Heat Shock Factor 1 Attenuates 4-Hydroxynonenal-mediated Apoptosis

Lipid peroxidation is a consequence of both normal physiology and oxidative stress that generates various reactive metabolites, a principal end product being 4-hydroxynonenal (HNE). As a diffusible electrophile, HNE reacts extensively with cellular nucleophiles. Consequently, HNE alters cellular signaling and activates the intrinsic apoptotic cascade. We have previously demonstrated that in addition to promoting apoptosis, HNE activates stress response pathways, including the antioxidant, endoplasmic reticulum stress, DNA damage, and heat shock responses. Here we demonstrate that activation of the heat shock response by HNE is dependent on the expression and nuclear translocation of heat shock factor 1 (HSF1), which promotes the expression of heat shock protein 40 (Hsp40) and Hsp70-1. Ectopic expression and immunoprecipitation of c-Myc-tagged Hsp70-1 indicates that HNE disrupts the inhibitory interaction between Hsp70-1 and HSF1, leading to the activation heat shock gene expression. Using siRNA to silence HSF1 expression, we observe that HSF1 is necessary for the induction of Hsp40 and Hsp70-1 by HNE, and the lack of Hsp expression is correlated with an increase in apoptosis. Nrf2, the transcription factor that mediates the antioxidant response, was also silenced using siRNA. Silencing Nrf2 also enhanced the cytotoxicity of HNE, but not as effectively as HSF1. Silencing HSF1 expression facilitates the activation of JNK pro-apoptotic signaling and selectively decreases expression of the anti-apoptotic Bcl-2 family member Bcl-XL. Overexpression of Bcl-XL attenuates HNE-mediated apoptosis in HSF1-silenced cells. Overall, activation of HSF1 and stabilization of Bcl-XL mediate a protective response that may contribute significantly to the cellular biology of lipid peroxidation.

The oxidation of membrane phospholipids is an inescapable consequence of normal cellular respiration. A major route for lipid oxidation involves the abstraction of bis-allylic hydrogen atoms from polyunsaturated fatty acids by reactive oxygen spe-cies. The resulting lipid radicals react readily with dioxygen to form lipid hydroperoxides, which are mainly reduced to their corresponding hydroxyl species. A number of xenobiotic metabolites, anti-neoplastic drugs, transition metals, and disease states promote increased levels of both reactive oxygen species and oxidized lipids. The decomposition of oxidized lipids yields a variety of diffusible electrophiles, including the ␣,␤unsaturated aldehyde, 4-hydroxynonenal (HNE). 2 Diffusible lipid electrophiles are capable of modifying nucleophilic sites on DNA and proteins throughout the cell. Modification of cellular targets by HNE is found in multiple disease states, including atherosclerosis, ischemia-reperfusion injury, Parkinson disease, and Alzheimer disease (1)(2)(3)(4)(5). Plasma conjugates of HNE have been observed at low micromolar concentrations in healthy adults, implying the detectable generation of HNE under normal physiological conditions, although absolute steady-state concentrations are less clear (6,7). At the cellular level, elevated levels of HNE promote cell cycle arrest and programmed cell death. Studies in a variety of cultured cell systems, including: neuronal (PC12); endothelial human umbilical vein endothelial cells; myeloid (K562, HL-60); lymphoid (MOLT-4, Reh); and colorectal cell lines (RKO) have verified the induction of apoptosis by HNE (8 -13). Reports have indicated that apoptosis is mediated in part by the activation of c-Jun N-terminal kinase (JNK) and that HNE toxicity is reduced by the JNK-specific inhibitor SP600125 (9, 14 -16).
In addition to initiating apoptosis, HNE activates various protective, stress response pathways. We previously used microarray analysis to explore transcriptional pathways activated by HNE in RKO cells and found that HNE induces the DNA damage, endoplasmic reticulum stress, antioxidant, and heat shock responses (17). Among the pathways activated by HNE, the heat shock response is particularly robust. The heat shock response mediates the induction of a highly conserved set of heat shock proteins (Hsps) (18). The inducible expression of Hsps is mediated by heat shock transcription factor 1 (HSF1), which translocates to the nucleus upon activation and enhances the expression of genes from promoters containing heat shock elements (HSE) (19,20). A principal function of Hsps is to chaperone other proteins, binding to nascent polypeptide chains as well as to unfolded and damaged proteins. Their function as protein chaperones aids in the recovery of cells from thermal and chemical-induced damage (21,22). Moreover, Hsps regulate a diverse set of signaling pathways via their interactions with "client" proteins (23). In addition, although Hsp70 does not interact directly with Bax, it does inhibit Bax translocation to the mitochondrion and attenuates stress-induced apoptosis (24).
Here we investigate activation of the heat shock response by HNE in detail. Treatment of a colorectal cancer cell line with HNE elicits a concentration-dependent activation of HSF1, associated with a robust increase in Hsp40 and Hsp70-1 (inducible Hsp70, Hsp72) expression. Utilizing an siRNA-based method to silence HSF1, we attenuate the heat shock response and examine the effect on cell survival. Our data indicate that heat shock response protects against HNE-induced apoptosis and that HSF1 activation is critical in mediating this effect. The degree of cytoprotection afforded by the heat shock response exceeds the anti-apoptotic effects of Nrf2 activation. We find that treatment with HNE causes both activation of JNK signaling and a specific reduction in Bcl-X L expression, both of which are exacerbated by silencing HSF1. Ectopic expression of Bcl-X L confers resistance to HNE-mediated apoptosis and partially reverses the sensitization to HNE caused by HSF1 siRNA. These results indicate a critical role for HSF1 activation in resistance to stress-mediated apoptosis and suggest a novel role for Hsp70-1 as a regulator of mitochondria-mediated cell death.

EXPERIMENTAL PROCEDURES
Cell Culture and Treatment-RKO cells, a poorly differentiated, p53(ϩ) colon carcinoma cell line widely used for studies on DNA damage and cellular signaling of electrophiles, was obtained from American Type Culture Collection and maintained in DMEM (Invitrogen) supplemented with 10% fetal bovine serum (Atlas), 1% antibiotic/antimycotic (Invitrogen), and 25 mM HEPES buffer. HNE (Axxora) was quantified by UV-visible spectroscopy using a molar extinction coefficient of 13,750 M Ϫ1 cm Ϫ1 at 223 nm. For addition to cell culture, a desired quantity of HNE in ethanol was dried down under an argon stream and reconstituted in Me 2 SO. The cells were treated in 10% fetal bovine serum-containing DMEM containing a final concentration of 0.5% Me 2 SO.
Viability Assays-RKO cells were seeded in 96-well plates at a density of 7.5ϫ 10 3 cells/well. After adhering overnight, the cells were treated with HNE or Me 2 SO. After 48 h of incubation, the medium was aspirated, washed once in Dulbecco's phosphate-buffered saline (D-PBS) and replaced with a solution of 2 M calcein-AM (Molecular Probes) in D-PBS. After 30 min, fluorescence was read using a SpectraMax multiwell plate reader (Molecular Devices) with an excitation of 494 nm and an emission of 517 nm.
siRNA Transfections-For siRNA transfections, RKO cells were seeded at 25% confluence in 10-cm dishes. After adhering overnight, the cells were washed once with D-PBS, and 4 ml of Opti-MEM (Invitrogen) was added per dish. The transfections were performed with 0.2 nmol of Stealth siRNA (HSF1 sense strand sequence, 5Ј-CGGAUUCAGGGAAGCAGCUG-GUGCA-3Ј; Nrf2 sense strand sequence, 5Ј-CCAGUUGA-CAGUGAACUCAUUAAAU-3Ј) and Lipofectamine 2000 (Invitrogen) per dish according to the manufacturer's instructions. After 24 h the cells were split at a ratio of 1:4 and transfected a second time the following day. The cells were split, allowed 24 h to adhere, and then treated with HNE or Me 2 SO.
Plasmid Constructs and Transfections-Clones for Bcl-X L (NCBI accession BC019307) and Hsp70-1 (NCBI accession BC009322) were obtained from the Vanderbilt Microarray Shared Resource clone collection. Bcl-X L was excised from pCMV-SPORT6 using the restriction endonuclease BglII and ligated into the BamHI site of pcDNA3.1 Hygro(ϩ). Hsp70-1 was excised from pOTB7 using MscI and XhoI. The excised fragment and a double-stranded oligonucleotide encoding the c-Myc tag (5ЈP-GGCATGGAGCAAAAGCTCATTTCTGAA-GAGGACTTGG-OH3Ј) were ligated into the EcoRV and XhoI sites of pcDNA3.1-Hygro(ϩ). All of the constructs were confirmed by sequencing at the Vanderbilt DNA Sequencing Facility. For stable transfection of Bcl-X L , RKO cells at 70% confluence in 10-cm dishes were transfected using either the Bcl-X L construct or empty pcDNA3.1 Hygro(ϩ) and Lipofectamine 2000. After 24 h the cells were split into DMEM containing 200 g/ml hygromycin, and stably expressing clones were selected. For transient transfection of c-Myc-tagged Hsp70-1, RKO cells were transfected with either the c-Myc-Hsp70-1 construct or empty pcDNA3.1-Hygro(ϩ) and Lipofectamine 2000, and the experiments were performed 24 h after transfection.
Protein Extraction and Western Blotting-For total protein extraction the cells were scraped in cold D-PBS and collected by centrifugation. Cold mammalian protein extraction reagent (M-PER) lysis buffer (Pierce) containing protease and phosphatase inhibitors (Sigma) was added to cell pellets. Protein lysates were cleared by centrifugation and stored at Ϫ80°C. For nuclear proteins, the cells were collected by centrifugation and then resuspended in 800 l of ice-cold buffer A (10 mM HEPES, pH 7.9, 10 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride), and 50 l of Igepal CA-630 (Nonidet P-40) was added. The nuclei were pelleted by centrifugation at 1000 ϫ g for 5 min at 4°C and washed twice with buffer A. The nuclear proteins were extracted by the addition of buffer B (10 mM HEPES, pH 7.9, 420 mM NaCl, 1.5 mM MgCl 2 , 0.1 mM EDTA, 0.1 mM EGTA, 25% glycerol, 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride) and centrifuged at 14,000 ϫ g for 10 min at 4°C to remove debris. Protein concentrations were determined using Bradford protein assay (Bio-Rad). For Western blot, equal amounts of protein were resolved by SDS-PAGE, transferred onto a 0.2-m nitrocellulose membrane, and incubated in blocking buffer (20 mM Tris, pH 7.6, 140 mM NaCl, 0.05% Tween 20, 5% nonfat dry milk) prior to the addition of primary antibody. Following incubation with primary and secondary antibodies, Luminol-based detection was performed. Primary antibodies were obtained from the following sources: HSF1, c-Myc tag, PARP, caspase 3, caspase 3 cleavage products, JNK, phospho-JNK, Bad, Bax, Bcl-2, and Bcl-X L from Cell Signaling; Hsp27, Hsp40, Hsp70, Hsp90, and Hsp105 from BD Biosciences; ␣-tubulin from Sigma-Aldrich; and Nrf2, actin, and all secondary antibodies from Santa Cruz Biotechnology.
Immunoprecipitation-RKO cells were transfected with empty pcDNA3.1-Hygro(ϩ) vector, or with the c-Myc-Hsp70-1 expression construct. After 24 h total proteins were collected in mammalian protein extraction reagent (M-PER) lysis buffer containing protease inhibitors. The lysates were cleared by centrifugation, and protein concentrations were determined by Bradford protein assay. Protein samples were treated with either 0.1% Me 2 SO (vehicle) or with 15 M or 45 M HNE for 1 h, and then 400 g of total protein was immunoprecipitated using the ProFound c-Myc tag immunoprecipitation kit (Pierce). Following elution in 25 l of SDS loading buffer, the samples were resolved by SDS-PAGE and analyzed by Western blot.
Luciferase Assay-An HSE-bearing luciferase construct was generated by insertion of a double-stranded oligonucleotide (5ЈP-TGCTAGCCTTCTGGAACCTTCTGGAACCTTCTG-GAACCTTCTGGAACCTTCTGGAACCTTCTGGAACCT-TCTTGGAAGGGCTCGAG-OH3Ј) into the SmaI site of pGL3-Promoter (Promega). The cells were seeded at a density of 25% confluence in 12-well plates and co-transfected with 200 ng of firefly luciferase reporter construct (pGL3-HSE) and 5 ng of Renilla luciferase (pRL-SV40)(Promega) using Lipofectamine 2000 and grown for an additional 24 h. The cells were treated for 8 h with Me 2 SO or HNE at indicated concentrations and then collected in passive lysis buffer. Analysis of luciferase was performed using a dual luciferase assay system (Promega). Luciferase activity was measured using a Monolight 3010 luminometer (Pharmingen).
RNA Extraction and Real Time PCR-Total RNA was collected using RNeasy RNA collection kit (Qiagen). Digestion of trace DNA was performed by incubation with DNase using DNA-free reagent (Ambion). 1 g of total RNA was used in each reverse transcription reaction with iScript reagent, (Bio-Rad). One-tenth of each reaction volume (2 l) was used per well in subsequent real time PCR analysis, using iQ SYBR Green Supermix (Bio-Rad). Real time reactions were performed using a Bio-Rad iCycler real time PCR machine. Standard curves were generated by the amplification of target sequences that were previously cloned into pGEM-T (Promega), in dilution series from 10 Ϫ1 to 10 Ϫ6 fmol/well.

HNE Induces Nuclear HSF1 Expression and Promotes Heat
Shock Protein Expression-Our previous observation that HNE activates the heat shock response led us to examine the underlying mechanism in detail. RKO cells were treated with either Me 2 SO (vehicle) or HNE for various times. Nuclear proteins were collected and analyzed for HSF1 expression by Western blot. Me 2 SO-treated cells showed little nuclear HSF1 expression. In contrast, the addition of 5, 15, or 45 M HNE induced a robust accumulation of HSF1 within the nucleus (Fig. 1). Expression of HSF1 in nuclear extracts was noticeable within 15 min following HNE addition. To examine the magnitude of heat shock protein induction, RKO cells were treated with 30 -60 M HNE and total protein extracts collected at 8 h. The lysates were analyzed by Western blot for the expression of various heat shock proteins. HNE induced a robust, concentration-dependent increase in the expression of both Hsp40 and Hsp70-1 (Fig. 2). Only modest expression of Hsp40 and Hsp70-1 was observed in Me 2 SO-treated cells. Also, HNE did not significantly influence the expression of other heat shock proteins, including Hsp27, Hsc70 (Hsp73), Hsp90, and Hsp105.
Interaction of Hsp70-1 with HSF1-Previous work has established that Hsp70 is a negative regulator of HSF1 transcriptional activity (25,26). Furthermore, others have shown that HNE modifies Hsp70-1 in vitro, hindering its function as a protein chaperone (27). Therefore, we hypothesized that HNE disrupts the inhibitory interaction between Hsp70-1 and HSF1. RKO cells were transiently transfected with either empty pcDNA3.1 Hygro(ϩ) or an N-terminally c-Myc-tagged Hsp70-1 expression construct. Western blotting revealed that  immunoprecipitation of c-Myc-Hsp70-1 resulted in co-immunoprecipitation of HSF1. Preincubation of protein lysates with HNE (15, 45 M) inhibited co-immunoprecipitation of HSF1 in a concentration-dependent manner (Fig. 3), suggesting that HNE-modified Hsp70-1 releases HSF1. Thus, data support our hypothesis that Hsp70-1 modification facilitates activation of the heat shock response in HNE-treated cells.
HSF1 Silencing Attenuates HSE-driven Gene Expression and Heat Shock Protein Induction-To investigate the biological consequences of HSF1 activation and the resulting induction of the heat shock response, an siRNA-based approach was used to selectively inhibit HSF1 expression. RKO cells were transfected with either an HSF1-specific siRNA or a negative control siRNA (scrambled sequence). Real time PCR was performed to determine the extent of HSF1 mRNA reduction. The level of HSF1 mRNA in cells transfected with HSF1 siRNA was reduced to ϳ20% of the level in control cells (Fig. 4A). Western blotting indicates complete silencing of HSF1 protein expression (Fig.  4B). A luciferase-based reporter construct also was employed to verify the inhibition of heat shock-mediated gene expression. Cells treated with siRNA (HSF1 or control) were subsequently transfected with the HSE-bearing firefly luciferase construct (pGL3-HSE). As a control, cells were co-transfected with a constitutive Renilla luciferase expression vector (pSV40-RL). Following treatment with 30 -60 M HNE, the samples were collected in passive lysis buffer. Analysis of normalized luciferase (firefly/Renillaactivity)showsthatHNEelicitedaconcentrationdependent increase in expression from the pGL3-HSE construct (Fig. 5A), and a maximal 21-fold increase in normalized luciferase was observed with 60 M HNE. In contrast, silencing HSF1 nearly abolished HNE-induced luciferase expression from pGL3-HSE. Inhibition of the heat shock response was also evaluated by Western blot. Induction of both Hsp40 and Hsp70-1 by HNE was robust in control siRNA-transfected cells (Fig. 5B); however, their expression was completely attenuated in cells transfected with HSF1 siRNA, further verifying successful knockdown of the heat shock response.
Nrf2 Silencing Suppresses Activation of the Antioxidant Response-The induction of heat shock proteins is a cytoprotective mechanism that enhances cell survival in the wake of   thermal or chemical stress. In addition to heat shock, HNE also activates the antioxidant response, an adaptive mechanism to defend against oxidative and electrophile-induced stress that is regulated by the transcription factor, Nrf2 (17,28,29). Active Nrf2 accumulates in the nucleus and drives gene expression from antioxidant response elements, up-regulating proteins that detoxify electrophiles and restore redox balance (30,31). Experiments in vitro and in vivo confirm that the Nrf2-mediated response protects cells from cytotoxicity caused by oxidative stress (32)(33)(34). Using siRNA, we silenced Nrf2 expression in RKO cells. To evaluate the degree of Nrf2 reduction, the cells were treated with 30 or 45 M HNE and assayed for nuclear expression of Nrf2 by Western blot. In control siRNA-transfected cells, HNE induced a robust nuclear accumulation of Nrf2. In contrast, Nrf2 siRNA-transfected cells had no detectible nuclear Nrf2 following HNE treatment, indicating successful attenuation of Nrf2 protein expression by siRNA (Fig. 6A). To confirm silencing of the antioxidant response, the level HO-1 protein expression was measured. Western blot shows that siRNA-mediated silencing of Nrf2 suppressed the HNE-induced expression of HO-1 (Fig. 6B).
Silencing of HSF1 or Nrf2 Enhances the Cytotoxicity of HNE-To evaluate the relative contributions of the heat shock and antioxidant responses in counteracting the cytotoxicity of HNE, siRNA was used to individually silence either HSF1 or Nrf2. RKO cells were subsequently treated with various concentrations of HNE in 96-well plates, and cell viability was assessed using calcein-AM. In both nontransfected and control siRNA-transfected cells, the lethal concentration-50 (LC 50 , the concentration at which viability is reduced by 50%) of HNE was 55 M. In contrast, silencing HSF1 or Nrf2 approximately doubled the toxicity of HNE, with LC 50 values of 24 and 33 M, respectively (Fig. 7).

Silencing HSF1 Facilitates HNE-induced Apoptosis-Expo-
sure of RKO cells to elevated concentrations of HNE induces apoptotic cell death, as evidenced by cytochrome c release, internucleosomal chromatin cleavage (DNA laddering), and cleavage of PARP and caspase 3 (13). Suppression of HSF1 levels by siRNA significantly increases apoptotic cell death. Following treatment with 45 M HNE, expression of the 85-kDa PARP fragment was elevated in HSF1 siRNA-treated cells compared with control (Fig. 8A). Levels of the catalytically active 17and 12-kDa fragments of cleaved caspase 3 were also elevated. In addition, HNE-induced expression of the pro-apoptotic proteins CHOP (GADD153) and ATF3, which are associated with the activation of endoplasmic reticulum stress-mediated apoptosis, were enhanced in HSF1-deficient cells (Fig. 8B). These data establish the crucial role of HSF1 in mediating multiple aspects of the anti-apoptotic response in cells exposed to HNE.
Silencing HSF1 Enhances JNK Phosphorylation and Decreases Bcl-X L Expression-To explore the mechanisms underlying HNE-induced cell death, we have evaluated the signaling processes controlling the initiation of apoptosis. Previous reports have correlated HNE-induced apoptosis with activation of the pro-apoptotic, stress-activated mitogen-activated protein kinase, JNK. In support of earlier reports, our data confirm that HNE (45 M) promotes sustained JNK phosphorylation on residues Thr-183/Tyr-185, corresponding to the catalytically active form of the protein. Moreover, this process appears to be inhibited by the heat shock response, because silencing of HSF1 drastically enhanced the level of phospho-JNK in HNE-treated cells and lowered the threshold for JNK phosphorylation from 45 to 15 M (Fig. 9).
The pro-apoptotic consequences of JNK activation are likely mediated through a variety of downstream pathways, one being the modulation of proteins that regulate the mitochondrial  (intrinsic) apoptotic cascade. Therefore, we have examined the expression of proteins regulating mitochondrial permeability, including Bad, Bax, Bcl-2, and Bcl-X L . Whereas HNE did not affect Bad, Bax, or Bcl-2 levels, Bcl-X L expression decreased significantly at pro-apoptotic concentrations of HNE (Fig. 10). Silencing HSF1 exacerbated the decrease in Bcl-X L caused by HNE treatment, suggesting that the decrease in Bcl-X L partly mediates the enhanced sensitivity of HSF1-deficient cells to HNE-mediated cell death. Interestingly, knockdown of HSF1 also appeared to decrease Bcl-X L expression in the Me 2 SOtreated control, implying a role for basal Hsp expression in the regulation of Bcl-X L levels.
To test the hypothesis that decreased Bcl-X L levels facilitate HNE-mediated apoptosis, Bcl-X L was stably overexpressed in RKO cells utilizing a hygromycin-selectable expression con-struct, and the cells were subsequently treated with 15-45 M HNE. Western blots show that overexpression of Bcl-X L attenuated apoptosis in HSF1-deficient cells, as evidenced by a reduction in PARP and caspase 3 cleavage products (Fig. 11), supporting a role for Bcl-X L down-regulation in HNEmediated cell death. These data also suggest a role for Hsp70-1 and Hsp40 as negative regulators of the intrinsic apoptotic pathway via stabilization of Bcl-X L levels.

DISCUSSION
HNE is an electrophilic lipid oxidation product and is one of the major constituents responsible for the toxic effects of lipid peroxidation (35). In addition to being cytotoxic, HNE promotes the activation of various signaling networks related to cell stress, including the DNA damage, endoplasmic reticulum stress, antioxidant, and heat shock response pathways (17). Previously, HNE has been shown to promote heat shock gene expression in the human hepatoma cell line, HepG2 (36). HNE has likewise been shown to induce the nuclear expression of an HSE-binding protein in HeLa cells (presumably, HSF1) (37). Here, we addressed the role of HSF1 in the cellular response to HNE and examined the mechanistic relationship between induction of the heat shock response and HNE-mediated apoptosis.
Our data reveal a role for the heat shock response in counteracting the pro-apoptotic effects to HNE (Fig. 12). Using siRNA against HSF1 significantly enhanced the induction of apoptosis. Hence, activation of the heat shock response significantly attenuated the degree of cell death at certain levels of HNE exposure. Based on our data, HSF1-deficient cells at 50 M HNE were Ͻ5% viable, compared with control cells, which were Ͼ50% viable. By using siRNA against Nrf2, we also silenced the antioxidant response, which likewise sensitized cells to HNE. Our data, however, indicate that the degree of cytoprotection afforded by the heat shock response exceeded the effect of the antioxidant response.
A principal step in the activation of HSF1 is translocation to the nucleus following heat shock or chemical stress (20,38). In unstressed cells, HSF1 is inhibited through its associations with Hsp90, Hsp70, and additional co-chaperones (25, 26, 39 -42). Mapping experiments have revealed that the interaction of HSF1 with Hsp70 is mediated by a 15-amino acid sequence within the C-terminal transactivation domain (43). A mechanism for the HNE-mediated activation of HSF1 has not been established. One possibility involves direct modification of Hsp90 or Hsp70 residues by HNE, thereby abolishing their inhibition of HSF1. In support of this hypothesis, HNE has been shown to inhibit the chaperone activities of both Hsp90 and Hsp70 in vitro. Moreover, inhibition of Hsp90 and Hsp70 was correlated with the identification of specific HNE amino acid adducts by tandem mass spectrometry analysis (27,44). Our  data using co-immunoprecipitation of c-Myc-tagged Hsp70 suggest that HNE treatment inhibits the interaction of Hsp70 with HSF1, which we suspect liberates HSF1 and facilitates its nuclear accumulation.
In addition to acting as protein chaperones, Hsps inhibit cell death by directly inhibiting of a variety of pro-apoptotic mediators. For example, Hsp70 associates with and inhibits the activation of JNK1 (45). By inhibiting JNK1, Hsp70 prevents Bid cleavage, thereby decreasing expression of the pro-apoptotic tBid fragment and consequent cytochrome c release (46). Our data showing enhanced JNK phosphorylation in HSF1-deficient cells support these observations. Apoptosis is also inhibited downstream of cytochrome c release by the direct interactions of Hsp70 and Hsp90 with apoptotic peptidase activating factor 1, thereby hindering formation of the apoptosome (47,48). Furthermore, Hsp70 inhibits apoptosis-inducing factor, a mitochondrial flavoprotein whose nuclear expression in apoptotic cells promotes PARP cleavage and chromatin condensation (49). Recently, Hsp70 was shown to inhibit the cleavage of the transcription factor GATA-1 by activated caspase 3, suggesting that the protection of client proteins from caspase-mediated proteolysis may be an additional means of inhibiting apoptosis (50).
Evidence suggests that Bcl-X L is critical in mediating the resistance of eukaryotic cells to chemical and redox stress (51,52). Previously, induction of the heat shock response was shown to enhance Bcl-X L expression, and conversely, silencing Hsp70 with antisense RNA lead to diminished levels of Bcl-X L (53,54). We found that pro-apoptotic concentrations of HNE caused a dramatic decrease in Bcl-X L expression, whereas Bcl-2, Bad, and Bax were unaffected. Moreover, silencing HSF1 exacerbated the decrease in Bcl-X L levels by HNE. The specific down-regulation of Bcl-X L is recently emerging as a consequence of exposure of cells to chemotherapeutics, including cisplatin, bischloronitrosourea (carmustine), retinoic acid as well as to UV light (55)(56)(57)(58). Our finding that HSF1-mediated gene expression stabilizes Bcl-X L levels implies that activation of the heat shock response may partly mediate chemotherapeutic resistance.
The consequences of heat shock response activation and Hsp induction likely extend beyond cytoprotection from electrophilic lipids. For example, Hsp70 has been shown to directly inhibit a variety of signaling pathways, including the lipopolysaccharide-induced innate immune response. As a conse-   12. Proposed model for cytoprotective activation of heat shock response. HNE disrupts the inhibitory association of Hsp70 (and possibly Hsp90) with HSF1, which accumulates in the nucleus and subsequently promotes the expression of HSE-containing genes. Our data demonstrate that silencing of HSF1 enhanced HNE-mediated JNK phosphorylation, which has been correlated with HNE-induced apoptosis. In addition, HSF1 reduction facilitated the HNE-mediated inhibition of Bcl-X L levels, implying a role for induction of the heat shock response in stabilizing Bcl-X L expression and the consequent inhibition of apoptosis. Furthermore, overexpression of Bcl-X L attenuated HNE-mediated apoptosis in HSF1-deficient cells.
Our findings indicate that a fundamental consequence of promoting HSF1-mediated gene expression is enhanced cell survival under otherwise lethal concentrations of HNE. In our experiments, the magnitude of cytoprotection exceeded the contribution of Nrf2 and induction of the antioxidant response. In view of the complex biology of HNE and similar endogenous metabolites, the role of HSF1 in a physiological context is undoubtedly significant. For example, the modification of DNA by HNE yields potentially mutagenic lesions such as N2-etheno-dG and cyclic N2-propano-dG adducts, which have been found at significant levels in disease states including chronic pancreatitis, ulcerative colitis, and Crohn disease (64 -68). In such a context, the combined effects of DNA damage and cytoprotection imply a heightened potential for the accumulation of mutations. Overall, through its ability to enhance cell survival by inhibiting pro-apoptotic pathways, we speculate that the heat shock response contributes significantly in the cellular biology and pathophysiology of HNE and similar oxidized lipid electrophiles.