysed by four cycles of freezing in an acetone–dry ice mix and thawing at 37°C and were then assayed for CAT enzyme activity, by using ¹⁴C-labeled chloramphenicol, according to Sambrook et al. ³⁵ Equal amounts of the total proteins were used in each assay. The acetylated chloramphenicol products of the assay reactions were separated by thin-layer chromatography on silica glass plates. 

Mutagenesis of the AP-1–Binding Site

The AP-1–binding site was mutated in the pK-CAT438 plasmid using a mutagenesis kit (ExSite; Stratagene, La Jolla, CA), according to the conditions recommended by the manufacturer. Primers used for mutagenesis were forward, 5′-AACTCTGATCATTGCCAATGG-3′, and reverse, 5′TCTAGACTCTGGGGAAGCATAAA-3′. AP-1–binding site in the reverse primer (GTGACT) was mutated to TCTAGA, a site for restriction endonuclease XbaI, which was later used for primary screening of the mutants. 

Identification of Phosphorylated c-Jun and Phosphorylated SAPK/JNK

Human lens epithelial cells (1.6 million) were treated with a bolus of H₂O₂ as before. The cell lysate from cells pre-exposed to H₂O₂ for 0, 5, 10, 15, 30, or 60 minutes were examined by Western blot for the presence of phospho-c-Jun (P-c-Jun) using anti-phospho-c-Jun antibody (Santa Cruz Biotechnology). The presence of the activated SAPK/JNK (phospho[P]-SAPK/JNK) in the cell extracts and the nuclear lysates was studied with anti-P-SAPK/JNK (Cell Signaling Technologies, Beverly, MA). 

Protein Determination

Protein content in the cell homogenate was determined using a BCA microprotein assay. ³⁷  

Results

AP-1 Binding to the Putative Binding Site in the 5′ Region of the hTTase Gene

Based on the hTTase sequence data of Park and Levine, ³⁶ the space that spans approximately 1300 bp from the transcription start point (TSP) in the 5′ region of the hTTase gene contains one copy of a putative AP-1–binding site in the position between −248 and −241 bp from the TSP. The sequence of this site (Fig. 1A) differs from the consensus AP-1–binding site (5′-TGACTCA-3′) by C-to-A transversion of one of the terminal nucleotides. To analyze the ability of AP-1 to bind the putative site in the hTTase 5′ region, we performed a GMSA with ³²P-labeled, double-stranded oligo containing the putative AP-1–binding site (Fig. 1A) and nuclear lysates from normal and H₂O₂-treated HLE B3 cells. As shown in the autoradiograph in Figure 1B , the slower-moving band depicts the binding complex formed between the AP-1 in the lysate and the labeled nucleotide probe. The faster-moving band depicts the nonspecific binding complex that formed. The DNA-binding ability of AP-1 was increased after 5 minutes of treatment with H₂O₂ (lane 5) and reached its maximum (lane 10) over the control (lane 0) at 10 minutes. The binding activity started to decrease after 15 minutes (lane 15), as the H₂O₂ in the medium was gradually detoxified by the cells. The binding activity at 30 minutes (lane 30) was lowered to its basal level after 60 minutes (lanes 60 and 120), when the oxidant was completely eliminated. A control reaction, which contained 100-fold excess of unlabeled oligo was included in the study (lane C) to distinguish between specific and nonspecific binding. The absence of the labeled AP-1–binding complex in this lane indicates that unlabeled probe had specifically competed for binding with AP-1. 

Verification of the Composition of AP-1–Binding Complex by Supershift Assay

To confirm that the bound protein in the GMSA (Fig. 1B) was indeed AP-1 transcription factor, we performed the supershift assay with anti c-Fos and anti c-Jun antibodies to locate c-Fos and c-Jun, the components of AP-1. Both antibodies were specific only to corresponding protein and had no cross-reactivity with other c-Fos and c-Jun family members. In Figure 1C , the retardation in gel mobility (from the GMSA in Fig. 1B ) of the AP-1/oligo/antibody complexes is demonstrated in lane 1 (anti-c-Jun antibody) and lane 2 (anti-c-Fos antibody) in comparison with AP-1/oligo complexes without the antibodies (lane 3). This indicates that AP-1, which responded to changes in cellular redox status by enhancing DNA binding consisted of c-Jun and c-Fos proteins. 

Effect of Oxidation and Reduction on DNA Binding by AP-1

In that DNA binding by AP-1 was shown to be enhanced by oxidative stress induced on intact cells, it was of interest to explore whether oxidation or reduction directly applied to the cell lysate would have any effect. We performed another GMSA (Fig. 1D) with cell lysates from nontreated HLE B3 cells after incubating the lysates with the oxidant H₂O₂ (lane 2), the reductant DTT (lane 3), or the disulfide-reducing enzymes TRx and TTase (lanes 4 and 5, respectively). All those agents and enzymes did not influence AP-1–binding activity in comparison with the control (nontreated, lane 1), indicating that the binding complex is insensitive to oxidation or reduction in vitro. 

Transient Appearance of P-c-Jun and P-SAPK/JNK in H₂O₂-Treated Lens Epithelial Cells

To investigate the possible mechanism for oxidative stress–stimulated AP-1 binding to the hTTase gene, we examined the presence of phosphorylated c-Jun or the activated c-Jun as a result of an upstream signal-transducing process. The cell lysate from cells pre-exposed to H₂O₂ for the period of 0, 5, 10, 15, 30, or 60 minutes were examined by Western blot for the presence of P-c-Jun, by using anti-P-c-Jun antibody. A transient increase in P-c-Jun was observed during oxidative stress (Fig. 2A) . The phosphorylation of c-Jun increased at 5 minutes, reached maximum at 15 minutes, and gradually subsided by 60 minutes. Probing the same filter with anti-G-3PD (Fig. 2A) indicated that the same amounts of proteins were used in the Western blot analysis. 

Similarly, activated SAPK/JNK or P-SAPK/JNK, which is the upstream protein kinase to P-c-Jun, was also found in the cell lysate (Fig. 2B) . Phosphorylation of SAPK/JNK was transiently increased between 5 and 60 minutes after exposure to H₂O₂ and began to decrease at 60 minutes, indicating that the stress-related signaling process was activated under the same conditions. Western blot with anti-G-3PD of the same samples in Figure 2B indicated that the same amounts of proteins were used on the gel. 

When the nuclear lysate of the H₂O₂–treated cells was analyzed for P-SAPK/JNK (or p54/46 JNK), it was found that P-SAPK/JNK was absent in the control, unlike the cytosolic samples, but appeared abruptly after 5 minutes of H₂O₂ treatment and then abruptly disappeared after 30 minutes of treatment. This suggests that this activated SAPK/JNK signaling component was translocated from the cytosol into the nucleus to activate its downstream target c-Jun (Fig. 3A) . The same membrane was probed for nonphosphorylated p46 JNK, to show that the P-SAPK/JNK was transferred to the nuclei but not phosphorylated in situ (Fig. 3B) . Again, the constant amount of G-3PD in the same samples as shown in Figure 3C confirms that all the samples shown in Figures 3A and 3B contained the same amount of proteins. 

Presence of Ref-1, a Transcriptional Coactivator of AP-1 and Its Association with the AP-1–Binding Complex

To investigate whether transcriptional coactivator p35 Ref-1 is also involved in the eye lens epithelium cells in AP-1–mediated gene activation, we examined the presence of Ref-1 in HLE B3 cells by Western blot analysis with anti-Ref-1 antibody. Lane 1 of Figure 4A clearly demonstrates the presence of Ref-1 with a molecular mass of 35 kDa in the nuclear lysate of HLE B3 cells. 

To study the possible involvement of Ref-1 in the AP-1–binding complex, we examined the presence of the AP-1/Ref-1 complex in the nuclear extract of the lens cells pretreated with H₂O₂ from 0 to 120 minutes, using immunoprecipitation with anti-c-Jun antibodies followed by Western blot analysis with anti-Ref-1 antibodies. The transient presence of Ref-1 in the AP-1/Ref-1 binding complex was clearly shown 5 minutes after the oxidative stress and reached its maximum by 15 minutes after treatment (Fig 4B , top). The Ref-1–positive band at 60 minutes had decreased almost to the same level as that of 5 minutes after treatment. Western blot was also performed with anti-c-Jun antibodies to ensure equal loading (Fig. 4B , bottom). This pattern coincided with the oxidative stress–induced upregulation of AP-1 binding in Figure 1B . 

Studies of the Regulatory Function of the AP-1–Binding site by Reporter Gene Assay

To investigate whether the AP-1 putative binding site at the 5′ region of the hTTase gene has the ability to drive the expression of a reporter gene, chloramphenicol-N-acetyltransferase (CAT), we used the full length of the 5′ region of hTTase and compared it with its truncated fragments, with and without the AP-1–binding site (Fig. 5A) . There were four upstream primers located at −1190, −776, −437, and −183 and one downstream primer located at +81 from the TSP. The downstream primer comprised the ATG-codon of the hTTase gene, which was mutated in the primer to avoid possible initiation of translation from that point in the constructs, with consequential out-of-frame translation of the reporter gene. The CAT gene contained its own ATG codon. The sizes of the fragments of hTTase were 1271, 857, 518, and 264 bp. The AP-1–binding site was present in the first three fragments but not in the 264-bp fragment. PCR fragments were cloned in the reporter construct in front of the CAT gene (Fig. 5B) . Another clone containing a 518-bp fragment with the mutated AP-1–binding site (for details see Materials and Methods) was also used (Fig. 5B) . Transfected cells were treated with 0.1 mM H₂O₂ for 0, 5, 15, or 30 minutes. After the treatment, the cells were lysed and used for CAT assay. 

The results of CAT activity as a function of H₂O₂ treatment time points (0, 5, 15, and 30 minutes) are shown in Figure 6 . The cells transfected with plasmids containing the reporter gene under the control of the 1271-, 857-, 518-, and 264-bp fragments, respectively, were used. The full-size 1271-bp as well as the truncated 859- and 520-bp fragments increased the expression of the CAT gene within 5 minutes after exposure to H₂O₂. The upregulation of CAT was above the control (0 time) at 5 minutes, was more significant at 15 minutes, and returned to the basal level at 30 minutes. However, the 264-bp fragment, which did not possess the AP-1–binding site, showed no upregulation of CAT expression under the same conditions. The effect of the mutation of AP-1 on its binding to DNA in the lens epithelial cells is clearly shown in Figure 6 , which shows that the fragment with the mutated AP-1–binding site did not increase the expression of the CAT gene. These results indicate that the AP-1–binding site within the 5′ region of hTTase is essential in controlling the gene’s expression. 

Discussion

Our earlier study showed that oxidative stress in HLE cells induces transient hTTase gene expression, which results in increased TTase enzyme activity and mRNA level, followed by a gradual downregulation of both when the oxidant in the culture media is totally detoxified. ¹⁶ Increased gene expression suggests that a certain transcription factor (or factors) was involved in this process. Analysis of the hTTase gene’s 5′ region revealed several putative transcription factor–binding sites. Presence of a sequence that is similar to the AP-1 transcription factor–binding site (Fig. 1A) prompted us to assume that it has a role in activation of the hTTase gene. To check this hypothesis, we performed a GMSA with ³²P-labeled double-stranded oligo containing a putative AP-1–binding site and nuclear lysates from normal and H₂O₂-treated HLE B3 cells (Fig. 1B) . Our results demonstrated that binding of AP-1 to the putative binding site located in the hTTase gene 5′ region occurred in an oxidation-dependent manner. The pattern of modulation of AP-1 binding shown in Figure 1B was very similar to the previously observed changes in TTase mRNA level under the same experimental conditions. ¹⁶ Furthermore, the hTTase gene promoter activity examined by the reporter assay demonstrated clearly that the plasmid containing either the intact hTTase 5′ region or a fragment of it containing the AP-1–binding site could successfully express the CAT gene in an oxidative stress–dependent manner, similar to the TTase expression in the H₂O₂-pretreated HLE B3 cells. On the contrary, the fragment devoid of the putative AP-1–binding site did not express the CAT gene in a similar manner (Figs. 5 6) . The mutation study also revealed that when the key sequence in the AP-1–binding site was substituted, the reporter gene’s expression no longer responded to oxidative stress, and thus activation of gene expression did not occur. All these findings strongly support our hypothesis that AP-1 may play a regulatory role in activation of the hTTase gene. As far as we know, this is the first evidence of AP-1’s association with the expression of the hTTase. 

Normally, the genes with expression that is regulated by the AP-1 transcription factor may contain one or more copies of the AP-1–binding site in their 5′ regions within the effective distance from the promoter (EDP). The number of copies plays a regulatory role. Therefore, the more copies present in the gene, the more active the expression. ^{38 39 40} In our search, only one AP-1–binding site was found in the 5′ region of the hTTase gene. The sequence of this site differed from the consensus AP-1–binding site by C-to-A transversion of one of the terminal nucleotides. This single base pair substitution in the AP-1–binding site may influence the transcription factor’s DNA-binding ability and may alter its activation response. ^{41 42} Our study revealed that in spite of this difference, AP-1 binding occurred in the HLE B3 cells and responded to oxidative stress. 

An important finding in the current study was that AP-1, which bound the hTTase 5′ region in an oxidative stress–dependent manner, consisted of c-Jun and c-Fos (Fig. 1C) . Furthermore, Ref-1, which is known to be involved in regulation of the transactivating function of AP-1, was also found in the HLE B3 cells. We also showed that Ref-1 could form complexes with AP-1 and that the pattern of this association under oxidative stress conditions (Fig. 4B) was similar to that of the AP-1 binding complex in the GMSA (Fig. 1B) . The contribution of other members of the c-Fos and c-Jun protein families to oxidation-dependent gene activation should be further examined. 

To confirm that oxidation to the intact cell is essential for AP-1–mediated upregulation of hTTase expression, we examined the effect of H₂O₂ in vitro by treating the nuclear lysate obtained from untreated cells directly with H₂O₂. We did not observe any change in the AP-1–binding complex (Fig. 1D) . In addition, the binding complex was not affected by direct treatment with a reductant (DTT) nor by treatment with the thiol-regulating enzymes TTase or TRx. These results agree with our previous findings that GSH-depleted HLE cells do not induce upregulation of hTTase gene expression. ¹⁶ Therefore, we conclude that the oxidative stress-induced upregulation signal must be initiated through an intact cell. 

How an intact HLE cell transduces such oxidative stress signals into the nucleus for the AP-1 to function is an intriguing question. Mammalian cells are known to have an intricate signal-transduction system to respond to various stimuli and to control appropriate biological functions. ⁴³ The main signaling system is the mitogen-activated protein kinase family, known as the MAPK superfamily. In addition to the traditional stimuli of growth factors and cytokines, recent studies have shown that oxidative stress can also stimulate the cells to promote a wide range of cellular functions, including proliferation, differentiation, and apoptosis. ⁴⁴ The c-Jun-N-terminal kinase or SAPK/JNK is one of the MAPK family members that responds to stress signals, such as UV and other oxidative stress stimuli. ⁴⁵ SAPK/JNK is known as a cytosolic protein, but when phosphorylated, it can be transported to the nucleus where it phosphorylates c-Jun and other downstream targets. ⁴⁶ We have recently shown that lens epithelial cells contain the complete signaling components of the MAPK superfamily, ⁴⁷ including SAPK/JNK. Our current results demonstrated that both SAPK/JNK and c-Jun were transiently phosphorylated in HLE B3 cells pre-exposed to a bolus of H₂O₂ (Figs. 2B and 2A , respectively). It is thus logical to assume that signaling through SAPK/JNK and its downstream target c-Jun is a plausible mechanism for AP-1 regulation under oxidative stress. Our data in Figure 3 support this hypothesis, because P-SAPK/JNK, which was absent in the nuclear lysate of the untreated HLE cells, abruptly appeared 5 minutes after oxidation treatment and then suddenly disappeared after 30 minutes. In contrast, P-SAPK/JNK was present in the cytosolic preparation (Fig. 2C) of the untreated cells and continued to appear 60 minutes after oxidation treatment. This suggests that the P-SAPK/JNK was translocated from cytosol to the nucleus, where it in turn phosphorylated c-Jun, leading to the activation of the AP-1 transcription factor. This translocation and subsequent phosphorylation of c-Jun and recruitment of Ref-1 may be the mechanism responsible for activation of the hTTase gene under oxidative stress conditions in our present study. 

In conclusion, our data indicate that expression of the hTTase gene, coding for an important redox-regulating enzyme, thioltransferase (glutaredoxin), under oxidative stress conditions may be regulated by the AP-1 transcription factor. To our knowledge, this is the first report of the regulatory role of AP-1 on the hTTase gene. The AP-1 transcription factor itself appears not to be a primary redox sensor in the cell, because its DNA-binding ability does not depend on direct oxidation and reduction, but is mediated by interaction with other transcriptional coactivators. We speculate that the mechanism for AP-1’s activation is mediated through redox signaling and that signal transduction resides mainly at the activation of SAPK/JNK, followed by the downstream target of c-Jun activation and the subsequent event of AP-1 transactivation of the hTTase gene. Further studies are under way to determine the primary H₂O₂ sensors and the involvement of the upstream signaling components in HLE cells. 

²  Present affiliation: Department of Medicine, University of California Los Angeles School of Medicine, Los Angeles, California.

 Presented at the annual meeting of the Association for Research in Vision and Ophthalmology, Fort Lauderdale, Florida, May 2001.

 Supported by Grant RO1-10595 from the National Eye Institute (MFL).

 Submitted for publication July 20, 2001; revised January 8, 2002; accepted February 4, 2002.

 Commercial relationships policy: N.

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

 

Figure 1.

 

The effect of H₂O₂ on DNA binding activity of AP-1 in human lens epithelium cells. (A) Putative AP-1–binding site containing double-stranded oligo from the human thioltransferase gene 5′ region. (B) Autoradiographic image of GMSA of AP-1 binding. Nuclear lysates were prepared from the cells treated for 0, 5, 10, 15, 30, 60, or 120 minutes with a bolus of 0.1 mM H₂O₂ and incubated with ³²P-labeled oligonucleotide containing a putative AP-1–binding site followed by analysis with GMSA. Lane C: a 100-fold molar excess of unlabeled specific oligonucleotide was added to the binding reaction. (C) Autoradiographic image of supershift assay with nuclear lysates from nontreated human lens epithelium cells. Total proteins (20 μg) from the nuclear lysates were incubated with anti-c-Jun antibody (lane 1) or anti-c-Fos antibody (lane 2), after incubation with ³²P-labeled oligonucleotide containing a putative AP-1–binding site, and analyzed with GMSA. Arrows: shifted bands in lanes 1 and 2 after positive complex was formed with anti-c-Jun and anti-c-Fos, respectively. Lane 3: no antibody. (D) The effect of direct oxidation, reduction and disulfide-reducing enzymes on DNA-binding activity of AP-1 in human lens epithelium cells. Total proteins of 20 μg from nuclear lysates were incubated with no additive (control, lane 1), 0.1 mM H₂O₂ (lane 2), 100 mM DTT (lane 3), 8.4 mU thioredoxin (lane 4), and 50 μU thioltransferase (lane 5) for 30 minutes at 30°C, after incubation with ³²P-labeled oligonucleotide containing a putative AP-1–binding site, and analyzed with GMSA. An autoradiographic image of the gel is shown.

Figure 2.

 

The effect of H₂O₂ on c-Jun and c-Jun N-terminal kinase (JNK) phosphorylation in human lens epithelial cells. Cellular lysates were prepared from the cells treated with bolus of 0.1 mM H₂O₂ for 0, 5, 10, 15, 30, or 60 minutes. Twenty micrograms of each lysate was separated in a 10% SDS-PAGE after the transfer to a nitrocellulose membrane and probing with anti-P-c-Jun or anti-phospho-JNK antibodies. The same membrane was reprobed with antibodies to a ubiquitous protein, G-3PD, to ensure that equal amounts of proteins were loaded on the gel. (A) P-c-Jun (top) and G-3PD control (bottom). (B) Phospho-JNK (top) and G-3PD control (bottom).

Figure 3.

 

The effect of H₂O₂ on phosphorylated p54/46 c-Jun N-terminal kinase (p-JNK) translocation to the nucleus in human lens epithelium cells. Nuclear lysates were prepared from the cells treated with a bolus of 0.1 mM H₂O₂ and incubated for 5, 10, 15, 30, 60, or 120 minutes, and (A) 20 μg of each nuclear lysate was separated in a 10% SDS-PAGE after transfer to a nitrocellulose membrane and probing with anti-phospho-JNK antibodies, which were not cross-reactive with nonphosphorylated JNK. (B) The same membrane was probed for nonphosphorylated p46 JNK to show that P-SAPK/JNK was transferred to the nuclei, but not phosphorylated in situ. (C) The same sample was reprobed with antibodies to a ubiquitous protein, G-3PD, to ensure that equal amounts of protein were loaded on the gel.

Figure 4.

 

Studies on the involvement of AP-1 cotransactivating factor, p35 Ref-1, in the HLE B3 cells. (A) Western blot with nuclear lysate from nontreated human lens epithelial cells and anti-Ref-1 antibody (lane 1). Twenty micrograms of nuclear lysate was separated in a 10% SDS- PAGE after the transfer to a nitrocellulose membrane and probing with anti-Ref-1 antibodies (arrow). Lane 2: protein molecular weight marker. (B) Ref-1/AP-1 interaction in HLE B3 cells under oxidative stress. Nuclear lysates were prepared from HLE B3 cells treated with 0.1 mM H₂O₂ for 0, 5, 10, 15, 30, or 60 minutes. Ref-1/AP-1 complexes from 200 lysate were immunoprecipitated using anti-c-Jun antibodies and separated on 8% SDS-PAGE gels. Western blot analysis was performed using anti-Ref-1 antibodies (top) or anti-c-Jun antibodies, to ensure equal loading (bottom).

Figure 5.

 

(A) PCR primer location in hTTase gene’s 5′ region and (B) the scheme of PCR fragments cloning in the reporter construct in front of the CAT gene. Plasmids pK-CAT/1190, pK-CAT/776, and pK-CAT/437 contained the AP-1–binding site. The binding site was mutated in pK-CAT/437Mut, and it was absent in pK-CAT/183.

Figure 6.

 

CAT assay with cell lysates from human lens epithelial cells transfected with reporter plasmids containing fragments of the hTTase gene’s 5′ region (Fig. 5) . HLE B3 cells were transfected with 2 μg of each plasmid, incubated for 24 hours, and treated with a bolus of 0.1 mM H₂O₂ for 0, 5, 15, or 30 minutes. The cells were lysed and ready for CAT assay. To assay for CAT enzyme activity, 50 μg of total cellular protein was incubated with ¹⁴C-labeled chloramphenicol in the presence of acetyl CoA for 2 hours, and the acetylated chloramphenicol was extracted with ethyl acetate and separated from other radioactive products of the reaction by thin-layer chromatography (TLC). The autoradiographic image of the TLC plates as shown represents a typical result from three separate experiments.

The authors thank Usha Andley of Washington University (St. Louis, MO) for providing the HLE B3 cells.