SH-SY5Y, HeLa and HEK293T cells were obtained from the American Type Culture Collection (ATCC). Cells were grown in DMEM supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin in a 5% CO₂ humidified atmosphere at 37°C. To induce cellular stress, cells were treated with 1 µg/ml doxorubicin or 0.3 mM H₂O₂ for 24 h.

Full-length cDNAs for human Hsp70 (Genbank: NM_005345.5) and human ARD1 (Genbank: NM_003491.3) were generated by PCR and subcloned into pCDNA3.1 (FLAG-ARD1) or pEGFP-C3 (GFP-Hsp70) vectors for cellular expression. For the construction of stable cell lines, cDNA constructs for Hsp70 and ARD1 were co-inserted into the pIRES vector, purchased from Clontech.

Transfection was carried out as described previously (20). HEK293T cell transfection used polyethyleneimine (PEI) at a ratio of 4:1 (ml PEI/mg plasmid DAN) in basal media overnight, followed by a change of media. For the establishment of stable cells, pEGFP-C3-Hsp70 and pIRES-GFP-Hsp70-FLAG-ARD1 plasmids were transfected into SH-SY5Y cells using PolyFect reagent (Qiagen), according to the manufacturer's instructions. Transfected cells were maintained in complete DMEM with G418 (500 µg/ml). After several days, the surviving colonies were selected and amplified.

Anti-cleaved caspase-3 antibody (#9661, 1:3,000), anti-PARP antibody (#9542, 1:3,000), anti-Apaf-1 (#8969, 1:2,000) antibody, anti-AIF antibody (#4642, 1:2,000), anti-Atg12 antibody (#4180, 1:3,000), anti-Beclin-1 antibody (#3495, 1:3,000), and anti-LC3A/B antibody (#12741, 1:200) were purchased from Cell Signaling Technology. Anti-Myc antibody (9E10, sc-40, 1:3,000) and anti-lamin A antibody (4A58, sc-71481, 1:2,000) were purchased from Santa Cruz. The anti-FLAG antibody (M2, F1804, 1:3,000), anti-tubulin antibody (DM1A, T9016, 1:3,000) and anti-β-actin (A2066, 1:3,000) antibody were purchased from Sigma. The anti-GFP antibody (ab6556, 1:3,000) was from Abcam.

Cultured cells were washed with PBS, homogenized in buffer A [10 mM HEPES (pH 7.4), 1.5 mM MgCl₂, 10 mM KCl, 0.5 mM DTT, 0.1% NP-40], and centri-fuged at 600 × g for 10 min at 4°C. The nuclear pellet was washed with buffer A, resuspended in buffer C [10 mM HEPES (pH 7.4), 1.5 mM MgCl₂, 0.5 mM DTT, 20% glycerol, 0.5 mM PMSF, 0.2 mM EDTA, and 420 mM NaCl], and centrifuged for 30 min at 15,000 rpm, before the supernatant was isolated.

Proteins were extracted using lysis buffer consisting of 20 mM Tris (pH 7.5), 150 mM NaCl, 0.1 mM EDTA, 0.2% Triton X-100 and a protease inhibitor cocktail (Roche). Then, 20 mg of cell extract was used for immunoblotting. For immunoprecipitation, 1 mg of protein was incubated with a corresponding antibody conjugated to A or G beads (Upstate) overnight at 4°C. Beads were washed three times with washing buffer containing 20 mM Tris (pH 7.5), 150 mM NaCl and 0.1 mM EDTA. Following SDS-PAGE, membranes were immunoblotted using the corresponding primary antibody overnight at 4°C. HRP-conjugated secondary antibodies were incubated with the membranes for 1 h at room temperature. Visualization was performed using ECL Plus (Intron) and LAS-4000 (GE Healthcare).

HeLa cells were seeded onto glass coverslips in 24-well plates and transfected with GFP-Hsp70 WT and K77R. After treatment of 0.3 mM H₂O₂ for 24 h, cells were fixed in 4% PFA for 20 min and permeabilized in 0.3% Triton X-100 in PBS for 5 min at room temperature. Then, cells were incubated with LC3A/B antibody (Cell Signaling Technology, #12741, 1:200) and visualized with Alexa 546-conjugated IgG (Molecular Probes). Nuclear staining was performed using Hoechst 33342 (Molecular Probes). The immunofluorescence was visualized using an Axiovert M200 microscope (Carl Zeiss).

Cell viability was calculated by measuring the amount of lactate dehydrogenase (LDH) released from the cells into the medium. Conditioned media from cultured cells were collected, and LDH activity was determined with an LDH assay kit (DoGen). Total cellular LDH activity was measured by solubilizing the cells with 0.2% Triton X-100.

Results are expressed as the means ± SD P-values were calculated by applying the two-tailed Student's t-test. A difference was considered statistically significant at P<0.05.

Hsp70 acetylation at residue K77 has been previously reported to protect cells from stress. To confirm the protective effect of Hsp70 acetylation, we treated Hsp70 wild-type (WT) and K77R mutant-expressing SH-SY5Y cells with doxorubicin and analyzed cell viability. Consistent with a previous report, overexpression of wild-type Hsp70 protected cells against doxorubicin-induced cell death, but this protective effect was eliminated by the presence of the K77R mutation (Fig 1A). The acetylation of Hsp70 at K77 is mediated by acetyltransferase ARD1 (19). To further validate the relevance of ARD1 in the protective function of Hsp70, ARD1 WT and a dominant-negative mutant (DN) that does not harbor acetyltransferase activity was co-expressed with Hsp70 constructs in SH-SY5Y cells, and cell viability was assessed after doxorubicin treatment (21). As expected, co-expression of the dominant-negative mutant ARD1 abolished the protective effect of Hsp70 WT (Fig. 1B), indicating that ARD1-mediated Hsp70 acetylation protects cancer cells from doxorubicin-induced cell death.

To further investigate the underlying mechanisms of Hsp70 acetylation-mediated cellular protection, we first analyzed caspase-dependent apoptosis. Doxorubicin treatment induced cleavage of caspase-3 and poly(ADP-ribose) polymerase (PARP), which are hallmarks of apoptosis (Fig. 2A). Consistent with the doxorubicin-induced cell death shown in Fig. 1, Hsp70 WT overexpression prevented doxorubicin-induced cleavage of caspase-3 and PARP, whereas the presence of the K77R mutation abolished the protective effect. Moreover, co-expression of the dominant-negative mutant ARD1 diminished Hsp70 acetylation-mediated inhibition of PARP and caspase cleavage (Fig. 2B).

Since Hsp70 is reported to inhibit apoptosome formation via binding with Apaf-1, a key molecule of caspase-dependent apoptosis, we next investigated whether acetylation at K77 can modulate Apaf-1 binding. Hsp70 WT or the K77R mutant was overexpressed in HEK293T cells and their affinity to Apaf-1 was assessed by co-immunoprecipitation. Consistent with previous reports, Hsp70 wild-type was co-immunoprecipitated with Apaf-1 in cell extracts. Interestingly, however, mutation at K77 abrogated its binding to Apaf-1, implying that Hsp70 acetylation contributes to Hsp70/Apaf-1 association, subsequently leading to inhibition of functional apoptosome assembly and caspase activation (Fig. 2C). These results suggest that Hsp70 acetylation enables the association of Apaf-1 with Hsp70 and prevents apoptosis.

Another cell death pathway in which Hsp70 is involved is AIF-dependent apoptosis. Hsp70 inhibits caspase-independent cell death by sequestering AIF and blocking its induction of apoptosis. To examine its relevance to caspase-independent apoptosis, we analyzed AIF binding to Hsp70 WT and the K77R mutant by co-immunoprecipitation (Fig. 3A). Interestingly, compared to Hsp70 WT, AIF binding to the K77R mutant was significantly reduced, implying that Hsp70 acetylation can interfere with AIF-mediated caspase-independent apoptosis.

We next verified whether this change in binding affinity was indeed relevant for AIF function. Nuclear translocation of AIF is a key final step in the AIF pathway (14). To validate the difference in binding affinity and whether it influences AIF nuclear translocation, Hsp70-expressing cells were treated with hydrogen peroxide and nuclear translocation of AIF was greatly increased (Fig. 3B). Moreover, Hsp70 WT reduced H₂O₂-induced nuclear translocation of AIF but the K77R mutant could not. These results demonstrate that Hsp70 K77 acetylation suppresses apoptosis by preventing AIF from nuclear translocation, consequently inhibiting caspase-independent cell death. Collectively, these findings indicate that Hsp70 acetylation protects cancer cells against apoptosis through inhibition of both caspase-dependent and -independent apoptotic pathways.

Autophagic cell death is another type of programmed cell death that is controlled by Hsp70. Recent studies of stress-induced Hsp70 acetylation and its modulatory role in autophagosome formation led us to hypothesize that Hsp70 acetylation at K77 can also regulate autophagy (22). To test this, we treated H₂O₂ to Hsp70 WT and K77R mutant-expressing cells and assessed autophagy-related proteins. Autophagy-related genes (Atg) are universal markers for autophagic induction (23). During stepwise autophagy induction, Beclin-1 (the mammalian orthologue of yeast Atg6)-mediated core complex formation and Atg12-Atg5 conjugation are key processes, in nucleation and elongation of autophagosome formation, respectively (24,25). Consistent with previous reports, Hsp70 WT overexpression decreased H₂O₂-induced Beclin-1 and Atg12-Atg5 conjugation, suggesting that Hsp70 plays a regulatory role in the modulation of autophagy (Fig. 4A). However, the Hsp70 K77R mutant could not prevent autophagy as demonstrated by an increase in Beclin-1 and Atg12-Atg5 conjugation. This observation implies that Hsp70 acetylation contributes to the attenuation of autophagic cell death.

We also analyzed microtubule-associated protein light chain 3 (LC3) to monitor autophagic induction. Upon autophagy, the unconjugated cytosolic form of LC3-I is converted to the phosphatidylethanolamine-conjugated form of LC3-II that forms the autophagosomal membrane (26,27). Therefore, the transition of LC3 from a diffusive cytoplasm pattern to the punctated membrane pattern is a hallmark of autophagy induction, indicating the formation of autophagic vacuoles (27). When compared to WT, the Hsp70 mutant caused an increase in LC3 expression and perinuclear autophagic vacuole formation (Fig. 4B). These results suggest that Hsp70 acetylation may play a role in the prevention of autophagic cell death.