Mechanisms and Physiological Significance of the Transport of the Glutathione Conjugate of 4-Hydroxynonenal in Human Lens Epithelial Cells

Abstract

purpose. To study the mechanism and physiological significance of the transport of the glutathione (GSH) conjugate of 4-hydroxynonenal (4-HNE) from lens epithelial cells.

methods. HLE B-3 cells were treated with [³H] 4-HNE and efflux of its GSH conjugate, [³H] GS-HNE, into the medium was quantitated and characterized by HPLC and mass spectrometry. Inside-out vesicles (IOVs) were prepared from HLE B-3 cell membranes. The kinetics of adenosine triphosphate (ATP)-dependent uptake of GS-HNE and dinitrophenyl S-glutathione (DNP-SG) by these IOVs and inhibition of GS-HNE uptake by anti-RLIP76 IgG was studied. Localization of RLIP76 was studied by immunogold electron microscopy and kinetics of the adenosine triphosphatase (ATPase) activity of purified RLIP76 was determined. 4-HNE–induced apoptosis was compared in HLE-B3 cells coated with anti-RLIP76 IgG or preimmune IgG, by caspase activation assay.

results. The results showed the presence of RLIP76 in plasma membranes of HLE B-3 cells and that it mediated ATP-dependent transport of GS-HNE as well as of DNP-SG. The transport was saturable with respect to GS-HNE (K _m = 8.4 μM), DNP-SG (100 μM) as well as to ATP (K _m 0.45 mM) and was sensitive to temperature and osmolarity of the medium. Anti-RLIP76 IgG inhibited approximately 65% of the transport of GS-HNE and DNP-SG, indicating that most of the transport was mediated by RLIP76. Compatible with its transport function, the EGTA- and ouabain-insensitive ATPase activity of purified RLIP76 was stimulated by DNP-SG and GS-HNE. 4-HNE–induced caspase activation in HLE-B3 cells was exacerbated when the transport of GS-HNE from these cells was blocked by anti-RLIP76 IgG.

conclusions. RLIP76 provides a defense against the deleterious effects of 4-HNE by transporting GS-HNE and can modulate apoptotic signaling by regulating the intracellular concentrations of 4-HNE.

Oxidative stress has been implicated in many age-related diseases of the eye, including cataractogenesis, ^{1 2 3 4} retinal damage, and age-related macular degeneration (ARMD). ⁵ Lipid peroxidation, a consequence of oxidative stress, has been implicated in the etiology of these degenerative disorders. ^{6 7 8} For example, 4-hydroxynonenal (4-HNE), a toxic and stable end product of lipid peroxidation, has been implicated in cataractogenesis ⁴ and the pathophysiology of the retina. ^{5 8} Recent studies suggest that depending on its intracellular concentration, 4-HNE can differentially modulate cell cycle signaling. ^{9 10 11} Therefore, the mechanisms responsible for maintaining the intracellular levels of 4-HNE are important, not only in the defense of ocular tissues against oxidative stress but also in cell cycle signaling. ^{9 10 11} Intracellular levels of 4-HNE are primarily controlled through its metabolism and transport of its metabolites because its formation occurs through an uncontrolled process. One of the predominant mechanisms of 4-HNE metabolism is through its conjugation to glutathione (GSH) catalyzed by glutathione S-transferases (GSTs). ^{12 13 14} Even though, 4-HNE is also metabolized by aldehyde dehydrogenase to 4-hydroxy-2-nonenoic acid ¹⁵ and by aldose reductase to 4-hydroxy-2-nonenol, ^{3 16} these pathways account for the metabolism of only a minor fraction of intracellular 4-HNE. In human ocular tissues, a specific GST isozyme hGST5.8 has been characterized ^{17 18} that predominantly catalyzes the conjugation of 4-HNE with GSH, leading to formation of the conjugate (GS-HNE). 

We have shown that mild transient oxidative stress applied to cultured retinal pigmented epithelial cells leads to the induction of hGST5.8 and GSH synthesis, even before the induction of other antioxidant enzymes such as catalase, superoxide dismutases, and glutathione peroxidases. ¹⁸ This finding strongly suggests that an accelerated detoxification of 4-HNE through its conjugation to GSH is one of the early adaptive responses of these cells to oxidative stress. ^{17 18} It is known that relatively higher concentrations of 4-HNE lead to toxicity and apoptosis in cells of different tissue origin, including human lens epithelial cells. ^{9 10 11} During oxidative stress, 4-HNE levels are increased by enhanced lipid peroxidation. ¹⁸ For maintaining the intracellular physiological levels of 4-HNE, the conjugation of 4-HNE to GSH and subsequent transport of the conjugate GS-HNE seem to be crucial, because accumulation of GS-HNE inhibits GSTs. This suggests that the intracellular levels of 4-HNE in human lens epithelial cells must be regulated through a coordinated action of GSTs, which catalyze its conjugation to GSH, and transporters, which mediate the efflux of GS-HNE. Such a mechanism may be crucial for defense against the deleterious effects of lipid peroxidation. 

GSH-conjugates including GS-HNE are transported from cells through an ATP-dependent, primary active transport. ^{19 20 21 22 23} The mechanism(s) for the transport of GS-HNE from human ocular tissues have not been investigated. However, in bovine lens epithelial cells ATP-dependent transport of the GSH-conjugate of 1-chloro-2,4-dinitrobenzene (S-(dinitrophenyl) glutathione; DNP-SG) has been demonstrated that suggests that a transporter or transporters that catalyze the energy-dependent efflux of GSH conjugates are present in these cells. ²⁴ The molecular nature of the transporter(s) that mediate this process is not understood, and should be investigated because of its role in the defense mechanism against lipid peroxidation and xenobiotics. We have recently shown that RLIP76, a previously reported guanosine triphosphatase (GTPase)–activating Ral-binding protein, ^{25 26 27 28} accounts for most of the transport of GSH conjugates from human erythrocytes. ^{22 28}  

In the present studies we investigated the role of RLIP76 in the transport of GS-HNE and the GSH conjugates of xenobiotics from the human lens epithelial cell line HLE B-3. Inside-out vesicles (IOVs) prepared from the plasma membranes of these cells were used to measure the ATP-dependent transport of these compounds and to study the kinetics of the transport. To determine whether this transport was specifically mediated by RLIP76, we investigated whether the polyclonal antibodies raised against RLIP76 inhibit the transport of GS-HNE. Localization of RLIP76 in these cells was examined by immunogold electron microscopy. Furthermore, RLIP76 from HLE B-3 cells was purified and its kinetic properties investigated. To determine the physiological significance of GS-HNE transport, we studied the effect of blocking the transport of GS-HNE on oxidative stress-induced apoptosis in HLE B-3 cells. 

The results of these studies demonstrated the presence of RLIP76 in the human lens epithelial cell membrane, and it catalyzed most of the adenosine triphosphate (ATP)–dependent transport of GSH conjugates, including GS-HNE, from these cells. 4-HNE induced apoptosis in HLE B-3 cells, and blocking of the transport of GS-HNE by coating the cells with anti-RLIP76 IgG led to potentiation of the apoptotic effect of 4-HNE. These studies suggest that RLIP76, in conjunction with GSTs, plays a crucial role in regulating the intracellular concentration of 4-HNE and signaling for apoptosis. 

Materials and Methods

Chemicals

[Glycine-2-³H] GSH (specific activity, 44 Ci/mmol) was purchased from New England Nuclear (Boston, MA). 4-HNE was purchased from the Cayman Chemical Co. (Ann Arbor, MI). Minimum essential medium (MEM), phosphate-buffered saline, fetal bovine serum, and gentamicin (50 mg/mL) were purchased from Invitrogen/Gibco (Grand Island, NY). All other chemicals and reagents were obtained from Sigma-Aldrich (St. Louis, MO). 

Cell Line

The human lens epithelial cell line HLE B-3 used in the present studies was kindly provided by Usha P. Andley (Department of Ophthalmology and Visual Science, Washington University School of Medicine, St. Louis, MO). ²⁹ The cell line was maintained in minimum essential medium containing 20% fetal bovine serum and 50 μg/mL gentamicin at 37°C and in a 5% CO₂ atmosphere. 

Antibodies

Polyclonal antibodies against recombinant full-length RLIP76 expressed in Escherichia coli used during the present studies were the same as those described by us previously. ²⁶ An IgG fraction from the antiserum purified by DE-52 chromatography and protein A affinity chromatography was used in these studies. The specificity of these antibodies was stringently established by Western blot studies showing that the antibodies did not recognize any other protein except RLIP76 in the cellular extract. They also did not recognize the known ATP-binding cassette (ABC) transport proteins, multidrug-resistance–associated protein (MRP1), and P-glycoprotein (Pgp). 

Synthesis of DNP-SG

[³H]-labeled and unlabeled DNP-SG was synthesized by the GST-catalyzed conjugation of 1-chloro-2,4-dinitrobenzene and GSH, as described by us previously. ³⁰  

Synthesis of GS-HNE

[³H] 4-hydroxy-2-trans-nonenal was synthesized, purified, and authenticated by the method of Chandra and Srivastava, ³¹ using fumaraldehyde dimethyl acetal as the starting material. Purified [³H] 4-HNE (1 μmol; specific activity, 1 μCi/mmol) was subjected to enzymatic conjugation with GSH (1 μmol) by using our previously described method. ²² Crude GS-HNE thus formed was purified by preparative thin layer chromatography on silica gel G plates, using butanol, acetic acid, and water (4:1:1) as the mobile phase and further purified by HPLC, as described previously. ^{18 22} For the synthesis of unlabeled GS-HNE, a similar procedure was used, except that [³H] 4-HNE was replaced by 4-HNE. 

Measurement of the Efflux of GS-HNE from Cells

HLE B-3 (3–4 × 10⁵) cells were incubated with [³H] 4-HNE (20 μM; specific activity 3750 cpm/nmol) in complete growth medium for 30 minutes at 37°C. Intracellular GSH and oxidized glutathione (GSSG) concentrations in HLE B-3 cell lysate, before and after 4-HNE treatment, were measured by previously reported methods. ^{32 33} After incubation, the medium was removed, and the cells were washed three times with PBS to remove any residual radioactivity and incubated in 5 mL of PBS at 37°C. A 1-mL aliquot of the medium was withdrawn at different time points, immediately frozen at −70°C, and lyophilized. Residue thus obtained was resuspended in 200 μL of 70% ethanol and a fixed volume of the soluble material, collected after centrifugation, was mixed with 10 mL of liquid scintillation fluid (Ecolite; ICN, New Haven, CT) and counted on a liquid scintillation counter (model LS-6800; Beckman Coulter, Fullerton, CA) to ascertain the efflux of radioactivity above the background. 

For the identification of GS-HNE, 20 μL of the 70% ethanol supernatant was subjected to HPLC analysis on a C₁₈ reversed-phase chromatography column, as described by us previously. ¹⁸ The radioactivity associated with different peaks was counted. We have shown that GS-HNE elutes as a major peak at the retention time of 28.5 minutes. ¹⁸ Therefore, the peak associated with most radioactivity collected at the retention time of 28.5 minutes was subjected to mass spectrometry analysis. 

Effect of Anti-RLIP76 IgG on GS-HNE Efflux

To investigate the effect of anti-RLIP76 IgG on the efflux of GS-HNE from HLE B-3 cells, nonpermeabilized cells were preincubated with 20 μg of anti-RLIP76 IgG for 1 hour at 37°C in complete growth medium. The control group of cells was incubated with equal concentration of preimmune IgG protein. These IgG-coated cells were then treated with the [³H] 4-HNE in a manner similar to that described earlier, and the efflux of [³H] GS-HNE was compared in the cells coated with anti-RLIP76 IgG and control cells coated with preimmune IgG. 

Preparation of IOVs from Human Lens Epithelial Cell Membranes

Plasma membrane vesicles of human lens epithelial cells (1 × 10⁹) were prepared according to a method described by us. ^{22 34 35} Briefly, cells were harvested from the culture medium by centrifugation and lysed by incubating in hypotonic buffer (0.5 mM sodium phosphate [pH 7.0], containing 0.1 mM EDTA and 0.1 mM phenylmethylsulfonyl fluoride [PMSF]) for 1.5 hours, followed by homogenization. After centrifugation of homogenate at 12,000g (10 minutes at 4°C), the supernatant was further centrifuged at 100,000g for 40 minutes at 4°C. The resultant pellet was suspended in reconstitution buffer containing 250 mM sucrose in 10 mM Tris-HCl (pH 7.4), 100 μM PMSF, and 50 μM t-butylated hydroxytoluene (BHT) and homogenized in a glass homogenizer on ice. The resultant homogenate was diluted to 10 mL in the reconstitution buffer, layered on top of a 38% sucrose (wt/vol) solution in 5 mM HEPES-KOH (pH 7.4), and centrifuged at 100,000g for 2 hours at 4°C in a swinging bucket rotor. The turbid layer formed at the interphase was collected, diluted in the reconstitution buffer, homogenized in a glass–glass homogenizer, and centrifuged at 100,000g for 40 minutes at 4°C. The resultant pellet, consisting of a mixture of the IOVs and the right-side-out vesicles (ROVs) was resuspended in 2 mL of reconstitution buffer and passed 20 times through a 27-gauge needle attached to a syringe, for vesiculation. IOVs were separated from ROVs by passing the mixture of vesicles over a wheat germ agglutinin column, which selectively bound ROVs and resulted in the enrichment of IOVs recovered in the flow-through fraction. Aliquots of suspension were then stored at −80°C. Vesicle protein was measured by the Bradford method, ³⁶ and the sidedness of the IOVs was determined by the acetylcholinesterase assay of Ellman et al. ³⁷  

Transport Measurements

The ATP-dependent transport of [³H] GS-HNE and [³H] DNP-SG in the IOVs prepared as described earlier was measured by a rapid filtration method previously described by us. ^{34 35} Briefly, the reaction mixture (120 μL) consisted of 4 to 12 μg IOV protein, 10 mM Tris-HCl (pH 7.4), 250 mM sucrose, and 10 mM MgCl₂. After the reaction mixture was preincubated at 37°C for 5 minutes with appropriate concentrations of the radiolabeled substrates, the reaction was started by adding 4 mM ATP, and the mixtures were further incubated for different lengths of time at 37°C. In control samples, 4 mM of 5′-adenosine monophosphate (AMP) was used in place of ATP. The uptake was stopped by rapid filtration of a 30-μL aliquot of the reaction mixture using a 96-well nitrocellulose plate (0.45 μm pore size). The filters were washed with cold buffer and dried in air. The dried filters were punched out, and the radioactivity associated with the IOVs was measured by placing these in 10 mL of liquid scintillation fluid (Ecolite; ICN). Each determination was performed in triplicate. Nonspecific binding of the radiolabeled substrate to the filtration membrane was determined for each experiment and subtracted to obtain the vesicular uptake of the substrate. 

Inhibition of Transport in the IOVs by Anti-RLIP76 IgG

IOVs (4–8 μg protein) were coated separately with preimmune and anti-RLIP76 IgG by incubating the vesicles with different concentrations of IgG protein (0–1 μg) for 1 hour at room temperature. The transport of GS-HNE was measured in these vesicles by the rapid filtration method described earlier. 

In Situ Detection of Apoptotic Signaling through Caspase Activation in Cells

HLE B-3 cells (1 × 10⁴) were grown on a chamber slide (Laboratory-Tek; Nalge Nunc International, Naperville, IL) and then preincubated for 1 hour with preimmune IgG or anti-RLIP76 IgG, at a final concentration of 20 μg/mL. After the cells were washed with PBS, they were treated with different concentrations of 4-HNE at 37°C for 30 minutes. The apoptotic cells were detected by staining with 10 μM caspase FITC-VAD-FMK in situ marker (Promega, Madison, WI) for 30 minutes in the dark. This marker contains the FITC conjugate of the caspase inhibitor VAD-FMK, which irreversibly binds to activated caspases. The FITC label allows for a single reagent addition to assay caspase activity for the detection of apoptotic cells. ³⁸ The slide was rinsed with PBS two times and fixed with 4% paraformaldehyde for 1 hour. After which, the slide was mounted with medium containing 4′,6′-diamino-2-phenylindole (DAPI; 1.5 μg/mL) and observed by fluorescence microscope (Olympus, Tokyo, Japan). All micrographs were taken at × 400 magnification. 

Purification of RLIP76

RLIP76 was purified from HLE B-3 cells by using DNP-SG affinity chromatography as described by us previously. ^{26 39} Briefly, HLE B-3 (3 × 10¹⁰) cells were collected and washed with PBS. The cell pellet was suspended in 5 mL of lysis buffer (10 mM Tris-HCl [pH 7.4], 1.4 mM β-mercaptoethanol, 1 mM EDTA, 100 μM PMSF, and 50 μM BHT) containing 1% polidocanol, sonicated three times (50 W, 30 seconds) on ice, and incubated for 16 hours at 4°C with gentle shaking. The cell lysate was centrifuged at 28,000g for 30 minutes, and the supernatant was mixed with DNP-SG Sepharose resin (2 mL packed volume) and allowed to bind for 12 hours at 4°C. The unbound protein was removed by centrifugation, and the resin was washed several times with lysis buffer containing 0.01% SDS and 150 mM NaCl until nothing measurable at optical density of 280 nm (OD₂₈₀) was observed in the bath. The bound protein was eluted with lysis buffer containing 10 mM ATP, 10 mM MgCl₂, 2 mM DNP-SG, and 0.025% polidocanol, with gentle shaking at room temperature. The eluted enzyme was dialyzed against lysis buffer containing 0.025% polidocanol. The purity of RLIP76 isolated from HLE B-3 cells was checked by SDS-PAGE analysis on gels containing 12.5% polyacrylamide. Protein on the gels was visualized by staining with Coomassie brilliant blue dye R-250. For Western blot analysis, proteins from the gel were blotted to a nitrocellulose membrane by using the method of Towbin et al., ⁴⁰ as described by us previously. ²⁶  

ATPase Activity

Ouabain- and EGTA-insensitive ATPase activity in the crude membranes prepared from HLE B-3 cells as well as in the purified RLIP76 was measured, as previously described by us. ³⁹ Aliquots of protein fraction containing 0.4 to 80 μg of protein were added to a 0.5-mL reaction mixture containing 50 mM Tris-HCl, (pH 7.4), 10 mM MgCl₂, 2 mM EGTA, 0.8 mM sodium phosphate, 2.8 mM β-mercaptoethanol, and 1 mM ouabain, and incubated for 5 minutes at 37°C. The reaction was started by addition of 1.6 mM [γ³²P] ATP, without or with different transport substrates of RLIP76. After incubation for 30 minutes at 37°C, the reaction was terminated by addition of 2.5 mL of a cold mixture of 1 M perchloric acid and 5% ammonium molybdate in water (4:1) and extracted with a 2.5-mL mixture of isobutanol-benzene (1:1). Radioactivity was quantified in the organic phase to determine the phosphate group cleaved from ATP. Basal ATPase activity was calculated by subtracting the radioactivity measured in the absence of protein (blank) from that obtained in the presence of protein. Substrate-stimulated ATPase activity was calculated by subtracting basal ATPase activity estimated in the absence of stimulator from that measured in its presence. Each assay was performed in triplicate. 

Immunogold Electron Microscopic Studies

HLE B-3 cells (5 × 10⁴) were grown in T-25 flasks in complete growth medium, and the cell monolayers were fixed in a mixture of 2.5% formaldehyde and 0.1% glutaraldehyde in 0.05 M cacodylate buffer (pH 7.2) containing 0.03% trinitrophenol and 0.03% CaCl₂ for 1 hour at room temperature. After cells were washed in 0.1 M cacodylate buffer, they were scraped off the flask and pelleted. The IOVs prepared from the HLE B-3 cell membranes were treated in a similar manner. For conventional transmission electron microscopy, pellets were postfixed in 1% OsO₄ in 0.1 M cacodylate buffer, en bloc stained with 2% uranyl acetate, dehydrated in a graded series of ethanol and embedded in epoxy resin (Poly/Bed 812; Polysciences, Warrington, PA). Ultrathin sections were cut on an ultramicrotome (Ultracut S; Leica, Deerfield, IL) stained with lead citrate, and examined in a electron microscope at 60 kV (model 201 or CM-100; Philips, Mahwah, NJ). 

For postembedding immunoelectron microscopy, cell pellets were stained en bloc with 2% aqueous uranyl acetate, dehydrated in 50% and 75% ethanol, and embedded in medium grade resin (LR White; Structure Probe, Inc., West Chester, PA). Ultrathin sections were collected on polyvinyl formal (Formvar)-carbon–coated nickel grids (Structure Probe, Inc.), incubated on blocking buffer (0.1% BSA and 0.01 M glycine in 0.05 M TBS) and then with primary antibodies against RLIP76 diluted 1:20 in 1% BSA in 0.05 M TBS (diluting buffer) for 1 hour at room temperature, followed by overnight incubation at 4°C. After a wash in blocking buffer, the grids were incubated for 1 hour at room temperature with 1:20 goat anti-rabbit IgG secondary antibodies conjugated with 15 nM colloidal gold (AuroProbe EM GAR G15 RPN 422; Amersham Life Science, Arlington Heights, IL). After they were washed with TBS and distilled water, the grids were fixed in 2% aqueous glutaraldehyde, washed, stained with uranyl acetate and lead citrate, and examined in an electron microscope. 

Results

Efflux of GS-HNE from HLE B-3 Cells

To establish that HLE B-3 cells transport GS-HNE, the cells were loaded with [³H] 4-HNE, the efflux of radioactivity from the cells was measured, and the metabolite of 4-HNE transported in medium was identified by HPLC and mass spectrometry. For these experiments, cells were harvested, and a fixed number of cells were allowed to become attached in complete growth medium overnight. These cells were then loaded with [³H] 4-HNE, which readily forms [³H] GS-HNE within cells. This was confirmed by measuring the GSH content of cells before and after incubation with 4-HNE. Before incubation with [³H] 4-HNE, GSH concentration in HLE B-3 cells was found to be 91 ± 6.9 ng/mg protein, which was reduced to 32.83 ± 2.2 ng/mg protein after 10 minutes of incubation with 4-HNE, indicating its utilization for conjugating 4-HNE. The ratio of GSH to GSSG (95:5) observed in untreated control cells did not change significantly in the 4-HNE–treated cells (94.4:5.6). These results are consistent with those in our previous studies, ¹⁸ showing that when K562 cells are incubated with 4-HNE, GS-HNE is readily formed, and 4-HNE does not cause any significant oxidation of GSH. The cells loaded with [³H] GS-HNE were then used to measure the efflux of GS-HNE into the medium. After removing the medium, the cells loaded with [³H] GS-HNE were washed three times with PBS to remove extraneous radioactivity and resuspended in PBS. The results of the measurement of the efflux of radioactivity from these cells (Fig. 1) showed that the efflux rate reached its maximum during the initial 10 minutes (174 ± 29 pmol/min per 10⁶ cells), which gradually tapered off. 

To confirm that the radioactivity of the medium was associated with GS-HNE, HPLC analysis of the medium collected after incubation of cells at different time intervals was performed. The results of HPLC indicated the presence of a major peak of radioactivity at the retention time of 28.5 minutes, which coincided with that of a standard sample of GS-HNE (Fig. 2) . This peak was further confirmed to be that of GS-HNE by mass spectrometry analysis, which showed a molecular ion peak at mass/charge (m/e) 462.3 (M-1) due to removal of a proton with peaks at m/e 444 (M-OH, removal of hydroxyl ion) and m/e 306 for glutathionyl moiety (Fig. 2 , inset). The fragmentation pattern derived from the mass spectrum indicated the formation of the hemiacetal of 3-(4-hydroxynonanyl) glutathione, as reported by us previously. ¹⁸ A minor peak also appeared at m/e 466, which perhaps corresponds to GS-dihydroxynonene (GS-DHN), a reduced form of GS-HNE, as reported by Srivastava et al. ¹⁶ Together, these results showed that in HLE B-3 cells, 4-HNE was readily converted to GS-HNE, which was eventually transported out of the cells. 

Effect of Anti-RLIP76 IgG on GS-HNE Efflux from HLE B-3 Cells

We have shown ^{18 22} that most transport of GS-HNE in human erythrocytes is mediated by RLIP76, a GTPase-activating protein characterized earlier ²⁶ and that this transport can be blocked by coating the cells with anti-RLIP76 IgG. To examine the possible role of RLIP76 in the transport of GS-HNE from HLE B-3 cells, we examined the effect of anti-RLIP76 IgG on the efflux of GS-HNE from these cells by comparing the efflux of GS-HNE in cells coated with anti-RLIP76 IgG with that in the control cells coated with preimmune IgG. The results in Table 1 show that approximately 40% of transport of GS-HNE conjugate was inhibited in cells treated with anti-RLIP76 IgG, whereas preimmune IgG had no effect on the efflux of GS-HNE. These results suggest that a significant portion of the transport of GS-HNE from HLE B-3 cells is mediated by RLIP76. This was further confirmed by studies with IOVs prepared from HLE B-3 cells. 

Transport Studies with IOVs

Characterization of IOVs.

Although the results presented thus far suggest that HLE B-3 cells transport GS-HNE and that this transport is mediated by RLIP76, these experiments do not establish that the efflux is due to a primary active ATP-dependent transport. Because IOVs provide a convenient system to study the kinetics of ATP-dependent transport, ^{34 35 41} we studied the ATP-dependent uptake of GS-HNE in IOVs prepared from HLE B-3 membranes. The membrane vesicles prepared from HLE-B3 cells and the IOVs were purified by passing the mixture of these vesicles (a mixture of ROVs and IOVs) over a column of wheat germ agglutinin bound to cynogen bromide (CNBr)-activated Sepharose 4B, which selectively binds the ROVs due to the presence of carbohydrates on outer surface. The purity of IOVs collected from column flow-through was estimated by measuring acetylcholinesterase activity, which indicated that the preparation contained approximately 55% vesicles as IOVs. This was consistent with the results of previous investigations using IOVs for transport studies. ^{34 35} We established that these IOVs were morphologically competent for transport experiments by fixing an aliquot of the IOVs in 3% formaldehyde and subjecting it to scanning electron microscopy, which indicated (Fig. 3) that the preparation contained sealed vesicles with bilayer membranes. The average radius of these vesicles was found to be 52 ± 27 nm (n = 30), which corresponded to an average volume of 5.38 × 10⁻¹³ μL per vesicle. Consistent with previous studies ³⁵ these vesicles were judged to be suitable for examining the kinetics of ATP-dependent uptake. 

ATP-Dependent Uptake of Glutathione Conjugates by IOVs.

The ATP-dependent uptake of DNP-SG, a model GSH conjugate of an electrophilic xenobiotic, and of GS-HNE by these IOVs was measured. Data presented in Figure 4 indicate that, in the presence of 5′-AMP, [³H] GS-HNE uptake (1.74 ± 0.11 pmol/10 min of incubation period) by IOVs was observed and increased to 5.10 ± 0.15pmol/10 min in the presence of ATP. However, no significant change in [³H] GS-HNE uptake was observed when methylene-5′-adenosine triphosphate (Met-ATP), a non-hydrolyzable analogue of ATP, was used in the reaction mixture in place of ATP. These results indicate that [³H] GS-HNE uptake by IOVs is an ATP-dependent process. Our results also showed that the uptake of [³H] DNP-SG and [³H] GS-HNE by the IOVs was linear in IOV protein concentration and time, for a period up to 20 minutes (Figs. 5A 5B) . The initial rate of ATP-dependent uptake of GS-HNE was found to be approximately 340 pmol/min per milligram protein. The uptake of both these GSH conjugates by the IOVs was sensitive to temperature and osmolarity of the incubating medium, and maximum uptake was observed at 37°C and at a sucrose concentration of 250 mM. The transport was dependent on the intravesicular space of IOVs, because increased sucrose concentrations resulted in a decreased uptake of substrate due to shrinkage of vesicle volume (Fig. 5C) . The uptake of [³H] GS-HNE by the IOVs was saturable (Fig. 6A) with respect to its concentration as well as that of ATP. Similarly, the uptake of [³H] DNP-SG, a nonphysiological GSH conjugate, was saturable (Fig. 6B) . The apparent K _ms of DNP-SG and GS-HNE were 100 and 8.4 μM, respectively. The K _m of ATP for the uptake of GS-HNE was 0.45 mM. These results show that the transporters had high affinity for the endogenous substrate GS-HNE compared with the xenobiotic DNP-SG and suggest a physiological role of the transporters in regulating the intracellular concentrations of 4-HNE. 

Effect of Anti-RLIP76 IgG on ATP-Dependent Uptake of GS-HNE by IOVs.

We have shown that RLIP76-mediated uptake of GS- HNE by IOVs prepared from erythrocytes is inhibited by anti-RLIP76 IgG. ²² To establish the role of RLIP76 in transport, ATP-dependent uptake of GS-HNE was compared in the IOVs coated with anti-RLIP76 IgG and those coated with preimmune IgG. IOVs were coated with anti-RLIP76 IgG or preimmune IgG by incubating with different concentrations of IgG proteins for 1 hour at room temperature before transport measurements. Results of transport experiments presented in Figure 5D show that, whereas preimmune IgG had no noticeable effect on the ATP-dependent uptake of GS-HNE, anti-RLIP76 IgG inhibited uptake in a dose-dependent manner, with maximum inhibition of approximately 65%. These results demonstrate that a major portion of ATP-dependent transport of GS-HNE in HLE-B3 cells was mediated by RLIP76. It is to be noted that inhibition of transport by anti-RLIP76 IgG in IOVs was significantly more than that observed with intact HLE B-3 cells. It is possible that as opposed to round, suspended IOVs, the attached cells were not uniformly coated by the antibodies, and consequently the inhibition of GS-HNE transport in HLE B-3 cells in situ by anti-RLIP76 IgG was less than that in the IOVs. 

Effect of the Blocking of GS-HNE on 4-HNE-Induced Caspase Activation for Apoptosis

It is known that 4-HNE induces apoptosis in a variety of human cell lines. ^{9 10 18} To determine whether 4-HNE causes apoptosis in HLE B-3 cells, and if so, whether the effect of 4-HNE is exacerbated by anti-RLIP76 IgG, we compared 4-HNE–induced apoptosis in HLE B-3 cells coated with preimmune IgG and RLIP76 IgG. The apoptotic cells were detected with an FITC-conjugated fluorescent marker of caspase, by fluorescence microscopy as described previously. ³⁸ Treatment with an increasing concentration of 4-HNE showed an increase in caspase activity suggesting a dose-dependent apoptotic effect of 4-HNE (Fig. 7) . Whereas 10 μM 4-HNE caused no apparent activation of caspase, measurable caspase activity was observed in cells treated with 15 μM 4-HNE, which increased in cells treated with 30 μM 4-HNE. Cells coated with preimmune IgG and treated with 4-HNE under identical conditions showed activation of caspase to a similar extent. However, when efflux of GS-HNE was blocked by coating the cells with anti-RLIP76 IgG, a strong exacerbation of 4-HNE–induced activation of caspase in HLE B-3 cells was observed indicating apoptosis even in the presence of 10 μM 4-HNE (Fig 8) . The results showed that blocking of the GS-HNE efflux by anti-RLIP76 IgG led to activation of caspase and eventual apoptosis. 

Purification and Characterization of RLIP76 from HLE B-3 Cells

We have shown that both recombinant and purified tissue RLIP76 contains ouabain- and EGTA-insensitive ATPase activity, which provides energy for transport. ^{26 27} This activity is stimulated by the substrates, or allocrites, of RLIP76 including DNP-SG, doxorubicin, GS-HNE, and leukotrienes. ^{22 23} Therefore, ouabain- and EGTA-insensitive ATPase activity in crude preparations of RLIP76 in HLE B-3 cell membranes as well as in purified of RLIP76 was measured, and the effect of the transport substrates of RLIP76 (DNP-SG, DOX, GS-HNE, and LTC4) on this activity was studied. RLIP76 was purified from these cells by DNP-SG affinity chromatography, as described previously, for the purification of bacterial lysates ²⁶ and human erythrocytes. ⁴² The homogeneity of the preparation was established by SDS-PAGE, Western blot analyses, N-terminal peptide sequencing, and amino acid composition, as described previously ²⁶ (data not shown). The results presented in Table 2 indicate that both crude membranes and purified preparation of RLIP76 had ouabain- and EGTA-insensitive ATPase activity, which was stimulated by its transport substrates. The basal ATPase activity of crude membrane was found to be 13 ± 3.30 nmol/min milligram protein whereas the specific activity of purified RLIP76 was found to be 149 ± 11.25 nmol/min per milligram protein. The stimulation of ATPase activity of the purified RLIP76 was found to be about twofold in the presence of the xenobiotic substrates DNP-SG and DOX. ^{26 27} The physiological substrates GS-HNE and LTC4 stimulated this activity by approximately threefold in crude membranes (Table 2) . LTC4 stimulated the ATPase activity of the crude membranes to a greater extent than did the purified preparation of RLIP76. The reasons for this were not explored during the present studies. These results confirm the presence of RLIP76 in these cells and that the kinetic properties related to transport function are similar to those reported previously ^{22 26 27 42} for RLIP76 purified from human erythrocytes. These results suggest that in human lens epithelial cells, RLIP76 is involved in the transport of GSH conjugates of xenobiotics as well as of physiologically important GSH conjugates, such as GS-HNE and leukotrienes. 

Immunolocalization of RLIP76

The results of our inhibition experiments with anti-RLIP76 IgG in intact HLE B-3 cells indicated the presence of RLIP76 epitopes on the plasma membrane surface. Therefore, we performed immunogold studies by electron microscopy with polyclonal antibodies against RLIP76 used to determine its localization. The studies showed the localization of RLIP76 in the plasma membrane, indicated by the frequent appearance of gold particles as clusters on the membrane (Fig. 9) . RLIP76 also was associated with membrane invaginations immediately adjacent to the plasma membrane, as shown in Figure 9B . The significance of this is not clear to us, but it could be relevant to the previously reported involvement of RLIP76 in formation of an exocyst complex. ⁴³  

Discussion

ATP-dependent transport of the GSH conjugates of electrophilic xenobiotics is essential for the protection of cells from these toxic compounds and is considered to be phase III of the detoxification process. ⁴⁴ The mechanisms responsible for the transport of these conjugates from the cells of ocular tissues are largely unknown. The present studies demonstrated the ATP-dependent transport of GS-HNE and DNP-SG from human lens epithelium and for the first time defined a distinct molecular mechanism responsible or this process. We demonstrated the efflux of the conjugate of 4-HNE and GSH from HLE-3B cells and showed that it can be inhibited by anti-RLIP76 IgG, suggesting that RLIP76 is the major transporter that mediates the efflux of GS-HNE. Furthermore, we characterized the kinetics of this transport by using IOVs. 

IOVs prepared from the plasma membranes provide a convenient system for the study of the kinetics of ATP-dependent transport, in which the concentrations of ATP and the transport substrates (allocrites) can be varied as opposed to the intact cells in which these parameters cannot be manipulated. The results of the present studies clearly demonstrated that the IOVs prepared from HLE B-3 cells catalyzed the ATP-dependent uptake of GS-HNE as well as of DNP-SG. The transport of these compounds in IOVs was driven by energy from hydrolysis of ATP, because the nonhydrolyzable analogue of ATP did not support the transport. The transport was saturable with respect to both ATP and the GSH conjugate concentrations and its optimal temperature was 37°C. At higher temperatures, the transport activity was abolished, suggesting that the transport is a protein-dependent physiological process. The ATP-dependent uptake of GS-HNE in IOVs is concentrated and is dependent on intravesicular space, as indicated by a diminished uptake at higher sucrose concentrations, resulting in an increased osmolarity and decreased intravesicular volume of the IOVs. 

The transport of DNP-SG has been demonstrated ²⁴ in bovine lens epithelium, but the molecular mechanisms responsible for the transport have not been characterized. Our studies demonstrate for the first time that in human lens epithelial cells, the ATP-dependent transport of the GSH conjugates of xeno- as well as endobiotics is mediated by RLIP76. As reported by us previously, ^{26 28} GSSG was not found to be a substrate of RLIP76. Our results show that approximately two thirds of the total transport of GS-HNE or DNP-SG in IOVs is blocked by anti-RLIP76 IgG, which indicates that RLIP76 is the major transporter of GSH conjugates in these cells. These results suggest that other transporters besides RLIP76 also contribute to a significant (about one third) portion of the transport activity. During the present studies, we did not examine the factors responsible for the residual transport activity. Further studies are needed to define these mechanisms. It has been shown that MRP1 can also mediate the transport of GS-HNE. ⁴⁵ It is possible that MRP1 or other transporters of the ABC protein family may contribute to the residual transport of GSH conjugates in these cells. Our results demonstrating the presence of RLIP76 in HLE-B3 membranes and that its EGTA- and ouabain-insensitive ATPase activity is stimulated in the presence of GSH conjugates are consistent with its function as a GSH conjugate transporter in these cells. It is to be noted that the K _m of RLIP76 for the ATP-dependent transport of GS-HNE (5 μM) and LTC₄ (100 nM) is much lower than that of the xenobiotic substrate DNP-SG (100 μM), suggesting a role of RLIP76 in the transport of the physiological substrates such as leukotrienes, which must be transported out of the cell to function, because their receptors are located on the cellular surface. 

The results of present studies also shed light on the physiological significance of RLIP76-mediated transport of GSH conjugates in protective mechanisms against lipid peroxidation and 4-HNE–mediated signaling in apoptosis. 4-HNE, the reactive but stable end product of lipid peroxidation, is highly toxic ⁴⁶ and has been implicated in oxidative stress-induced cataractogenesis. ^{4 8} 4-HNE affects cell cycle signaling, and at higher concentrations it causes apoptosis in a variety of cells from diverse origins. ^{11 47 48} Its intracellular concentration during oxidative stress is controlled through its metabolism and the transport of metabolites. ^{16 18} 4-HNE is mainly detoxified by its conjugation to GSH to form GS-HNE by the reaction catalyzed by GSTs, and the presence of hGST5.8, an isozyme with substrate preference for 4-HNE, has been demonstrated in human ocular tissues. ^{17 18} As indicated by our studies, HLE B-3 cells readily convert 4-HNE to GS-HNE without having any significant effect on the ratio of GSH to GSSG in these cells. However, it is known that GSH conjugates inhibit GSTs and, to sustain the GST-mediated conjugation of 4-HNE, the conjugate (GS-HNE) must be transported from cells. The results of the present studies show that GS-HNE and other GSH-conjugates are transported by HLE B-3 cells in an ATP-dependent manner and that this transport is mediated by RLIP76. Our results strongly suggest that in coordination with GSTs, RLIP76 contributes to detoxification of 4-HNE and plays a major role in defense mechanisms against the oxidative stress that leads to lipid peroxidation and the toxic effects of 4-HNE. This idea is supported by our results showing that blockage of the RLIP76-mediated efflux of GS-HNE by anti-RLIP76 IgG exacerbates the apoptotic effect of 4-HNE. Thus, RLIP76-mediated transport of GSH conjugates appears to be crucial for protecting cells against the deleterious effect of 4-HNE generated during lipid peroxidation. 

Our previous studies have shown that during oxidant-induced cataractogenesis apoptosis is observed in lens epithelial cells and that this may be caused by 4-HNE. ^{49 50} The finding in the present studies that 4-HNE induced apoptosis in HLE B-3 cells are consistent with this idea. The results showing exacerbation of 4-HNE–induced apoptosis by anti-RLIP76 further suggest that increased concentrations of 4-HNE promotes apoptosis in HLE B-3 cells and that RLIP76 in conjunction with hGST5.8 keeps the concentration of 4-HNE within the threshold of toxicity. These observations are also consistent with the results of our previous studies that show that the induction of hGST5.8 and RLIP76 in cells is an early adaptive response of K562 cells to oxidative stress. ¹⁸ Taken together, the results of the present studies underscore the significance of RLIP76 in the defense of HLE B-3 cells against oxidative stress and particularly against the toxic effects of 4-HNE. By transporting GS-HNE, RLIP76 sustains GST-mediated conjugation of 4-HNE to GSH and in the absence of this transport function of RLIP76, accumulation of GS-HNE causes inhibition of GST, resulting in increased concentrations of intracellular 4-HNE. Furthermore, RLIP76 can provide protection against xenobiotics, as its substrate specificity for pharmacological agents appears to be wider than that of MRP or Pgp. ²⁸  

Previous studies have shown that GS-HNE can be further metabolized to its corresponding alcohol GS-DHN by aldose reductase and GS-DHN has been suggested to be involved in signaling processes. ^{15 16} The present studies were primarily focused on the transport of GS-HNE, and we did not explore the subsequent metabolism of this conjugate. We have identified only GS-HNE as the major compound transported in the medium, but our results show several minor peaks of radioactivity indicating formation of GS-DHN and other metabolites of GS-HNE. These minor metabolites were not further characterized during the present studies. Nevertheless, our results clearly demonstrate that GSH conjugates of 4-HNE and/or its metabolites are transported by RLIP76, which plays an important role in regulating intracellular concentrations of 4-HNE and is a key component of the defense mechanisms against oxidative stress in lens epithelial cells. Furthermore, RLIP76 protects cells from electrophilic xenobiotics by transporting their GSH conjugates and may also be relevant to the functions of physiological GSH conjugates such as leukotrienes, which must be transported out of cells to perform their biological functions. 

 Supported in part by National Eye Institute Grants EY04396, GM 32304 (YCA), CA77495 (SA), and HL65416 (PJB).

 Submitted for publication January 16, 2003; revised March 17, 2003; accepted March 28, 2003.

 Disclosure: R. Sharma, None; Y. Yang, None; A. Sharma, None; S. Dwivedi, None; V.L. Popov, None; P.J. Boor, None; S.S. Singhal, None; S. Awasthi, None; Y.C. Awasthi, 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 1.

 

Time course of the efflux of [³H] GS-HNE from HLE B-3 cells: HLE B-3 (2 × 10⁵) cells were treated with [³H] 4-HNE (20 μM; specific activity, 3750 cpm/nmol) for 30 minutes at 37°C. After incubation, cells were washed with PBS three times to remove any unbound radioactivity. The cells were then incubated at 37°C in PBS (5 mL), and aliquots were withdrawn at different time points and freeze dried. The residue was dissolved in 200 μL of 70% ethanol, and the aliquots of the ethanol-soluble fraction were counted in triplicate. Data are the mean ± SD of results in three separate experiments.

Figure 2.

 

Identification of [³H] GS-HNE in the medium after incubating the cells with [³H] 4-HNE. The ethanol-soluble fraction, obtained as described in Figure 1 , was subjected to HPLC analysis on a reversed-phase C₁₈ column (3.9 × 150 mm) using a linear gradient from 0.1% trifluoroacetic acid [TFA] to 100% acetonitrile containing 0.1% TFA in 60 minutes at a flow rate of 1 mL/min. One-millimeter fractions were collected and counted for radioactivity. Maximum radioactivity was found to be associated with [³H] GS-HNE eluted at the retention time of 28.5 minutes. The HPLC profile of GS-HNE was isolated from the medium and superimposed on the radioactivity profile. The elution of GS-HNE was confirmed by coelution of an authentic sample of this compound as well as by mass spectrometry analysis (inset).

Table 1.

 

Efflux of GS-HNE by HLE B-3 Cells Coated with Preimmune and RLIP76 IgG

Treatment	GS-HNE Efflux (nmol/106 cells per 10 min ± SD)
None (control)	1.74 ± 0.29
Preimmune IgG	1.68 ± 0.31
RLIP76 IgG	1.04 ± 0.19

Figure 3.

 

Electron micrographs of IOVs prepared from HLE B-3 cell membranes. (A) IOVs derived from human lens epithelial cells consisted of regular-sized vesicles (v) with bilayered membranes and some membrane fragments present. (B) Higher power of vesicles shows the bilayered nature of lipid membranes indicated by (arrows). Bar: (A) 0.5 μm; (B) 0.1 μm.

Figure 4.

 

ATP-dependent uptake of [3H] GS-HNE by IOVs prepared from HLE B-3 cells was compared in the presence of AMP, ATP, and Met-ATP. IOVs (8 μg protein) were preequilibrated with [³H] GS-HNE (10 μM, specific activity, 100 cpm/pmol) in transport buffer for 5 minutes at 37°C, and transport was initiated by adding equimolar (4 mM) portions of AMP, ATP, and Met-ATP separately to the reaction mixture. Aliquots of reaction mixture (30 μL) were filtered in triplicate after 10 minutes, and the radioactivity remaining on the nitrocellulose filters in the 96-well plate was quantified. Data are expressed in picomoles per 10 minutes (mean ± SD, n = 3).

Figure 5.

 

(A) Dependence of [³H] GS-HNE transport on vesicle protein concentrations. The transport buffer containing 10 μM [³H] GS-HNE and IOVs (2–12 μg protein) was allowed to equilibrate for 5 minutes at 37°C, and transport was initiated by addition of 4 mM of either ATP or AMP. Aliquots of reaction mixture were filtered after 10 minutes, and the radioactivity remaining on the nitrocellulose filter of each well in a 96-well plate was quantified. ATP-dependent uptake by IOVs was determined by subtracting uptake observed at 10 minutes after addition of 4 mM AMP from uptake observed after addition of 4 mM ATP. (B) The time course of the ATP-dependent uptake of [³H] GS-HNE by IOVs. (C) Effect of osmolarity on the transport of [³H] GS-HNE. (D) Inhibition of GS-HNE transport by anti-RLIP76 IgG. IOVs (8 μg) were coated separately with anti-RLIP76 IgG and preimmune IgG (0–1 μg IgG protein), and ATP-dependent transport of GS-HNE was compared in these IgG-coated vesicles. Data are the mean ± SD of results in two separate experiments performed in triplicate.

Figure 6.

 

Kinetics of ATP-dependent transport in HLE B-3 IOVs with varying concentrations of (A) GS-HNE (1–20 μM) and (B) DNP-SG (25–200 μM). Uptake of both the substrates by IOVs was separately measured as described in Figure 2 .

Figure 7.

 

4-HNE induced activation of caspase in HLE B-3 cells. HLE B-3 cells (1 × 10⁴) were grown on chamber slides. Cells were treated with 4-HNE (5–30 μM) at 37°C for 30 minutes. The apoptotic cells were detected by staining with 10 μM of the caspase FITC-VAD-FMK in situ marker for 30 minutes in the dark. The slides were rinsed with PBS two times, fixed in 4% paraformaldehyde for 1, mounted with medium containing DAPI (1.5 μg/mL), and observed under a fluorescence microscope. All photographs were taken at ×400 magnification. (A–D) Nuclear staining of cells with DAPI (E–H) green fluorescence of activated caspase observed in apoptotic cells after treatment with different concentrations of 4-HNE.

Figure 8.

 

Exacerbation of 4-HNE–induced caspase activation in HLE B-3 cells coated with anti-RLIP76 IgG. HLE B-3 cells (1 × 10⁴) were separately coated with preimmune IgG and anti-RLIP76 IgG (20 μg/mL) for 1 hour at 37°C. After a wash in PBS, cells were treated with 4-HNE (10 μM) for 30 minutes. Activation of caspase in preimmune and anti-RLIP76–coated IgG after treatment with 4-HNE was detected by the caspase FITC-VAD-FMK in situ marker, as described in Figure 7 . (A) Nuclear staining with DAPI in preimmune IgG-coated cells. Nuclear staining of cells in anti-RLIP76–coated cells after treatment with 4-HNE. (B) Green fluorescence observed due to activation of caspase in preimmune IgG-coated cells. Green fluorescence observed in cells coated with anti-RLIP76 IgG, after treatment with 4-HNE.

Table 2.

 

ATPase Activity in the Membrane Vesicles and Purified RLIP76 from HLE B-3 Cells and Stimulation by Different Substrates of RLIP76

Source of RLIP76	ATPase Activity (nmol/min per mg protein ± SD)
Membrane vesicles	13.0 ± 3.31
Membrane vesicles+DNPSG	19.0 ± 1.62
Membrane vesicles+GS-HNE	48.3 ± 1.5
Purified RLIP76 from HLE B-3 cells	148.2 ± 11.25
Purified RLIP76+DNP-SG	341 ± 8.78
Purified RLIP76+GS-HNE	419 ± 16.42

Figure 9.

 

(A) Localization of RLIP76 to the plasma membrane of human lens in two adjacent epithelial cells. Collections of gold particles localized on or immediately adjacent to the membrane, and frequently appeared in clusters (arrow). (B) RLIP76 also appeared to localize immediately adjacent to the plasma membrane and was associated with membrane invaginations Bar: (A) 0.1 μm; (B) 0.5 μm.