Article2 November 1998free access Hsp70 exerts its anti-apoptotic function downstream of caspase-3-like proteases Marja Jäättelä Corresponding Author Marja Jäättelä Apoptosis Laboratory, Institute of Cancer Biology, Danish Cancer Society, Strandboulevarden 49, DK-2100 Copenhagen, Denmark Search for more papers by this author Dorte Wissing Dorte Wissing Apoptosis Laboratory, Institute of Cancer Biology, Danish Cancer Society, Strandboulevarden 49, DK-2100 Copenhagen, Denmark Search for more papers by this author Klaus Kokholm Klaus Kokholm Apoptosis Laboratory, Institute of Cancer Biology, Danish Cancer Society, Strandboulevarden 49, DK-2100 Copenhagen, Denmark Search for more papers by this author Tuula Kallunki Tuula Kallunki Apoptosis Laboratory, Institute of Cancer Biology, Danish Cancer Society, Strandboulevarden 49, DK-2100 Copenhagen, Denmark Search for more papers by this author Mikala Egeblad Mikala Egeblad Apoptosis Laboratory, Institute of Cancer Biology, Danish Cancer Society, Strandboulevarden 49, DK-2100 Copenhagen, Denmark Search for more papers by this author Marja Jäättelä Corresponding Author Marja Jäättelä Apoptosis Laboratory, Institute of Cancer Biology, Danish Cancer Society, Strandboulevarden 49, DK-2100 Copenhagen, Denmark Search for more papers by this author Dorte Wissing Dorte Wissing Apoptosis Laboratory, Institute of Cancer Biology, Danish Cancer Society, Strandboulevarden 49, DK-2100 Copenhagen, Denmark Search for more papers by this author Klaus Kokholm Klaus Kokholm Apoptosis Laboratory, Institute of Cancer Biology, Danish Cancer Society, Strandboulevarden 49, DK-2100 Copenhagen, Denmark Search for more papers by this author Tuula Kallunki Tuula Kallunki Apoptosis Laboratory, Institute of Cancer Biology, Danish Cancer Society, Strandboulevarden 49, DK-2100 Copenhagen, Denmark Search for more papers by this author Mikala Egeblad Mikala Egeblad Apoptosis Laboratory, Institute of Cancer Biology, Danish Cancer Society, Strandboulevarden 49, DK-2100 Copenhagen, Denmark Search for more papers by this author Author Information Marja Jäättelä 1, Dorte Wissing1, Klaus Kokholm1, Tuula Kallunki1 and Mikala Egeblad1 1Apoptosis Laboratory, Institute of Cancer Biology, Danish Cancer Society, Strandboulevarden 49, DK-2100 Copenhagen, Denmark *Corresponding author. E-mail: [email protected] The EMBO Journal (1998)17:6124-6134https://doi.org/10.1093/emboj/17.21.6124 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The major heat shock protein, Hsp70, is an effective inhibitor of apoptosis. To study its mechanism of action, we created tumor cell lines with altered Hsp70 levels. The expression levels of Hsp70 in the cells obtained correlated well with their survival following treatments with tumor necrosis factor, staurosporine and doxorubicin. Surprisingly, the surviving Hsp70-expressing cells responded to the apoptotic stimuli by activation of stress-activated protein kinases, generation of free radicals, early disruption of mitochondrial transmembrane potential, release of cytochrome c from mitochondria and activation of caspase-3-like proteases in a manner essentially similar to that of the dying cells with low Hsp70 levels. However, Hsp70 inhibited late caspase-dependent events such as activation of cytosolic phospholipase A2 and changes in nuclear morphology. Furthermore, Hsp70 conferred significant protection against cell death induced by enforced expression of caspase-3. Thus, Hsp70 rescues cells from apoptosis later in the death signaling pathway than any known anti-apoptotic protein, making it a tempting target for therapeutic interventions. Introduction Tissue homeostasis is dependent on the dynamic balance between cell proliferation and apoptosis (Kerr et al., 1972; reviewed by White, 1996; Nagata, 1997). Until recently, cancer researchers have focused most of their attention on pathways within the cell that induce proliferation, but the realization that malignant tumors may grow because cells have lost the ability to die has provided a new direction for investigation. Apoptosis occurs in tumors responding to tumor necrosis factor (TNF) and related cytokines, cytotoxic lymphocytes, hormone ablation, starvation, irradiation and chemotherapy (reviewed by Kerr et al., 1994; White, 1996; Nagata, 1997). In spite of the wide variety of stimuli that can activate apoptosis, the signaling pathways induced by them converge into a common death pathway either at a mitochondrial step culminating in the release of cytochrome c from mitochondria to the cytosol (X.Liu et al., 1996; Kluck et al., 1997; Li et al., 1997; Yang et al., 1997; Bossy-Wetzel et al., 1998) or at the following step at which Asp-Glu-Val-Asp (DEVD)-specific caspase-3-like cysteine proteases, also called the executioners of apoptosis, are activated (X.Liu et al., 1996; Li et al., 1997). Active caspases cleave a limited set of cellular proteins. Caspase-mediated cleavage often inhibits the function of their substrates but, interestingly, some enzymes, such as protein kinase C δ (PKCδ) (Ghayur et al., 1996), MEKK1 (Cardone et al., 1997), DNA fragmentation factor (Liu et al., 1997), PITSLRE kinases (Beyaer et al., 1997), p21-activated kinase hPAK65 (Lee et al., 1997), Mst1 kinase (Graves et al., 1998) and cytosolic phospholipase A2 (cPLA2) (Wissing et al., 1997), are actually activated by caspases either directly by the cleavage of the enzyme or indirectly by activation/inactivation of regulatory proteins. These and other as yet unknown caspase-activated enzymes could be the actual downstream effectors of apoptosis, whose activation might result in the typical apoptotic morphology including membrane blebbing, nuclear condensation and cell shrinkage. Tumor cells express several proteins that suppress apoptosis. Of special interest in cancer research are those that function at the common part of the apoptotic signaling pathway and thus inhibit apoptosis induced by a wide range of stimuli. Increased activity of such protective proteins, such as anti-apoptotic members of the Bcl-2 protein family (Vaux et al., 1988; Boise et al., 1993), members of the inhibitor of apoptosis protein (IAP) family (Rothe et al., 1995; Deveraux et al., 1997) and Hsp70 (Jäättelä et al., 1992; Mosser et al., 1997), may result in aggressively growing and therapy-resistant tumors. Recent results have elucidated the mechanisms of action of some of the anti-apoptotic proteins. Bcl-2 inhibits the fall in the mitochondrial transmembrane potential (Δψm) and the release of cytochrome c from the mitochondria (Shimizu et al., 1996; Kluck et al., 1997; Yang et al., 1997), whereas IAPs function by inhibiting the caspase activity (Deveraux et al., 1997). The mechanism of action by which Hsp70 exerts its anti-apoptotic effect is obscure. Its role in cellular resistance against stress was first suggested based on the correlation between its synthesis and the development of resistance to elevated temperatures (Li and Werb, 1982). More direct evidence supporting the role of Hsp70 in thermotolerance came from studies showing that inhibition of its expression or neutralization of its function rendered cells extremely sensitive to heat (Jonston and Kucey, 1988; Riabolow et al., 1988). Recent results have made it clear that Hsp70 is a general anti-apoptotic protein, which protects cells from cytotoxicity induced by TNF, monocytes, oxidative stress, chemotherapeutic agents, ceramide and radiation (Jäättelä et al., 1992; Jäättelä and Wissing, 1993; Simon et al., 1995; Karlseder et al., 1996; Mosser et al., 1997). To study at which step of the apoptotic signaling pathway Hsp70 functions, we used two model systems; inhibition of Hsp70 expression by antisense hsp70 cDNA in ME–180 cervix carcinoma cells with high constitutive Hsp70 expression, and overexpression of Hsp70 in WEHI–S fibrosarcoma cells with low endogenous Hsp70 levels. Interestingly, the analysis of death signaling pathways in these cells revealed that while Hsp70 protected cells from apoptosis, it did not inhibit activation of stress-activated protein kinases (SAPK or c-Jun N-terminal kinase, JNK), the early mitochondrial changes or the activation of caspase-3-like proteases. Hsp70 did, however, inhibit the function of a caspase-activated cPLA2, the ensuing late Δψm collapse and apoptotic changes in nuclear morphology. Remarkably, Hsp70-expressing cells in which cytochrome c was released from mitochondria and downstream caspases were activated acquired normal growth shortly after the otherwise lethal stimuli. These data indicate that ‘the point of no return’ in apoptosis signaling is further downstream than previously believed. Results Hsp70 protects cells against TNF- and drug-induced apoptosis ME–180 human cervical carcinoma cells express high levels of Hsp70 constitutively (Figure 1A). To reduce the Hsp70 expression, we transfected them with three different antisense hsp70 cDNA constructs. G418-resistant cell pools (ME–ashsp1-P, ME–ashsp2-P and ME–ashsp3-P) and single-cell clones (ME–ashsp1-1, ME–ashsp1-2, ME–ashsp2-1 and ME–ashsp3-1) together with their vector-transfected controls (ME–vector-P and ME–vector-1) were then tested for Hsp70 expression by immunoblot analysis (Figure 1A). All three antisense constructs reduced Hsp70 protein levels as compared with vector-transfected control cells. The expression of the closely related Hsc70 was not affected. The antisense Hsp70 transfectants were significantly more susceptible than vector-transfected control cells to the cytotoxicity induced by TNF as analyzed by a 48 h cell viability assay (Figure 1A). The single cell clone ME–ashsp2-1 and the pool ME–ashsp3-P with lowest Hsp70 levels were the most TNF sensitive. In addition to TNF sensitivity, low Hsp70 expression correlated with the sensitivity of ME–180 transfectants to death induced by staurosporine and doxorubicin (Figure 1B). Figure 1.Inhibition of Hsp70 synthesis sensitizes ME–180 cells to TNF, staurosporine and doxorubicin. (A) Single-cell clones (ME–ashsp1-1, ME–ashsp1-2, ME–ashsp2-1 and ME–ashsp3-1) and pools (ME–ashsp1-P, ME–ashsp2-P and ME–ashsp3-P) of ME–180 cells transfected with antisense Hsp70 constructs and their respective controls (ME–vector-1 and ME–vector-P) were plated on 96-well plates at a density of 8000 cells/0.1 ml of RPMI medium supplemented with 0.5% FCS and, 24 h later, treated with TNF for 48 h. The survival of TNF-treated cells is presented as a percentage of the survival of untreated cells as measured by an MTT assay. The experiment was repeated four times with essentially the same results. The amount of Hsp70 and Hsc70 in cell lysates from ∼3×105 cells was determined by immunoblot. (B) The sensitivity of cells to 100 nM staurosporine and 500 ng/ml of doxorubicin was analyzed by a 48 h MTT assay as above. The experiment was repeated twice with essentially the same results. (C) Approximately 1000 cells were plated in 96-well plates in RPMI + 0.5% FCS. After 24 h incubation, 10 ng/ml of TNF was added and the cell density was measured at the indicated times by an MTT assay. The experiment was repeated once with essentially the same results. The means of triplicate samples ± SD are shown (A–C). Download figure Download PowerPoint Unstressed WEHI–S murine fibrosarcoma cells express very low levels of Hsp70. To obtain another model system to study the mechanism of action of Hsp70, we transfected WEHI–S cells with human hsp70 cDNA. Two single-cell clones with the highest Hsp70 expression levels (WEHI-hsp-2 and WEHI-hsp-18) and their vector-transfected control clones (WEHI-vector-2 and WEHI-vector-3) were tested for their sensitivity to TNF, staurosporine and doxorubicin in an 18 h cell viability assay. As expected based on earlier results (Jäättelä et al., 1992; Simon et al., 1995), Hsp70 expression in the transfectants obtained correlated with the resistance against cell death induced by TNF and doxorubicin (Figure 2). Moreover, Hsp70 conferred resistance against staurosporine, further indicating that Hsp70 protects cells from a wide range of inducers of apoptosis (Figure 2). The protective effect of Hsp70 against TNF, staurosporine and doxorubicin in both ME–180 and WEHI–S cells was confirmed by a chromium release assay and by counting of cells with apoptotic morphology (data not shown). Figure 2.Overexpression of Hsp70 renders WEHI–S cells resistant to TNF, staurosporine and doxorubicin. (A) Immunoblot of Hsp70 protein expression in vector-transfected (WEHI-vector-2 and WEHI-vector-3) and Hsp70-transfected (WEHI-hsp-2 and WEHI-hsp-18) single-cell clones of WEHI–S cells. (B) Cells were plated on 96-well plates at a density of 15 000 cells/0.1 ml of complete medium. After 24 h, 100 pg/ml of TNF, 100 nM staurosporine or 1 μg/ml of doxorubicin were added and the survival of the cells was measured 18 h later by an MTT assay. The experiment was repeated three times with essentially the same results. (C) Approximately 2000 cells were plated in 96-well plates in RPMI medium containing 5% FCS. After 24 h incubation, 100 pg/ml of TNF was added and the cell density was measured at the indicated times by an MTT assay. The experiment was repeated once with essentially the same results. The means of triplicate samples ± SD are shown (B and C). Download figure Download PowerPoint To test whether the Hsp70-associated protection from apoptosis described above was due to a delay in the cell death, we followed the growth of TNF-treated ME–180 and WEHI–S transfectants for 96 and 60 h, respectively. Remarkably, ∼80% of the Hsp70-expressing ME–vector-P and WEHI-hsp-2 cells, which survived the TNF treatment, reached a normal growth rate after 48 and 35 h, respectively (Figures 1C and 2C). Hsp70 does not affect apoptosis-associated activation of SAPK/JNK ME–180 and WEHI–S cells with altered levels of Hsp70 were then used to study the mechanism of the anti-apoptotic action of Hsp70. The AP-1-like transcription factors, which are activated by SAPK/JNK-mediated phosphorylation of c-Jun, have been suggested to mediate TNF- and stress-induced apoptosis (Xia et al., 1995; Verheij et al., 1996). Therefore, we studied the ability of Hsp70 to interfere with the SAPK/JNK pathway by a kinase assay employing c-Jun as a substrate and by an AP-1-responsive luciferase reporter gene assay. TNF activated SAPK/JNK in both ME–180 and WEHI–S transfectants, whereas only WEHI–S transfectants responded to staurosporine. In all cases, the SAPK/JNK activation was rapid and transient, reaching a peak at 30–60 min. In repeated experiments, no significant difference could be seen in the TNF-induced SAPK/JNK activities of ME–vector-2 and ME–ashsp2-1 cells, the maximal inductions being 72- and 58-fold, respectively (Figure 3A). WEHI-vector-2 and WEHI-hsp-2 cells responded to 100 pg/ml of TNF with up to 4.7- and 4.9-fold induction in SAPK/JNK activity, and to 100 nM staurosporine with up to 41- and 32-fold induction, respectively (data not shown). In conclusion, Hsp70 had no significant effect on the activation of SAPK/JNK. Surprisingly, neither TNF- nor staurosporine-induced activation of SAPK/JNK led to the activation of AP-1-like transcription factors as measured by a luciferase assay (Figure 3B). Contrary to apoptotic stimuli, phorbol 12-myristate-13-acetate (PMA) activated AP-1-like transcription factors (Figure 3B). Figure 3.Hsp70 has no effect on apoptosis-associated activation of SAPK/JNK. (A) Cells were treated with 10 ng/ml of TNF for the indicated times prior to harvesting and measuring the SAPK/JNK activity using the solid-phase kinase assay. (B) Luciferase activity was measured in cells transiently transfected with pAP(4)-LUC plasmid containing four AP-1-binding sites upstream of the P.pyralis luciferase-coding sequence. Two days after the transfection, cells were stimulated for 5 h with 100 ng/ml of PMA, 20 ng/ml (ME–180 transfectants) or 1 ng/ml (WEHI transfectants) of TNF, or 100 nM staurosporine. The values represent means of a duplicate experiment ± SD. The experiment was repeated once with essentially the same results. Download figure Download PowerPoint Hsp70 does not affect apoptosis-associated early mitochondrial changes but blocks the late fall in Δψm Heat shock has been shown to protect mitochondria from hydrogen peroxide-induced damage (Polla et al., 1996). Thus, we studied the ability of Hsp70 to protect cells from apoptosis-associated mitochondrial alterations, i.e. free radical generation, reduction in the Δψm and release of cytochrome c from mitochondria. As shown in Figure 4A, staurosporine induced a similar early free radical burst in vector-transfected and Hsp70-overexpressing WEHI–S cells. Treatment of WEHI–S transfectants with up to 1 ng/ml of TNF and ME–180 transfectants with up to 100 ng/ml of TNF or 100 nM staurosporine failed to induce measurable generation of free radicals. Figure 4.Hsp70 has no effect on the generation of free radicals or the early reduction in Δψm, but inhibits the late Δψm fall. (A) Generation of free radicals in untreated cells and cells treated with 100 nM staurosporine was analyzed by flow cytometry analysis of DHR123-labeled cells. The data are presented as mean fluorescence intensity. (B and C) The Δψm was analyzed by flow cytometry analysis of RH123-labeled cells. The data are presented as the percentage of the mean fluorescence intensity in staurosporine-treated cells as compared with untreated cells (B) or as the percentage of TNF-treated cells with reduced Δψm (C). Download figure Download PowerPoint In some apoptosis models, the reduction in the Δψm is an early event which precedes the release of caspase-activating factors from mitochondria, whereas in others the cytochrome c release occurs independently of the fall in the Δψm (Zamzami et al., 1995; Susin et al., 1997; Bossy-Wetzel et al., 1998). In models used in this study, only ME–180 transfectants treated with staurosporine responded with a reduction in Δψm prior to the cell death. Hsp70 had, however, no effect on this early staurosporine-induced Δψm fall (Figure 4B). In staurosporine-treated WEHI–S cells, Δψm fell first at the time of cell death, ∼10–12 h after the addition of the drug. TNF induced a similar late Δψm disruption, which occurred simultaneously with the cell death, ∼2–3 and 12 h after the release of cytochrome c in WEHI–S and ME–180 cells, respectively. As shown for TNF-treated ME–180 transfectants, the late Δψm fall was inhibited by Hsp70 in a manner similar to cell death (Figure 4C). The release of cytochrome c from mitochondria into the cytosol is a crucial step in apoptosis signaling as it results in the activation of downstream caspases (Li et al., 1997; Bossy-Wetzel et al., 1998). Treatment of ME–180 cells with TNF or staurosporine resulted in a clear reduction in the mitochondrial cytochrome c levels at ∼8 and 5 h, respectively. The disappearance of cytochrome c from the mitochondrial fraction was not due to a complete destruction of the mitochondria as the level of another mitochondrial protein, Grp75, remained unchanged in this fraction. Surprisingly, Hsp70 had no effect on either TNF- (Figure 5A) or staurosporine-induced (Table I) cytochrome c efflux. WEHI–S cells also responded to TNF and staurosporine by a clear reduction in mitochondrial cytochrome c levels at ∼1.5 and 2.5 h, respectively. Hsp70 expression levels in WEHI–S cells had no effect on apoptosis-associated release of cytochrome c (Figure 5B, Table I). Figure 5.Hsp70 does not inhibit TNF-induced release of cytochrome c from mitochondria. ME–180 transfectants were treated with 10 ng/ml of TNF (A) and WEHI–S transfectants with 100 pg/ml of TNF (B) for the indicated times before harvesting. Mitochondrial extracts were then prepared and analyzed for cytochrome c (cc) and mitochondrial heat shock protein (grp75) by immunoblot. Download figure Download PowerPoint Table 1. A summary of the effects of Hsp70 on TNF- and staurosporine-induced cellular events Eventa WEHI-hsp/WEHI-vector (%) ME–vector/ME–ashsp (%) TNF Staurosporine TNF Staurosporine JNK activation 138 ± 36 (2)b 103 ± 25 (2) 99 ± 3 (2) N.A. ROI release N.A.c 96 ± 11 (3) N.A. N.A. Early Δψm fall N.A. N.A. N.A. 110 ± 2 (2) Cytochrome c release 100 ± 0 (3) 100 ± 0 (2) 97 ± 9 (3) 100 (1) DEVDase activation 106 ± 4 (8) 115 ± 14 (4) 117 ± 9 (8) 114 ± 8 (4) AA release 15 ± 4 (4) 45 ± 5 (4) 35 ± 8 (4) 51 ± 9 (4) Nuclear changes 0 ± 0 (2) 27 ± 24 (2) N.D.d N.D. Late Δψm fall 15 ± 5 (2) 29 ± 1 (2) 33 ± 12 (2) N.A. Death 25 ± (10) 55 ± 11 (6) 30 ± 6 (10) 65 ± 5 (6) a TNF- and staurosporine-induced cellular events are presented in the order of their appearance. b The effect of Hsp70 expression on the induction of the indicated events is expressed as a percentage ± SD of the TNF- or staurosporine-induced change in cells with high Hsp70 expression as compared with the change in cells with low Hsp70 expression. The calculations were made using values representing the maximal inductions of signaling events occurring before the onset of death, except for values for late Δψm fall, which occurred at the time at which cells began to die, and death, which was monitored 18 (WEHI–S) or 48 h after the addition of TNF or staurosporine. The number of experiments performed is given in parentheses. For methods used, see the legends to Figures 1,2,3,4,5,6 and 10. c N.A., not applicable as the indicated event does not occur. d N.D., not detected. Hsp70-mediated protection from apoptosis occurs downstream of caspase-3-like proteases The next possible step at which Hsp70 could inhibit apoptosis signaling is the cytochrome c-mediated activation of downstream caspases. Thus, we tested the effect of Hsp70 on the activation of DEVD-specific caspases using a colorimetric probe Ac-DEVD-pNA. Surprisingly, treatment of ME–180 transfectants with TNF, staurosporine or doxorubicin induced as strong an activation of DEVD-specific caspases in apoptosis-resistant control cells with high Hsp70 expression as in apoptosis-sensitive antisense Hsp70 transfectants (Figure 6A–C). TNF- but not staurosporine- or doxorubicin-induced activation of caspases was delayed by ∼4 h in control cells as compared with the antisense Hsp70 transfectants. WEHI–S cells also responded to apoptotic stimuli by activation of DEVD-specific caspases. In spite of the strong protective effect of Hsp70, it could not attenuate the activation of DEVD-specific caspases, but delayed their activation by TNF and staurosporine by ∼1 and 4 h, respectively (Figure 6D–F). Contrary to Hsp70, both Bcl-2, which protects ME–180 cells (Jäättelä et al., 1995), and CrmA, which protects WEHI–S cells from TNF-mediated apoptosis (Wissing et al., 1997), inhibited the activation of caspases (Figure 6). Figure 6.Hsp70 does not inhibit TNF-, staurosporine- or doxorubicin-induced activation of DEVD-specific caspases. ME–180 transfectants (A–C) and WEHI–S transfectants (D–F) were treated with the indicated concentrations of TNF (A and D), staurosporine (B and E) or doxorubicin (C and F). The DEVDase activity of the cell lysates was determined based on their ability to cleave Ac-DEVD-pNA. The mean values of triplicate samples ± SD are shown. The experiment was repeated three times (A and D) or once (B, C, E and F) with essentially the same results. Download figure Download PowerPoint As the assay system used measures the ability of caspases to cleave an artificial peptide target, Hsp70 could bind to important caspase substrates in vivo and thereby inhibit their cleavage. Therefore, we studied the effect of Hsp70 on TNF-induced cleavage of known targets for DEVD–specific caspases. As shown in Figure 7, poly(ADP-ribose) polymerase (PARP), cPLA2 and PKCδ are cleaved to the same extent in ME–180 cells with high and low Hsp70 expression. Similarly, PARP and cPLA2 cleavage were unaffected by Hsp70 in WEHI–S transfectants (data not shown). These data indicate either that Hsp70 protects cells from the consequences of proteolysis by caspase-3-like proteases or that apoptosis is mediated by a caspase-independent mechanism. To test the latter hypothesis, we first studied the ability of a specific tetrapeptide inhibitor of caspase-3-like proteases, Ac–DEVD-aldehyde (DEVD), to interfere with death signaling. As DEVD conferred a clear protection from cell death in both ME–180 and WEHI–S transfectants with low Hsp70 expression levels, the activity of DEVD-specific caspases must be necessary for the cytotoxicity in the model systems used (Figure 8). Figure 7.Hsp70 has no effect on the TNF-induced cleavage of caspase substrates. Subconfluent ME–180 transfectants were treated with 10 ng/ml of TNF for the indicated times, whole cell extracts were prepared and an equal amount of protein from each extract was analyzed for PARP, PKCδ and cPLA2 by immunoblot. Download figure Download PowerPoint Figure 8.DEVD-specific caspases are mediators of cell death induced by TNF, staurosporine and doxorubicin. The effect of DEVD on the sensitivity of ME–ashsp2-1 cells to 10 ng/ml of TNF (24 h), 100 nM staurosporine (24 h) and 1 μg/ml of doxorubicin (28 h) and that of WEHI-vector-2 cells to 1 ng/ml of TNF (12 h), 100 nM staurosporine (12 h) and 1 μg/ml of doxorubicin (22 h) was analyzed by a chromium release assay. When indicated (+ DEVD), DEVD at the final concentration of 200 (ME–ashsp2-1) or 50 μM (WEHI-vector-2) was added to the cells 1 h before the addition of the cytotoxic agent. The mean values of triplicate samples ± SD are shown. The experiment was repeated once with essentially the same results. Download figure Download PowerPoint Hsp70 protects WEHI–S cells from death induced by enforced expression of caspase-3 The above data suggest that Hsp70 protects cells from caspase-3-like proteases. Thus, we established a model system to investigate directly the ability of Hsp70 to protect cells from enforced expression of caspase-3. Caspase-3 cDNA and/or an empty vector were co-transfected with β-galactosidase cDNA into WEHI–S cells, and cells were incubated for 24 h in the presence or absence of a caspase–3 inhibitor, DEVD. Thereafter, cells were fixed and stained for β-galactosidase and the number of surviving transfectants was determined by counting β-galactosidase-positive blue cells. DEVD had no effect on the survival of vector-transfected cells but increased that of cells transfected with caspase-3 by >100% (Figure 9). These data indicate that enforced expression of caspase-3 leads to the activation of caspase-3 and to the death of WEHI–S cells. Remarkably, co-transfection of Hsp70 cDNA together with caspase-3 cDNA reduced the ability of DEVD to increase the survival of transfected cells to 16%, indicating that Hsp70 protects cells from caspase-3 almost as effectively as DEVD (Figure 9). Co-transfection of a cDNA encoding CrmA, a known caspase inhibitor, together with caspase-3 cDNA resulted in a protection from caspase-3-induced death similar to co-transfection with Hsp70, whereas a functionally defective mutant of CrmA could not protect WEHI–S cells from caspase-3. Thus, Hsp70 protects WEHI–S cells from caspase-3-induced cell death as effectively as CrmA. Figure 9.Hsp70, CrmA and DEVD protect WEHI–S cells from caspase-3-induced cell death. WEHI–S cells were transiently transfected with 3 μg of pcDNA3 or pcDNA-caspase-3 (caspase-3) and 6 μg of pSV2neo, pSV-Hsp70-tag (Hsp70), pcDNA-CrmA (CrmA) or pcDNA-CrmA-mut (CrmA-m) together with 2 μg of pCMV-βGal by electroporation. Transfected cells were incubated for 24 h in the absence (− DEVD) or presence (+ DEVD) of 75 μM DEVD, fixed and stained for β-galactosidase. The number of blue cells was determined by counting 2000 cells in five blindly chosen fields. The means of five determinations ± SD are shown. The experiment was repeated once with essentially the same results. Download figure Download PowerPoint Hsp70 inhibits caspase-3-mediated apoptotic events The ability of Hsp70 to inhibit caspase-3-induced cell death without inhibiting its proteolytic activity suggests that Hsp70 protects cells from the consequences of caspase-3-mediated proteolysis. We have shown earlier that cPLA2, which mediates the release of arachidonic acid (AA) from the sn-2 position of phospholipids, is one of the caspase-activated mediators of TNF-induced cell death (Wissing et al., 1997). As shown in Figure 10A, staurosporine and doxorubicin also induced a caspase-dependent release of AA from ME–ashsp3-P and WEHI-vector-2 cells with low Hsp70 expression before the onset of cell death. Interestingly, the AA release induced by TNF, staurosporine and doxorubicin was significantly reduced in Hsp70-expressing ME–vector-P and WEHI-hsp-2 cells as compared with low Hsp70-expressing ME–ashsp3-P and WEHI-vector-2 cells, respectively (Figure 10A). Figure 10.DEVD and Hsp70 inhibit apoptosis-associated arachidonic acid release (A) and nuclear changes (B). (A) [3H]Arachidonic acid-labeled cells were treated with 100 pg/ml of TNF for 4 h (WEHI transfectants), 10 ng/ml of TNF for 16 h (ME–180 transfectants), 100 nM staurosporine for 5 h or 1 μg/ml of doxorubicin for 16 h. The release of radioactivity into the supernatant was measured and is presented as a percentage of