Accelerated metabolism and exclusion of 4-hydroxynonenal through induction of RLIP76 and hGST5.8 is an early adaptive response of cells to heat and oxidative stress.

To explore the role of lipid peroxidation (LPO) products in the initial phase of stress mediated signaling, we studied the effect of mild, transient oxidative or heat stress on parameters that regulate the cellular concentration of 4-hydroxynonenal (4-HNE). When K562 cells were exposed to mild heat shock (42 degrees C, 30 min) or oxidative stress (50 microM H2O2, 20 min) and allowed to recover for 2 h, there was a severalfold induction of hGST5.8, which catalyzes the formation of glutathione-4-HNE conjugate (GS-HNE), and RLIP76, which mediates the transport of GS-HNE from cells (Awasthi, S., Cheng, J., Singhal, S. S., Saini, M. K., Pandya, U., Pikula, S., Bandorowicz-Pikula, J., Singh, S. V., Zimniak, P., and Awasthi, Y. C. (2000) Biochemistry 39, 9327-9334). Enhanced LPO was observed in stressed cells, but the major antioxidant enzymes and HSP70 remained unaffected. The stressed cells showed higher GS-HNE-conjugating activity and increased efflux of GS-HNE. Stress-pre-conditioned cells with induced hGST5.8 and RLIP76 acquired resistance to 4-HNE and H2O2-mediated apoptosis by suppressing a sustained activation of c-Jun N-terminal kinase and caspase 3. The protective effect of stress pre-conditioning against apoptosis was abrogated by coating the cells with anti-RLIP76 IgG, which inhibited the efflux of GS-HNE from cells, indicating that the cells acquired resistance to apoptosis by metabolizing and excluding 4-HNE at a higher rate. Induction of hGST5.8 and RLIP76 by mild, transient stress and the resulting resistance of stress-pre-conditioned cells to apoptosis appears to be a general phenomenon since it was not limited to K562 cells but was also evident in lung cancer cells, H-69, H-226, human leukemia cells, HL-60, and human retinal pigmented epithelial cells. These results strongly suggest a role of LPO products, particularly 4-HNE, in the initial phase of stress mediated signaling.

4-Hydroxy-t-2,3-nonenal (4-HNE), 1 a highly reactive but relatively stable end product of lipid peroxidation (LPO), has drawn a great deal of attention in recent years because of its possible involvement in signaling mechanisms. 4-HNE has been shown to cause apoptosis (1)(2)(3)(4)(5), differentiation (6 -9), and induction of enzymes including c-Jun kinase (JNK)/stress-activated protein kinase (1,10), protein kinase C (11), adenylate cyclase (12), and phospholipase C (13). Available evidence suggests that depending upon its intracellular concentration, 4-HNE may differentially affect the cell cycle regulation. For example, it has been shown that at low concentrations, 4-HNE causes proliferation of aortic smooth muscle cells (14) and K562 cells (6), but at relatively higher concentrations, it causes differentiation and apoptosis in these cells (6). Thus, the intracellular concentrations of 4-HNE must be stringently controlled. Because the formation of 4-HNE results from LPO, an uncontrolled process depending on the levels of cellular redox status, the intracellular levels of 4-HNE must be controlled through its metabolism and elimination of the metabolites from cells. Glutathione S-transferase (GST) mediated conjugation of 4-HNE to glutathione (GSH), resulting in the formation of the GSHconjugate (GS-HNE), is the major pathway for its metabolism (15). GSH conjugates are known to be transported out of cells through an ATP-dependent primary active efflux mechanism (16 -19). In humans, GST isozymes designated as hGST5.8 (20 -22) and hGSTA4-4 2 (23,24) preferentially conjugate 4-HNE to GSH, and GS-HNE thus formed is transported across the membrane by transport proteins, including RLIP76, a Ral binding GTPase-activating protein (25), which has been shown to account for the ATP-dependent transport of GS-HNE in K562 cells and human erythrocytes (26 -29). Thus, the levels of expression and activities of hGST5.8 or/and hGSTA4-4 and RLIP76 may be the major determinants of the intracellular concentrations of 4-HNE.
In aerobic organisms, reactive oxygen species such as O 2 Ϫ , H 2 O 2 , OH ⅐ are continually generated, and their overproduction during stress (oxidative, chemical, heat) conditions causes adverse effects. Under most of these stress conditions LPO is increased with an expected increase in the formation of 4-HNE. Pro-apoptotic agents such as H 2 O 2 or a variety of xenobiotics (e.g. doxorubicin) also lead to the generation of reactive oxygen species and induce LPO, suggesting a possible role of LPO products in apoptotic signaling. This idea is consistent with the results of our recent studies showing that attenuation of LPO in K562 by transfection with hGSTA2-3, which specifically reduces lipid hydroperoxides, blocks H 2 O 2 -induced apoptosis in these cells (30). A clear link between LPO and stress-mediated apoptotic signaling is, however, not established. We reasoned that if 4-HNE derived from LPO is involved in stress-mediated signaling for apoptosis, then one of the early responses of the cells may be to eliminate 4-HNE through its enhanced conjugation to GSH catalyzed by hGST5.8 and the subsequent transport of the conjugate, GS-HNE, by RLIP76. Therefore, during the present studies we have examined the effect of relatively low levels of heat (30 min, 42°C) and oxidative (50 M H 2 O 2 , 20 min exposure) stress on the expression and functions of hGST5.8 and RLIP76. K562 cells were chosen for these experiments because these cells constitutively express hGST5.8 (31) as well as RLIP76 and, when transfected with RLIP76, show enhanced efflux of GS-HNE (27). Results of these studies demonstrate that the induction of RLIP76 and hGST5.8 is an early adaptive response of cells exposed to mild heat shock (42°C, 30 min) or low levels of transient oxidative stress (50 M H 2 O 2 , 20 min), and the cells pre-exposed to these stress conditions suppress H 2 O 2 -and 4-HNE-mediated activation of JNK and caspase 3 and acquire resistance to apoptosis. These findings suggest that hGST5.8 and RLIP76 play regulatory roles in mechanisms in stress-mediated apoptosis.
Antibodies-Polyclonal antibodies raised in rabbits against the Alpha, Mu, and Pi classes of human GSTs were the same as used in our previous studies (30). Polyclonal antibodies against recombinant hG-STA4-4 were raised in chicken, and the specificity of these antibodies was stringently established (22). Polyclonal antibodies against recombinant mGSTA4-4, the mouse ortholog of hGST5.8, were raised in rabbits, and their specificity only to hGST5.8 among human GSTs has been established (20 -22, 32). Polyclonal antibodies raised in rabbit against recombinant RLIP76 were the same as these used in our previous studies (27,28). IgG fractions from all these antibodies purified through DE-52 and protein A-agarose columns were used. Anti-heat shock protein 70 (HSP70) goat polyclonal antibodies (K20) was obtained from Santa Cruz Biotechnology (Santa Cruz, CA). The anti-phospho-c-Jun (Ser-63) antibodies and GST-c-Jun-(1-89) fusion protein were obtained from New England Biolabs, Inc. Antibodies against poly(ADPribose) polymerase (PARP) (H-250) were obtained from Santa Cruz Biotechnology (Santa Cruz, CA), and those against caspase 3 were obtained from Pharmingen, San Diego, CA. Antibodies against ␤-actin were purchased form Sigma.
Cell Lines and Cultures-Human leukemia K562 and HL-60, human small cell lung cancer cell line H-69, and human non-small cell lung cancer cell line H-226 were obtained from the American Type Culture Collection (ATCC, Manassas, VA). All cultures were maintained at 37°C in a humidified atmosphere of 5% CO 2 and 95% air. K562, HL-60, and H-69 cells were grown as suspension cultures in RPMI 1640 medium supplemented with 10% (v/v) fetal calf serum and 1% penicillin/ streptomycin. Monolayer cultures of H-226 cells were maintained in RPMI 1640 medium supplemented with 10% (v/v) heat-inactivated fetal bovine serum, 1% (v/v) penicillin/streptomycin solution, 2 mM L-glutamine, 10 mM HEPES, 1 mM sodium pyruvate, 4.5 g/liter glucose, and 1.5 g/liter sodium bicarbonate. Cultures of simian virus 40-transformed fetal male retinal pigmented epithelial (RPE) cells (33) were the same as those used in our previous studies (34) and were grown in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, 1% (v/v) penicillin/streptomycin, 10 mM HEPES, pH 7.4. The cells were trypsinized and passaged every 5 days.
Heat Shock and H 2 O 2 Treatment-For heat shock the cells were exposed to 42°C for different time periods, brought back to 37°C, and allowed to recover for 2 h at 37°C before use in further experiments. For the transient exposure to H 2 O 2 , the cells were treated with 50 M H 2 O 2 in the medium for 20 min, after which cells were pelleted, washed with H 2 O 2 -free medium, resuspended in medium, and allowed to recover for 2 h at 37°C.
SDS-PAGE and Western Blot Analysis-The cells (50 ϫ 10 6 ), after specified treatment(s), were pelleted, washed, resuspended in 10 mM Tris-HCl, pH 7.4, containing 0.5 mM phenylmethylsulfonyl fluoride, 1 mM dithiothreitol, 1 g/ml leupeptin, and lysed by sonication. Cell lysates containing 25-100 g of proteins were subjected to SDS-polyacrylamide gel electrophoresis in 12% gels according to the method of Laemmli (35). Western blot analysis was performed essentially according to the method of Towbin et al. (36). Immunoblots were developed with the ECL (chemiluminescence) reagents from Pierce by following the manufacturer's instructions. Throughout these studies protein was determined by the method of Bradford (37).
Lipid Peroxidation-LPO was measured by determining the thiobarbituric acid-reactive substance as described by Wagner et al. (38). For each determination, 1 ϫ 10 7 cells collected by centrifugation at 500 ϫ g for 10 min and washed twice with PBS were used. Cells were resuspended in 1 ml of 10 mM potassium phosphate buffer, pH 7.0, containing 0.4 mM butylated hydroxytoluene, vortexed vigorously, and immediately used for thiobarbituric acid-reactive substance assay.
Determination of Intracellular Malonaldehyde and 4-HNE Levels-Malonaldehyde (MDA) and 4-HNE levels were determined using Biotech LPO-586 TM kit (Oxis International, Portland, OR) according to the manufacturer's instructions. For each determination, 5 ϫ 10 7 cells were collected by centrifugation at 500 ϫ g for 10 min and washed twice with PBS. The harvested cells were resuspended in 0.2 ml of 10 mM potassium phosphate buffer, pH 7.0, containing 4 mM butylated hydroxytoluene and vortexed vigorously. To each sample, 650 l of N-methyl-2phenylinodole and 150 l of 12 N HCl (for MDA determination) or 15.4 M methanesulfonic acid (for 4-HNE plus MDA determination) were added, and the reaction mixture was vortexed and incubated at 45°C for 60 min. The reaction mixture was centrifuged at 15,000 ϫ g for 10 min, and absorbance of the supernatant was read at 586 nm. Standards of MDA or 4-HNE were prepared from the hydrolysis of 1,1,3,3-tetramethoxypropane in HCl or 4-HNE diethylacetal in methanesulfonic acid, respectively. An extinction coefficient of 1.1 ϫ 10 5 M Ϫ1 cm Ϫ1 , determined from the standard curves of MDA and 4-HNE, was used, and the values were expressed as pmol of MDA or 4-HNE/mg of protein.
GST Purification-K562 cells (50 ϫ 10 6 ) treated with heat shock or H 2 O 2 as described above were harvested by centrifugation and washed with PBS. The cells were resuspended in 10 mM potassium phosphate buffer, pH 7.0 (buffer A), containing 1.4 mM 2-mercaptoethanol, lysed by sonication, and centrifuged for 45 min at 28,000 ϫ g at 4°C. Total GSTs were purified from the 28,000 ϫ g supernatants using GSH affinity chromatography as described in our previous studies (30).
Enzyme Assays-Catalase, glutathione peroxidase (GPx), and glutathione reductase activities were determined in 28,000 ϫ g supernatants of cell homogenates prepared in buffer A containing 1.4 mM 2-mercaptoethnol. For determination of superoxide dismutase (SOD) activity and GSH levels, 2-mercaptoethnol was excluded from homogenizing buffer. GST activity toward CDNB was determined by the method of Habig et al. (39) and that toward 4-HNE was determined according to the procedure described by Alin et al. (40). GPx activity toward cumene hydroperoxide or H 2 O 2 was measured according to the method described by us earlier (30,41). Catalase and SOD activities were determined by the methods described by Beers and Sizer (42) and Paoletti and Mocali (43), respectively. Glutathione reductase activity was determined by the method of Carlberg and Mannervik (44), and GSH was measured as non-protein thiols according to the method of Beutler et al. (45).
Synthesis of GS-HNE-Unlabeled and 4-[ 3 H]GS-HNE were synthesized enzymatically using purified human GSTs. Briefly, 1 mol of GSH hGST5.8 and RLIP76 Induction Suppresses JNK Activation was incubated with 10 mol of 4-[ 3 H]HNE in 50 mM potassium phosphate buffer, pH 7.2, and two units GST at 37°C for 30 min. GS-HNE conjugate was purified by thin layer chromatography over silica gel G plates as described in Xiao et al. (46). The authenticity of GS-HNE was confirmed through high performance liquid chromatography (HPLC) and mass spectra as described previously (15). The specific activity of [ 3 H]GS-HNE was determined by measuring radioactivity in a liquid scintillation counter (Beckman LS-6800), and the concentration of GS-HNE was determined by the previously described colorometic assay (47).
GS-HNE Transport-In these experiments, 50 ϫ 10 6 cells were incubated at 37°C for 10 min in 2 ml of PBS buffer containing 20 M 4-[ 3 H]HNE (specific activity 3,800 cpm/nmol). The cells loaded with 4-[ 3 H]HNE were harvested by centrifugation at 500 ϫ g, washed with PBS (2 ϫ 2 ml), resuspended in 2 ml of medium, and incubated for 2 h at 37°C for measuring the transport. After incubation, the cells were harvested, the media was quantitatively separated, and the radioactivity associated with the media and that retained within the cells was determined. Determinations were made in triplicate for the controls (untreated), heat shock-treated and H 2 O 2 -treated, and antibody-coated cells in parallel experiments under identical conditions using equal number of cells. For isolation and characterization of GS-HNE, the medium was lyophilized and extracted with 200 l of 70% ethanol, and the extract was used for HPLC analysis to isolate and characterize [ 3 H]GS-HNE as described above for synthetic GS-HNE.
TUNEL Assay for Apoptosis-The cells (5 ϫ 10 6 ) were allowed to recover for 2 h after a 30-min heat shock and then treated with 20 M 4-HNE for 2 h. Subsequently, the cells were washed twice with PBS and resuspended in 5 ml of PBS at a density of 1 ϫ 10 6 cells/ml. Cell aliquots (100 l containing 1 ϫ 10 5 cells) were layered onto poly-L-lysine-coated slides using Cytospin (500 ϫ g for 5 min) and fixed by treating with 4% paraformaldehyde for 20 min at 4°C, and a TUNEL assay was performed using the fluorescein apoptosis detection system (Promega) according to the protocol provided by the manufacturer. In the experiments to determine the effect of RLIP76 antibodies on heat shock protection against 4-HNE-induced apoptosis, pre-immune IgG or anti-RLIP76 IgG was added in the media (final concentration 20 g/ml) after 1 h of the recovery period. The cells were allowed to recover for an additional 1 h and incubated with 20 M 4-HNE for 2 h, and apoptosis was detected by the TUNEL assay.
DNA Laddering Assay-To detect apoptosis by DNA-laddering assay, 5 ϫ 10 6 cells (control or stress-pre-conditioned) were incubated with 20 M 4-HNE for 2 h at 37°C. Cells were pelleted, washed twice with PBS, and resuspended in 200 l of PBS, and the genomic DNA was isolated using QIAamp DNA blood mini kit (Qiagen) according to the manufacturer's instructions. The DNA-laddering assay was performed as described by us previously (30).
JNK Assay, Caspase 3 Activation, and PARP Cleavage-The JNK assay was performed essentially according to the method of Uchida et al. (1), with slight modifications as described by us previously (30). Caspase 3 activation and PARP cleavage were determined by Western blot analysis. Briefly, total cell lysates containing 25-50 g of protein were separated by SDS-polyacrylamide gel electrophoresis (12% gels) and transferred onto nitrocellulose membranes (Bio-Rad). Immunoblots were developed with the ECL according to the manufacturer's instructions using the antibodies against caspase 3 and PARP (H-250).
Cytotoxicity Assay-The sensitivity of the cells to 4-HNE was measured using the MTT assay (48) as described by us previously (30).

RESULTS
Effect of Heat Shock on LPO-K562 cells exposed to 42°C for 30 min or 50 M H 2 O 2 for 20 min followed by a 2-h recovery did not show any apparent effects on their gross morphology or vitality as measured by an MTT assay. Comparison of the extent of LPO in the control and heat shock-treated cells by the conventional thiobarbituric acid-reactive substance assay showed that LPO was increased by about 50 and 72% in heat shock and H 2 O 2 -treated cells (Fig. 1), respectively. 4-HNE levels in the untreated control cells and heat shock-treated cells were found to be 44.68 Ϯ 1.37 pmol/mg of protein and 68.26 Ϯ 3.46 pmol/mg of protein (n ϭ 3), respectively. These results indicated that increased LPO in heat shock-treated cells resulted in a proportional increase in 4-HNE levels.
Effect of Heat Shock on GSTs-Western blot analysis of the lysates of cells exposed to 42°C for 30 min and allowed to recover at 37°C for 2 h showed no noticeable effect of heat shock on the expression of the cationic Alpha-class isozymes hGSTA1-1, hGSTA2-2, hGSTA3-3, the Mu-class isozymes, and the Pi-class isozyme, GSTP1-1 (data not presented). The GST activity of heat shock-treated cells toward CDNB (Table I) was not significantly changed as compared with the control, which was consistent with the results showing no effect of heat shock on expression of the Alpha-, Mu-, or Pi-class GSTs, which account for the bulk of GST activity of K562 cells toward CDNB (31). On the other hand, heat shock caused a transient induction (about 3-fold) of the 4-HNE-metabolizing GST isozyme, hGST5.8, in cells exposed to a 30-min heat shock that declined gradually in cells exposed to longer periods of heat shock (Fig.  2, A and E). Consistent with the induction of hGST5.8 protein, an increase in the GST activity toward 4-HNE was also observed in heat shock-treated cells (Table I). In human tissues two immunologically distinct GST isozymes, hGSTA4-4 (23, 24) and hGST5.8 (20 -22), with high catalytic efficiency for 4-HNE have been reported. Western blot analysis of the control or heat shock-treated K562 cells using specific hGSTA4-4 antibodies showed no detectable expression of hGSTA4-4 in these cells (Fig. 2B). In heat shock-treated cells, the activities of glutathione peroxidase toward H 2 O 2 or cumene hydroperoxide, catalase, superoxide dismutase, and glutathione reductase remained unaltered (Table II). However, GSH levels in heat shock-treated cells (78.6 Ϯ 0.85 ng/mg of protein in control versus 151.1 Ϯ 3.0 ng/mg of protein in heat shock-treated cells, n ϭ 3) were increased significantly (p Ͻ 0.01).
Effect of Heat Shock and Oxidative Stress on RLIP76 and HSP70 -GSH conjugates (e.g. GS-HNE) of electrophilic compounds inhibit GSTs, and therefore, these conjugates must be eliminated from the cells to sustain GSH conjugation reactions. It has been recently shown that RLIP76 mediates the transport of GS-HNE in K562 cells (27) and erythrocytes (29). We therefore studied the effect of heat shock on the expression of RLIP76 and the transport of GS-HNE. Western blot analysis (Fig. 2C) showed that RLIP76 was induced by about 3.7-fold in cells subjected to a 30-min of heat shock (Fig. 2E). Similar to hGST5.8, the levels of induction of RLIP76 declined in cells exposed to longer periods of heat shock (Fig. 2, C and E). RLIP76 is known to show a band at 95 kDa in SDS gels even hGST5.8 and RLIP76 Induction Suppresses JNK Activation though its calculated molecular mass is 76 kDa (25). Previous studies show that RLIP76 readily undergoes proteolytic degradation to yield fragments of varying molecular masses (27)(28)(29). Results presented in Fig. 2C were consistent with these studies and a proportional increase in the intensities of the 95 kDa, and the lower molecular mass peptides were observed in cells exposed to 30 min of heat shock. We also compared the expression of HSP70, RLIP76, and hGST5.8 in cells exposed to heat shock for increasing time periods (0.5-5 h). Western blot analysis (Fig. 2D) showed no significant effect on the expression of HSP70 in cells exposed to 42°C for a period up to 1 h. However, a gradual increase in the expression of HSP70 was observed in cells exposed to heat shock for more than 1 h. This was in contrast to the effect of heat shock on hGST5.8 and RLIP76, where maximal induction was observed in cells subjected only to a 30-min heat shock that gradually declined with longer periods of heat shock (Fig. 2, A, C, and D). Our results showing a robust and sustained activation of HSP70 on prolonged heat shock are consistent with previous studies demonstrating the protective role of HSP70 against stress (49 -54). Results presented in Fig. 3 showed that the effect of mild transient oxidative stress (50 M H 2 O 2 , 20 min) on the levels of hGST5.8, RLIP76, and HSP70 was similar to that of heat shock, indicating that oxidative stress or heat shock affected the expression of these proteins in a similar manner.
Effect of Heat Shock on Transport of GS-HNE-To investigate the functional consequences of increased overexpression of RLIP76 in heat shock-treated cells, we compared the transport of GS-HNE in the control and heat shock-treated cells. In these experiments, the control and heat shock or H 2 O 2 -treated cells (after a 2-h recovery period) were loaded with 4-[ 3 H]HNE by incubating cells with 20 M 4-[ 3 H]HNE in PBS for 10 min at 37°C. We have previously shown that 4-HNE is readily taken up by the cells and is quickly converted to GS-HNE (27). The cells were pelleted and washed 3 times with PBS to remove extracellular 4-HNE. After loading, cells were resuspended in media and incubated at 37°C for 2 h, after which [ 3 H]GS-HNE was quantitated in the medium and the cells. The results presented in Table III showed that the amount of GS-HNE transported from cells subjected to heat shock for 30 min was about 3-fold higher compared with untreated controls. These results are consistent with previous studies showing increased efflux of GSH-conjugate from cells after heat shock (55). A gradual decline in the efflux of GS-HNE was observed in the cells subjected to increasing periods of heat shock (data not presented), which was consistent with the decreasing RLIP76 expression in cells upon prolonged heat shock (Fig. 2, C and E). In cells coated with anti-RLIP76 IgG, the transport of GS-HNE was inhibited (Table III) by about 65%, whereas the pre-immune serum did not have any noticeable effect on transport. These results demonstrated that in these cells, RLIP76 was the major transport protein for the efflux of GS-HNE, which was consistent with our previous studies showing that RLIP76 ac- -free medium, and re-stored in the medium for 2 h at 37°C. After 2 h of rest, cells were pelleted by centrifugation at 500 ϫ g for 5 min at 4°C, resuspended in buffer A, and lysed by sonication. The lysates were centrifuged for 45 min at 28,000 ϫ g at 4°C, GSTs were purified from the supernatants by affinity chromatography on a column of GSH linked to epoxy-activated Sepharose 6B, and the GST activities of purified GSTs toward CDNB and 4-HNE were determined as described under "Experimental Procedures." Values are the means Ϯ S.D. (n ϭ 31). One unit of GST catalyzed the conjugation of 1 mol of the electrophilic substrate/min at 25°C for CDNB, and of 1 mol of 4-HNE at 30°C.  2. Effect of mild, transient heat shock on the expression of hGST5.8, RLIP76, and HSP70 in K562 cells. K562 cells in normal media were separately exposed to 42°C for 0, 0.5, 1, 2, or 5 h, brought back to 37°C, and allowed to recover for 2 h. The cells were then pelleted and homogenized in lysis buffer (10 mM Tris-HCl, pH 7.4, containing 5 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride, and 1 g/ml leupeptin). The homogenates containing 25 g (for RLIP76 and HSP70) or 100 g of protein (for hGST5.8) were subjected to Western blot analysis using the primary antibodies against mGSTA4-4, which are highly specific for hGST5.8 (A); hGSTA4-4 (B); RLIP76 (C); and HSP70 (D). The blots were developed by ECL method. A and B: lanes 1, positive controls, rec-mGSTA4-4 and rec-hGSTA4-4, respectively; lanes 2, untreated control cells; lanes 3-6, cells exposed to 42°C for 0.5, 1, 2, and 5 h, respectively. In C and D, lanes 1, untreated control cells; lanes 2-5, cells exposed to 42°C for 0.5, 1, 2, and 5 h, respectively. E, bar graph showing the intensities of hGST5.8, RLIP76, and HSP70 bands, determined by densitometric analysis using ImageQuant Software Version 3.3 and normalized with the intensity of ␤-actin bands.

hGST5.8 and RLIP76 Induction Suppresses JNK Activation
counted for more than two-thirds of the transport of the electrophile-GSH conjugates from human erythrocytes (29).
Identification of GS-HNE in the Medium-Analysis of the authentic GS-HNE on a reverse phase HPLC column showed four major peaks at retention times of 22.0, 24.2, 28.7, and 31.7 min (Fig. 4A). This was consistent with previous studies showing multiple peaks for GS-HNE due to its diastereoisomers (26). The conjugation of 4-HNE to GSH results in two chiral carbon atoms, and the subsequent ring closure of GS-HNE yields another chiral carbon, giving rise to the possibility of up to a maximum of eight diastereoisomers of GS-HNE (26). To determine whether or not the radioactivity in the medium represented [ 3 H]GS-HNE, the components of the medium were analyzed. The media collected from the control and heattreated cells were separately lyophilized and extracted with 70% ethanol, and the extracts were subjected to HPLC. The results of HPLC analysis (Fig. 4B) showed the presence of a major peak at 28.5 min in the media from the control and treated cells, and it coincided with the peak of radioactivity. This peak corresponded to the hemiacetal form of GS-HNE. The structure of the hemiacetal form of GS-HNE was confirmed by mass spectrometry (Fig. 5), where it showed an M-1 peak at m/e 462 corresponding to its molecular weight and m/e peaks at 306 and 444, which corresponded to M-1 peak for GSH and M-17 (OH) for the hemiacetal. Taken together, these results showed that the cells transported GS-HNE and that the rate of its transport in heat shock-treated cells expressing higher levels of RLIP76 was higher as compared with that in untreated controls.
The Effect of Heat Shock on 4-HNE and H 2 O 2 Induced Cytotoxicity and Apoptosis-The result of MTT assays to assess the

TABLE II
Effect of heat shock on GSH levels and the enzymes involved in antioxidant functions 1 ϫ 10 6 K562 cells with or without heat shock (for 30 min) were washed with PBS and homogenized at 4°C by sonication (3 ϫ 5 s, 40 W) in 500 l of 10 mM potassium phosphate buffer, pH 7.0. Except for the GSH and SOD assays, 1.4 mM 2-mercaptoethanol was included in the lysis buffer. Homogenates were centrifuged at 28,000 ϫ g (10 min), and aliquots of the supernatants were used for analysis. Values are the means Ϯ S.D. of determinations, given in parentheses. CUOOH, cumene hydroperoxide.  cells) were exposed to 42°C for 30 min and allowed to recover for 2 h in medium at 37°C. The cells were pelleted and re-incubated for 10 min at 37°C in 2 ml of medium containing 20 M 4-[ 3 H]HNE. The cells were pelleted and washed twice with 2 ml of PBS. The supernatants and washings were discarded, and the cells were incubated at 37°C for 2 h in 2 ml of 4-HNE-free medium, after which radioactivity was determined in the medium.
b Statistically significant differences between treated and control cells evaluated by Student's t test (P Ͻ 0.05).
c For H 2 O 2 treatment, the cells were incubated for 20 min at 37°C in media containing 50 M H 2 O 2 . After incubation, the cells were pelleted, washed free of H 2 O 2 , and incubated in H 2 O 2 -free medium at 37°C for 2 h, after which the radioactivity was measured in the medium. d For treatment with antibodies, the cells, after heat shock treatment were allowed to recover for 1 h, and respective IgGs were added (20 g/ml medium) and incubated at 37°C for an additional 1 h. The cells were pelleted, and [ 3 H]GS-HNE transport was measured as described above. Values are the means Ϯ S.D. (n ϭ 3 separate experiments).

hGST5.8 and RLIP76 Induction Suppresses JNK Activation
cytotoxicity of H 2 O 2 and 4-HNE showed that the cells preexposed to heat shock were significantly more resistant to the cytotoxic effects of H 2 O 2 (about 1.7-fold) and 4-HNE (about 2-fold) compared with untreated control cells. Results presented in Fig. 6 demonstrated that stress-pre-conditioned cells acquired resistance against 4-HNE-induced apoptosis. Although a 2-h treatment with 20 M 4-HNE caused apoptosis in the control cells (Fig. 6A, lane 3), no apoptosis was observed in stress-pre-conditioned cells (Fig. 6A, lanes 5 and 7). The resistance of the stress-pre-conditioned cells to 4-HNE-induced apoptosis was confirmed by the results of experiments showing that PARP (a substrate of caspase 3) cleavage was observed only in the control cells treated with 4-HNE (Fig. 6B, lane 2) but not in stress-pre-conditioned cells (Fig. 6B, lanes 4 and 6).

4-HNE metabolism and exclusion of its GSH conjugate, acquired resistance to apoptosis caused by H 2 O 2 or by 4-HNE through blockage of caspase 3 activation. Anti-RLIP76 Antibodies Abolish the Protective Effect of Heat Shock Pre-conditioning against 4-HNE-induced Apoptosis-We
have previously demonstrated that the majority of the ATP-dependent transport activity of erythrocyte membrane toward GS-HNE and other GS conjugates is inhibited (65%) by the polyclonal antibodies against recombinant RLIP76 (29). We reasoned that if the protection to heat-pre-conditioned cells from 4-HNE-induced apoptosis was due to induction of RLIP76 and increased efflux of GS-HNE from the cells, it should be abolished by inhibiting GS-HNE efflux by RLIP76 antibodies. Therefore, heat shock-treated cells were incubated with anti-RLIP76 IgG after a 1-h of recovery period, allowed to recover for an additional 1 h, and then tested for 4-HNE-induced apoptosis using TUNEL assay. Results presented in Fig. 7 showed that heat shock alone did not cause apoptosis (Fig. 7A). Treatment with 4-HNE caused apoptosis in control cells (Fig.  7B), whereas heat shock-pre-conditioned cells were resistant to 4-HNE-induced apoptosis (Fig. 7C). In contrast, in heat shockpre-conditioned cells coated with anti-RLIP76 IgG, 4-HNEinduced apoptosis was observed (Fig. 7D), indicating that these antibodies abrogated the protective effect of heat shock preconditioning. In cells coated with the preimmune IgG, the protective effect of heat shock pre-conditioning against 4-HNEinduced apoptosis was retained (data not presented). These results taken together with inhibition of GS-HNE transport by anti-RLIP76 antibodies (Table III) strongly suggested that the protection provided to the heat-pre-conditioned cells was directly linked to the induction of RLIP76 and its associated transport function.
Effect of Heat Shock and Transient Oxidative Stress on JNK Activation-A number of cellular stress conditions including heat shock and oxidative stress activate the JNK cascade, and this signaling pathway has been implicated in mediating the apoptotic process. Recent studies suggest that a sustained activation of JNK leads to apoptosis in different cell lines (30, 56 -59). Stress-pre-conditioned cells in our studies acquired significant resistance to 4-HNE-and H 2 O 2 -induced apoptosis compared with control cells. Therefore, we compared 4-HNEmediated activation of JNK in control and stress-pre-condi-FIG. 7. Effect of anti-RLIP76 IgG on 4-HNE-mediated apoptosis in heat shock-pre-conditioned cells. Aliquots (50 -100 l) containing 1ϳ2 ϫ 10 6 cells were fixed onto poly-L-lysine-coated slides by Cytospin at 500 ϫ g for 5 min, and the TUNEL apoptosis assay was performed as described under "Experimental Procedures." The slides were analyzed by fluorescence microscope (Nikon Eclipse 600, Japan) using a standard fluorescein filter (EX 450 -490, DM 505, BA 520, B-2A). Photomicrographs at 400ϫ magnification are presented. Apoptotic cells showed characteristic green fluorescence. A, control K562 cells pre-treated with heat shock (42°C, 30 min) and allowed to recover for 2 h at 37°C. B, control cells without heat shock pre-treatment, incubated with 20 M 4-HNE for 2 h. C, cells pre-treated with heat shock, allowed to recover for 2 h at 37°C, followed by incubation in medium containing 20 M 4-HNE for 2 h at 37°C. D, heat shock-pretreated cells, allowed to recover for 1 h at 37°C, after which anti-RLIP76 IgG was added to medium (20 g/ml final concentration) and incubated for an additional 1 h. Cells were then incubated for 2 h at 37°C in medium containing 20 M 4-HNE. hGST5.8 and RLIP76 Induction Suppresses JNK Activation tioned cells. For these experiments, the cells were stressed (heat or H 2 O 2 ) as described above and allowed to recover in normal medium, and JNK activity was monitored during the recovery period. As shown in Fig. 8A, both heat and H 2 O 2 treatments alone caused an immediate transient activation of JNK that was prominent at 0 h during the recovery period ( Fig.  8A) but returned to basal levels after 2 h of recovery. In other experiments, the control and stressed-pre-conditioned cells were separately incubated in medium containing 20 M 4-HNE or 100 M H 2 O 2 for 2 h, allowed to recover in normal medium, and monitored for JNK activation during the recovery period. The results of these experiments presented in Fig. 8, B and C, showed that in control cells both 4-HNE and H 2 O 2 caused a sustained activation of JNK that persisted for at least 8 h. In heat-and H 2 O 2 -pre-conditioned cells, even though 4-HNE and H 2 O 2 initially caused JNK activation, the activity gradually declined and returned to basal levels within 8 h of the recovery period. Together, these results showed that in stress-pre-conditioned cells, 4-HNE-and H 2 O 2 -mediated activation of JNK was suppressed, perhaps due to the enhanced metabolism and exclusion of 4-HNE.
Effect of Heat Shock on hGST5. 8

and RLIP76 Expression and 4-HNE-induced Apoptosis in Different Cell
Lines-Early induction of hGST5.8 and RLIP76 in response to heat shock was not limited to K562 cells, as the results presented in Fig.  9 clearly showed an early induction of these proteins in response to mild transient heat shock in various other cell lines. Our results showed early induction of hGST5.8 and RLIP76 by heat shock in human leukemia cell line HL-60 (Fig. 9, A and B, lanes 2), human retinal pigmented epithelial cells (Fig. 9, A and B, lanes 4), human non-small cell lung cancer cell line H-226 (Fig. 9, A and B, lanes 6), and human small cell lung cancer cell line H-69 (Fig. 9, A and B, lanes 8).
We also compared the apoptotic effect of 4-HNE in RPE, HL-60, H-226, and H-69 cells with or without heat shock preconditioning. The results of these experiments showed that incubation of control HL-60 and H-69 cells with 10 M 4-HNE for 2 h caused apoptosis only in the control cells (Fig.  10, lanes 2) and not in stress-pre-conditioned cells (Fig. 10,  lanes 3). Likewise, 20 M 4-HNE caused apoptosis in control RPE and H-226 cells (Fig. 10, lanes 2) but not in stress-preconditioned cells. Collectively these results suggested that an early induction of hGST5.8 and RLIP76 in response to heat shock appears to be a generalized phenomenon and that the cells pre-conditioned with a mild transient heat shock acquire resistance to apoptosis caused by the LPO product, 4-HNE. DISCUSSION The activation of JNK/stress-activated protein kinase cascade upon exposure to heat, UV radiation, or inflammatory cytokines and its involvement in apoptotic signaling has been extensively studied. Available evidence suggests that activation of these kinases in response to stress is mediated by GTPase-activating proteins, in particular Rac and Cdc42 (60 -62). It has been suggested that activated Cdc42 binds to and activates PAK65, which activates stress-activated protein kinase kinase/JNK kinase (SEK/JNKK) by phosphorylating JNKK at the serine residue 219 (63). In turn, activated stress-activated protein kinase kinase/JNK kinase (SEK/ JNKK) causes activation of JNK/stress-activated protein kinase, which binds and phosphorylates c-Jun at specific residues. The initial events leading to the activation of JNK/ stress-activated protein kinase and the nature of chemical species initiating this cascade in response to stress are, however, poorly defined.
The results of the present studies shed light on the mechanisms of the response of the cells to stress in the initial phase and demonstrate a role of lipid peroxidation products, particularly 4-HNE, in stress-mediated signaling. We demonstrate that in cells subjected to a mild, transient stress (heat or oxidative), lipid peroxidation ensues, resulting in an increase in the cellular levels of 4-HNE. The cells respond to this increase in 4-HNE levels by acquiring the capability to exclude 4-HNE from their environment at a faster rate. This response occurs earlier than the induction of heat shock proteins or the antioxidant enzymes, which are known to be induced by stress. Within 30 min of exposure to mild stress, cells showed about a 50% increase in the 4-HNE levels that was accompanied by a concomitant activation of JNK, whose sustained activation has been suggested to be a prerequisite for activation of caspase 3 and subsequent apoptosis (64,65). However, within a 2-h resting period after the transient stress, the cells were able to exclude 4-HNE by transporting its GSH conjugate (GS-HNE) at a severalfold higher rate as compared with the control cells. The cells acquire this capability through a rapid induction of hGST5.8 and RLIP76 Induction Suppresses JNK Activation hGST 5.8 and RLIP76, which, respectively, catalyze the conjugation of 4-HNE to GSH and subsequent efflux of GS-HNE. This is indicated by our results showing increased expression of hGST5.8 and RLIP76 proteins, increased GST activity toward 4-HNE, and an accelerated transport of GS-HNE in the cells exposed to a mild, transient stress. As a result of the accelerated exclusion of 4-HNE-GSH conjugate, within a 2-h resting period after stress, 4-HNE levels were restored to normal physiological or basal levels, and the JNK activity was restored to constitutive levels. These results strongly suggest a role of 4-HNE in the activation of JNK either through directly acting on JNK or through interaction(s) with the kinases upstream to JNK. This postulate needs to be substantiated by further studies.
Our results showing acquisition of a limited resistance to H 2 O 2 -and 4-HNE-mediated apoptosis in stress-pre-conditioned cells overexpressing hGST 5.8 and RLIP76 further suggest the role of 4-HNE in stress-mediated signaling mechanisms. Our studies show that when 4-HNE is included in the medium for 2 h, the control cells undergo apoptosis that is accompanied by a sustained activation of JNK and caspase 3, whereas apoptosis or a sustained activation of JNK and caspase 3 is not observed in stress-pre-conditioned cells. Stress pre-conditioning either by heat shock or H 2 O 2 by itself caused transient activation of JNK, which returned to normal levels within 2 h when constitutive levels of 4-HNE were restored in these cells, probably due to the induction of hGST5.8 and RLIP76 and accelerated 4-HNE metabolism and transport of GS-HNE. Stress-pre-conditioned cells also acquire resistance to apoptosis by H 2 O 2 . It is to be noted that the stressed-preconditioned cells should have similar capabilities to decompose H 2 O 2 because the levels of the antioxidant enzymes, catalase, GPx, and SOD in the controls and conditioned cells are similar. Therefore, the resistance of pre-conditioned cells to H 2 O 2 -induced apoptosis may also be attributed to their ability to metabolize and transport 4-HNE at a faster rate. This contention is supported by our results showing that the resistance of stress-pre-conditioned cells to 4-HNE-induced apoptosis can be abrogated by coating the cells with anti-RLIP76 IgG, which inhibit the transport of GS-HNE. Collectively, these results demonstrate that the intracellular concentration of 4-HNE or its conjugate GS-HNE may be one of the major determinants for the signaling mechanisms leading to the activation of JNK, caspase 3, and subsequent apoptosis.
Activation of JNK is involved in stress-mediated apoptosis. Our results show that only a 50% increase in 4-HNE levels over its constitutive physiologic levels appears to initiate JNK activation. It has been reported that relatively low concentrations of 4-HNE promote proliferation of aortic smooth muscle cells (14) and that K562 cells transfected with mGSTA4-4, which have lower levels of 4-HNE, grow at a faster rate than the vector-transfected controls (6). These findings imply that at least in some cell types, low concentrations of 4-HNE may promote proliferation, whereas relatively higher concentrations of 4-HNE may promote apoptosis. Therefore, the mechanisms that determine the intracellular concentrations of 4-HNE should be crucial in the early phases of signaling. Our observations showing an early induction of hGST 5.8 and RLIP76 in response to stress support this idea. Furthermore, our results demonstrate that tumor cells of diverse origins respond to stress in a similar manner, and an early induction of hGST 5.8 and RLIP76 is observed in all the cell lines tested in the present work. This suggests that in the initial phase, the cells in general respond to stress (which causes lipid peroxidation) by up-regulating the mechanisms for metabolism and exclusion of 4-HNE to avoid apoptosis. The results of our studies, showing that only a small increase in the concentrations of 4-HNE leads to the activation of JNK, which is quickly suppressed to basal levels when 4-HNE levels are restored to normal, suggest that there is a narrow range of constitutive levels of 4-HNE, above which the apoptotic cascade is initiated. It is possible that suppression of the 4-HNE levels below the FIG. 9. Effect of heat shock on the expression of hGST5.8 and RLIP76 in different human cell lines. Human leukemia HL-60, human small cell lung cancer cell line H-69, human non-small cell lung cancer cell line H-226, and human RPE cells were treated with heat shock at 42°C for 30 min. The cells were allowed to recover at 37°C for 2 h and lysed, and the homogenates containing 25 g (for RLIP76) or 100 g of protein (for hGST5.8) were subjected to SDS-polyacrylamide gel electrophoresis in 12% gels and Western blot analysis using antibodies against mGSTA4-4 (A) and anti-RLIP76 (B) antibodies. The blots were developed with the ECL method. In both panels: lanes 1 and 2, control and heat shock-treated HL-60 cells, respectively; lanes 3 and 4, control and heat shock-treated RPE cells, respectively; lanes 5 and 6, control and heat shock-treated H-226 cells, respectively; lanes 7 and 8, control and heat shock-treated H-69 cells, respectively. hGST5.8 and RLIP76 Induction Suppresses JNK Activation constitutive levels may promote proliferation, such as that observed in K562 cells transfected with mGSTA4-4 having sub-basal levels of 4-HNE (6). Further studies are needed to substantiate these possibilities.
GSTs account for the metabolism of a major fraction of cellular 4-HNE (15). Recent studies show that in human tissues, two immunologically distinct GST isozymes with high catalytic efficiency (k cat /K m of about 2500 -3000 s Ϫ1 ⅐ mM Ϫ1 ) for 4-HNE are expressed in relatively low abundance in a tissue-specific manner (22). One of these isozymes, hGSTA4-4, has been cloned, but the gene encoding hGST5. 8 has not yet been cloned. In tumor cells used in the present studies, detectable expression of hGSTA4-4 was not observed either in control cells or in the stressed cells. Our results showed that low levels of hGST5.8 were constitutively present in all these cells, and a mild stress caused an early and transient induction of this isozyme. These results suggest that hGST5.8 is induced to interrupt 4-HNE-mediated signaling for apoptosis, and it is rapidly degraded after performing this function, as indicated by our results showing that within 5 h after stress, hGST5.8 expression returns to the usual low constitutive levels. This may explain the difficulty in cloning of hGST 5.8 from human libraries, because even though partial primary structure of hGST 5.8 purified from various human tissues and cell lines has been determined and its high catalytic efficiency toward 4-HNE has been established (20 -22), attempts to clone its cDNA have not been successful, perhaps due to its low abundance in cDNA libraries. Recent studies (66) identifying a pseudogene corresponding to the peptide sequences of hGST5.8 may be helpful in its cloning.
RLIP76 is a GTPase-activating protein that is believed to bridge the Ral and Rho pathways (25). It is a Ral effector protein with GTPase-activating protein activity for Cdc42/ Rac proteins, but its role in the signaling mechanism is not clearly understood. It has recently been demonstrated that RLIP76 catalyzes the ATP-dependent transport of GSH conjugates including GS-HNE and leukotrienes (27,29). In the present studies, we have focused only on the transport function of RLIP76 and demonstrated an increase in the transport of GS-HNE from the cells with induced RLIP76, which determines the intracellular levels of 4-HNE. Involvement of GTPase-activating protein, particularly Rac and Cdc42, in apoptotic signaling is known (67)(68)(69). Thus, the interactions of RLIP76 with CDC42/Rac may also potentially affect the stress-mediated signaling. Further studies are required to define the significance of RLIP76 induction as an early response to stress.