DAPK interacts with HSF1 in vivo and in vitro. (A) A model of the DAPK kinase domain in complex with a substrate peptide comprising HSF1 residues 227–233. Left panel, the backbone topology of DAPK is shown in shades from blue to green and the bound ATP is shown in stick presentation (coloring according to the atom types). The key residues Arg227 and Ser230 of the HSF-1 peptide and Glu100 of DAPK are shown in stick presentation. The Ser230 sidechain is located in proximity to the terminal phosphoryl group of the ATP and Arg227 interacts with Glu100 and the ribose oxygens of ATP (magenta arrow). Right panel, space-filled representation of the complex (same view as in the left panel). The kinase domain is shown in cyan and the key interacting residues are colored according to their atom type. (B) Recombinant DAPK–GST catalytic domain or recombinant GST proteins, immobilized on glutathione–agarose beads, were incubated with equal amounts of recombinant HSF1–His in binding buffer. Western blotting indicated that the proteins precipitated with glutathione–agarose beads (‘B’) and 10% of unbound target protein from the supernatant (SN) fraction using specific antibody. After probing with the first antibody, the membrane was stripped and reblotted with the next antibody. The asterisk indicates a nonspecific band for the GST antibody. M, biotinylated molecular mass marker. (C) HCT116 cells were transiently transfected with DAPK–HA (2 µg) construct. The transfection medium was changed after 6 h, and the cells were supplied with fresh TNF-containing medium. Cells were harvested at the indicated time-points after TNF stimulus. DAPK was immunoprecipitated (IP) using a specific antibody, and the bound products were immunoblotted with the indicated antibodies. IgG beads were incubated with DAPK-specific antibody and used as a negative control. Input controls (5% of the extract used for immunoprecipitation) are shown in the lower panel. Blots from representative experiments are shown (n = 3). Ctrl, control. (D) HCT116 cells were treated with 0.66 ng/ml TNF for 24 h and fixed with 3.7% paraformaldehyde. The subcellular location of endogenous pHSF1^Ser230 (red) and DAPK (green) was analyzed by confocal microscopy. DAPI staining (blue) was used to identify nuclei. Arrows indicate fragmented chromatin. Scale bar: 12 µm. n = 3 for all experiments.

DAPK phosphorylates HSF1 in vivo, in vitro and in silico. (A) HCT116 cells were transfected with wild-type (WT) DAPK, DAPK K42A dead mutant, S308A and S308D mutants. Empty vector (EV) was used as a negative control. Immunoprecipitation (IP) was performed using a DAPK-specific antibody. Magnetic beads were washed and incubated with recombinant HSF1–His (1 µg) in the presence of 7.5 µCi [γ-³²P]ATP and 200 µM ATP in DAPK kinase buffer for 30 min at 30°C. The samples were subjected to separation by SDS-PAGE, followed by autoradiography. (B) Protein expression level of wild-type DAPK and K42A, S308A and S308D mutants and empty vector in transfected HCT116 cells was assessed by western blotting using a DAPK-specific antibody. GAPDH was used as a loading control. (C) Recombinant HSF1–His (1 µg) was incubated with a GST-tagged catalytic domain of DAPK (0.5 µg) and ATP (200 µM) in DAPK kinase buffer, in the absence or presence of DAPK inhibitor (DI; 10 and 30 µM) at 30°C for 30 min. The products of the in vitro phosphorylation reaction were divided into two parts. Half was resolved by SDS-PAGE and analyzed by western blotting, and the other half was used as a loading control. One membrane was used for all indicated antibodies. The lower panel represents Coomassie Brilliant Blue R-250 staining and indicates equal protein loading. The first lane contains a prestained protein marker (PeqGold V, Peqlab); the second lane contains a biotinylated marker (‘M’). Input indicates the amount of protein used in the in vitro phosphorylation reaction. Incubation of recombinant GST and HSF1–His in the presence of cold ATP was used as a negative control. All experiments were repeated at least three times, and representative data are shown. (D) Interface of the HSF1–DAPK complex with the DAPK inhibitor. HSF1 and DAPK are shown in green and pink, respectively. DAPK inhibitor is shown in blue. Residues within 4 Å of the inhibitor are shown in stick representation. (E) Conformational changes in the DAPK catalytic domain upon inhibitor binding. Superposed structures of the DAPK catalytic domain available in PDB ID: 1JKS (blue), inhibitor-bound conformation of DAPK (green) and catalytic domain with inhibitor bound to HSF1 (pink) are superposed. DAPK inhibitor is shown in stick representation. Regions in DAPK showing conformational rearrangements are highlighted (red circles).

HSF1 is triggering DAPK-dependent apoptosis by TNF. (A) Scatter plots of flow cytometric analysis of annexin-V and propidium iodide staining of untreated HCT116 cells (ctrl), HCT116 cells subjected to 0.66 ng/ml TNF treatment (TNF), HCT116 cells transfected with either HSF1–HA cDNA (HSF1) or with HSF1 S230A or S230E mutant plasmids. pHACE-HA cDNA was used for HCT116 cell transfection as an empty vector (EV). Also analyzed were HCT116 cells transfected with either HSF1 siRNA (siHSF1) or scrambled siRNA, used as negative control (siNC). TNF treatment was performed at 6 h after transfection. The apoptosis rate was assessed as the sum of the lower right quadrant (early apoptosis) and upper right quadrant (late apoptosis). (B) Quantification of the mean annexin-V and propidium iodide fluorescence of the cell population (n = 3; ±s.e.m.). (C) HCT116 cells were transfected with 2 µg of wild-type HSF1–HA or S230A and S230E mutants. Empty vector was used as a negative control. Alternatively, 100 nM HSF1 siRNA or scrambled siRNA was transfected. At 6 h after transfection, fresh TNF-containing medium (or medium without TNF) was added. At 24 h after TNF incubation, cells were harvested, lysates were resolved by SDS-PAGE and probed with the specific indicated antibodies. GAPDH was used as a loading control. Blots from representative experiments are shown (n = 3).

The regulatory domain of HSF1 is involved in the pro-apoptotic response to TNF. (A) Upper panel, functional domains and potential DAPK phosphorylation motifs in human HSF1. DBD, DNA-binding domain; HR-A/B, harbour domain/leucine zipper trimerization domain; RD, regulatory domain; C, heptad repeat region C; AD, activation domain. Blue arrows show transcription-inhibitory phosphorylation; red arrows show transcription-inducing phosphorylation; white arrows indicate residues for which phosphorylation has unknown effects on gene transcription. Lower panel, schematic representation of the HSF1 cluster deletion mutant (CDM), containing partially excluded sequence of the regulatory domain (amino acids 250–370). WT, wild type. (B) HCT116 cells were transfected with 2 µg of HSF1-CDM–HA construct. Equal amounts of protein were resolved by SDS-PAGE and probed with the HSF1-specific antibody. The asterisk indicates endogenous HSF1. GAPDH was used as a loading control. Ctrl, control. (C) HCT116 cells were co-transfected with 1 µg of HSF1-CDM–HA and DAPK–FLAG constructs and treated with 0.66 ng/ml of TNF for 12 h. The subcellular localization of exogenous pHSF1^Ser230 (red) and DAPK (green) was visualized by confocal microscopy. DAPI staining (blue) was used to identify nuclei. Arrows indicate fragmented chromatin. Scale bar: 12 µm. n = 2 for all experiments. (D) Scatter plots of flow cytometric analysis of annexin-V and propidium iodide staining of untreated HCT116 cells (Ctrl), HCT116 cells subjected to TNF treatment (TNF) and HCT116 cells transfected with either HSF1-CDM–HA cDNA (CDM) or empty vector (EV). TNF treatment was performed at the same time for all samples, 6 h after cDNA transfection, for 24 h. The apoptosis rate is determined as the sum of the lower right quadrant (early apoptosis) and upper right quadrant (late apoptosis). (E) Quantification of the mean annexin-V and propidium iodide fluorescence of the cell population (n = 3; ±s.e.m.).

pHSF1^Ser230 binds to the DAPK promoter. (A) Regulatory elements in the proximal human DAPK promoter. The numbering of nucleotides is relative to the transcriptional start site (+1). (B) HEK293 cells were transiently transfected with 2 µg of dual Gaussia Luciferase (GLuc)/secreted alkaline phosphatase (SeAP) constructs: GAPDH-PG04 (positive control), NEG-PG04 (negative control), DAPK promoter-PG04, HSF1–HA and empty vector (EV, pHACE, 2 µg) constructs. All luciferase assay results are expressed as relative light units (RLU) and presented as a ratio of the mean GLuc reporter to secreted alkaline phosphatase luminescence signals. The RLU values obtained from co-transfected cell samples were normalized to those co-transfected with the DAPK promoter-PG04 reporter and the pHACE empty vector. Experiments were performed in triplicate and repeated three times; representative experiments are depicted. Data show the mean±s.d.; *P<0.05; **P<0.01 (unpaired two-tailed t-test). (C) Upper panel, ChIP assay using anti-pHSF1^Ser230 was performed on HCT116 cell lysates. The DAPK promoter was identified by PCR amplification of the DNA fragments precipitated with anti-pHSF1^Ser230 antibody. Input represents one-tenth of cleared supernatant. IgG immunoprecipitates (IP) were used as negative controls. Lower panel, immunoprecipitated DNA and input DNA were quantified by real-time PCR with primers specific for the HSE in the DAPK promoter (F1/R1). Data are represented as the mean±s.d.; *P<0.05 (unpaired two-tailed t-test). Ctrl, control. (D) Schematic representation of the primers designed for chromatin immunoprecipitation. The HSE in the DAPK promoter is illustrated, and the HSE region was amplified by using the F1/R1 primers. Negative control regions without HSE motifs are amplified using the upstream F2/R2 or downstream F3/R3 primer pairs. The boxed white region indicates the primary transcript, and the black region indicates the ATG codon as the start of translation. (E) Immunoprecipitated and input DNA were quantified by real-time PCR with HSE-specific primers (F1/R1) and HSE-nonspecific primers (F2/R2 and F3/R3) in the DAPK promoter after 48 h of TNF treatment. Data are represented as the mean±s.d.; *P<0.05 (unpaired two-tailed Student's t-test). All experiments were repeated at least three times.

Pattern of DAPK and pHSF1^Ser230 expression in human colorectal tumor tissue as determined by immunohistochemistry on formalin-fixed paraffin-embedded tissue samples. Tumor 1. (A) Hematoxylin and eosin (HE)-staining of colorectal cancer showed a cribriform pattern and mucous secretion. (B) Immunohistochemical analysis using the DAPK-specific antibody indicated moderate to strong cytoplasmic expression of DAPK in the tumor cells. (C) Immunohistochemical analysis using the pHSF1^Ser230-specific antibody showed moderate expression of pHSF1^Ser230 in both the cytoplasm and the nuclei of tumor cells. Tumor 2. (D) HE staining of rectal cancer showed a predominant tubular pattern. (E) Immunohistochemical analysis using the DAPK-specific antibody indicated complete absence of cytoplasmic DAPK. (F) Immunohistochemical analysis using the pHSF1^Ser230-specific antibody demonstrated lack of pHSF1^Ser230 in the tumor cells. Tumor 3. (G) HE staining of colorectal cancer showed poorly formed fused glands and extensive necrosis. (H) Strong cytoplasmic expression of DAPK is seen in this sample. (I) The same tumor lacked pHSF1^Ser230 expression (except for nonspecific staining within secretion). Tumor 4. (J) HE staining of moderately differentiated colorectal carcinoma showed dilated glands and necrosis. (K) DAPK was not expressed in the cytoplasm of this tumor (faint non-specific nuclear staining). (L) The same sample showed pHSF1^Ser230 expression within the apical cytoplasm of tumor cells. (M) This conventional sample showed a gradual increase in DAPK expression from the tumor center (upper left, ‘C’) to the invasion front (lower right, arrows). (N) The same sample showed a similar pattern for pHSF1^Ser230 (as for DAPK in M). Original magnification for all photomicrographs is ×200.