Subcellular localization of Daxx determines its ... - Wiley Online Library

FEBS Letters 581 (2007) 843–852

Subcellular localization of Daxx determines its opposing functions in ischemic cell death Yong-Sam Junga, Hye-Young Kima, Yong J. Leeb, Eunhee Kima,* a

Department of Bioscience and Biotechnology, Chungnam National University, 220 Gung-dong, Yuseong-gu, Daejeon 305-764, Republic of Korea b Department of Surgery and Pharmacology, University of Pittsburgh, PA 15213, USA Received 1 November 2006; revised 3 January 2007; accepted 22 January 2007 Available online 2 February 2007 Edited by Varda Rotter

Abstract This study examined the role of Daxx in ischemic stress. Upon ischemic stress, nuclear export of Daxx to the cytoplasm was observed in primary myocytes as well as in various cell lines. Daxx silencing using siRNAs was detrimental in tethering PML-nuclear body (PML-NB) constituents together. Overexpression of Daxx (W621A) caused nuclear export of p53 independently of PML and promoted ischemic cell death via activation of JNK. Conversely, overexpression of Daxx (S667A) prevented dissociation of PML-NB constituents and protected cells from ischemic death. Collectively, our results demonstrate that the subcellular localization of Daxx determines its role in ischemic cell death. 2007 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. Keywords: Daxx; Cardiac myocyte; Ischemia; JNK; PML; Nuclear body; Cell death

1. Introduction The death-associated protein, Daxx, is mainly known for its death receptor adaptor role at the cell surface [1,2]. Daxx binds to Fas and potentiates Fas-mediated apoptosis in the cytoplasm via the apoptosis signal-regulating kinase 1 (ASK1) fi c-Jun N-terminal kinase (JNK) pathway, independently of the FADD/caspase-8 pathway [3–7]. Daxx also binds to the transforming growth factor beta (TGFb) receptor to mediate JNK activation [8]. According to immunohistochemical studies, however, Daxx is mainly observed in the nucleus [1,9], where its nuclear-speckle distribution coincides with that of promyelocytic leukemia protein-nuclear bodies (PML-NBs) [10,11]. Recently, Daxx trafficking between subcellular compartments has been reported [12], which may explain the apparently contradictory data regarding the subcellular localization of Daxx. Daxx shuttles between the nucleus and cytoplasm. The nuclear export of Daxx is induced by various stimuli such as glu*

Corresponding author. Fax: +82 42 821 7542. E-mail address: [email protected] (E. Kim). Abbreviations: ASK1, apoptosis signal-regulating kinase 1; CH, chemical hypoxia; JNK, c-Jun N-terminal kinase; LMB, leptomycin B; siRNA, small interference RNA; PML-NB, promyelocytic leukemia protein-nuclear body; RASSF1C, Ras-associated domain family 1C protein; scRNA, control scrambled siRNA

cose deprivation, oxidative stress, and Fas stimulation [7,13– 15]. Daxx nuclear export depends on the phosphorylation status of Ser-667 by the homeodomain-interacting protein kinase 1 (HIPK1) [15]. Conversely, interferon a/b stimulates Daxx import in a tyrosine kinase 2 (Tyk2)-regulated manner [16,17]. Nuclear Daxx can also be redistributed within subnuclear compartments. Daxx is present within PML-NBs, but can also leave PML-NBs and move through the nucleoplasm [18,19]. Time-dependent shuttling of Daxx between centromeres and PML-NBs during cell-cycle progression has been reported [9,20–22]. Although the functional significance of these observations remains to be determined, Daxx might have different roles depending on its subcellular localization. In the present study, we examined the role of Daxx in ischemic stress. To further understand the controversial roles of Daxx, we used Daxx mutants that were confined to specific subcellular locations. We found that most Daxx is localized to the nuclei of resting cells, where it co-localizes with PML and p53. In response to ischemic stress, Daxx is exported to the cytoplasm and ischemic cell death is mediated via JNK activation. Our data indicate that Daxx is a scaffolding protein and, is essential for the structural maintenance of PML-NBs. Finally, our findings provide a possible solution to the controversy surrounding the function of Daxx.

2. Materials and methods 2.1. Cell culture and transient transfection Rat embryonic heart-derived H9c2, rat skeletal muscle L6 and human cervical carcinoma HeLa cells were maintained in Dulbecco’s modified Eagle’s medium (DMEM) with 10% heat-inactivated fetal bovine serum (FBS) and antibiotics (Gibco BRL). H9c2 cells were transfected using the Nucleofector system (Amaxa). 1 · 106 cells were suspended in 100 ll Nucleofector Solution Kit R (Amaxa) containing 2 lg of plasmid in a 2 mm electroporation cuvette. H9c2 cells were transfected with the electrical setting T-20. Transfection into HeLa was performed using Metafectene (Biontex Laboratories) according to the manufacturer’s protocol. Primary rat neonatal cardiac myocyte cultures were prepared from 1 to 2 day-old rats as described previously [23]. 2.2. Experimental simulations of ischemia In the case of chemical hypoxia (CH), cells were washed once with phosphate-buffered saline (PBS) and placed in the metabolic inhibition buffer (106 mM NaCl, 4.4 mM KCl, 1 mM MgCl2 Æ 6H2O, 38 mM NaHCO3, 2.5 mM CaCl2, 20 mM 2-deoxy-D -glucose, 1 mM NaCN, pH 6.6) for indicated time periods [24]. In the case of using an anaerobic chamber, cultures were incubated in serum-free, glucose-free DMEM using an anaerobic chamber (Forma Scientific) at 37 C with 5% CO2, 10% H2 and 85% N2.

0014-5793/$32.00 2007 Federation of European Biochemical Societies. Published by Elsevier B.V. All rights reserved. doi:10.1016/j.febslet.2007.01.055

844 2.3. Small interference RNA 21-Nucleotide RNA oligonucleotides were synthesized from Qiagen Inc. (Qiagen). The small interference RNA (siRNA) sequence for rat Daxx (si-rDaxx) was chosen by referring to the previously reported human Daxx siRNA sequence [25,26]. The siRNAs used in this study were si-rDaxx, 5 0 -GGAGUUGGAUCUGUCAGAAdTdT-3 0 (sense), 5 0 -UUCUGACAGAUCCAACUCCdTdT-3 0 (antisense); si-rPML, 5 0 -CGUGUUCUUCGAGAGCCUGdTdT-3 0 (sense), 5 0 -CAGGCUCUCGAAGAACACGdTdT-3 0 (antisense); si-rp53, 5 0 -GCAUGAACCGCCGGCCCAUdTdT-3 0 (sense), 5 0 -AUGGGCCGGCGGUUCAUGCdTdT (antisense); and scrambled RNA (scRNA), 5 0 -GCGCGCUUUGUAGGAUUCGdTdT-3 0 (sense), 5 0 -CGAAUCCUACAAAGCGCGCdTdT-3 0 (antisense). The scRNA sequence represents a 21-nucleotide RNA oligonucleotide with no significant homology to any mammalian gene sequence and therefore serves as a non-silencing control [27]. The sense and antisense siRNAs were annealed according to manufacturer’s instructions. 2.4. cDNA constructs and mutagenesis Daxx was subcloned into pFLAG-CMV-2. Daxx substitution mutants, Daxx (W621A) and Daxx (S667A), were constructed with the QuikChange site-directed mutagenesis kit as previously described [15].

Y.-S. Jung et al. / FEBS Letters 581 (2007) 843–852 2.5. Immunofluorescence microscopy Cells grown on glass cover slips were fixed in 4% paraformaldehyde/ PBS (pH 7.3) for 15 min at room temperature, permeabilized in 0.5% Triton X-100 in PBS for 15 min at room temperature. The fixed cells were processed for immunofluorescence. Polyclonal rabbit anti-Daxx, goat anti-Daxx, and rabbit anti-p53 antibodies were purchased from Santa Cruz biotechnology, and anti-SUMO1 monoclonal antibody was purchased from Zymed Laboratories. Anti-PML monoclonal antibody 5E10 was a generous gift from Dr. R. van Driel (University of Amsterdam, Amsterdam, Netherlands) [28]. Transfected H9c2 and HeLa cells were plated on cover slips in culture medium for 12 h, cells were incubated for the indicated times in metabolic inhibition buffer. Cover slips were examined on a Zeiss LSM510 (Carl Zeiss) or a BioRad Radiance 2000 laser scanning confocal microscope (BioRad) using 40· or 60· objective. Bars represent 10 lm in all micrographs. 2.6. DAPI DNA staining assay Slightly modified DAPI staining was performed as described previously [29]. Cells were grown on cover slips in six well culture plates. The fixed cells were permeabilized with 0.5% Triton X-100 in PBS for 15 min, followed by staining with 2 lg/ml of DAPI. Then the cells were observed and counted under fluorescence microscope.

Fig. 1. Daxx nuclear export in response to ischemic injury. (a, b) Cultures of H9c2 cells were subjected to CH for the indicated times (0, 0.5, 1, and 2 h). The confocal microscopic images (a) and Western blot analysis (b) of treated cells are shown. (c) H9c2 cells were subjected to anaerobic conditions for the indicated times (0, 1, 2, and 4 h). Cells were then incubated with a rabbit polyclonal Daxx antibody. (a–c) Bound primary antibody was labeled with a fluorescein-5-isothiocyanate (FITC)-conjugated anti-rabbit secondary antibody, and nuclei were visualized using propidium iodide (PI). WB, Western blot.

Y.-S. Jung et al. / FEBS Letters 581 (2007) 843–852

845

Fig. 2. Trafficking of PML-NB constituents in response to ischemic stress. Immunofluorescence microscopy was performed in H9c2 cells using mouse anti-PML (5E10) (a), rabbit anti-p53 (b) and mouse anti-SUMO1 (c), together with the anti-Daxx antibody. a–c (first columns): Red represents PML (a), p53 (b), and SUMO1 (c). a–c (second columns): Green represents Daxx. a–c (third columns): The red and green images were merged. (d) Rat L6 myoblast cells, before and after CH, were immunostained with anti-PML (5E10) and anti-Daxx antibodies. a–d (last columns): In double-label colocalization experiment, control immunofluorescence image is shown to exclude the possible artifact by double-label experiment. (e) Primary rat neonatal cardiac myocytes were prepared as described in Section 2. Cells were then fixed and assessed for the immunofluorescence of Daxx (green) and PML (red). (f) H9c2 cells were subjected to CH treatment for 1.5 h. Whole-cell lysates were then subjected to immunoprecipitation using anti-PML monoclonal antibody 5E10. The blot was probed with anti-p53 monoclonal antibody (DO-1). The same membrane was stripped and reprobed repeatedly with rabbit anti-Daxx and anti-PML (5E10) to detect Daxx and PML, respectively. IgG, immunoglobulin G; IP, immunoprecipitation; WCL, whole-cell lysate.

846


Fig. 2 (continued)

2.7. FACS analysis Dead cells were determined by DNA content analysis using PI staining. For transient transfection studies, cells were cotransfected with the green fluorescent protein (GFP) expression vector pEGFP-C2, and the profiles were determined for the GFP-gated cells. The transfected H9c2 cells were trypsinized, resuspended in 70% cold ethanol, fixed overnight at 20 C, washed twice with PBS, subjected to RNase (QIAGEN) digestion (1 mg/ml), and stained with PI (Sigma) for 30 min at room temperature. The stained cells were analyzed by EPICS XL flow cytometer (Beckmann Coulter) by gating on an area versus width dot plot to exclude debris and aggregates. 2.8. JNK assay The level of JNK phosphorylation was measured after CH with or without SP600125 treatment (an inhibitor of JNK, TOCRIS). In vitro kinase assays were performed as previously reported [30]. 2.9. Immunoprecipitation and Western analysis To test for interactions among PML, Daxx and p53 in H9c2 cells, cells were lysed in mammalian lysis buffer and immunoprecipitated with anti-PML monoclonal antibody 5E10. Immunoprecipitates were washed and separated by SDS–PAGE. The membrane was probed with anti-PML monoclonal antibody 5E10, rabbit anti-Daxx antibody, and anti-p53 monoclonal antibody (DO-1) followed by HRP-conjugated secondary antibodies (Santa Cruz). Proteins in the membranes were then visualized using the enhanced ECL reagent (Amersham Biosciences). 2.10. Reporter assay H9c2 cells were transfected with 2 lg of si-rDaxx, 1 lg of pFLAGCMV-2-p53 plasmid, 0.2 lg of CMV-b-galactosidase plasmid, and 0.2 lg of pG13-luciferase plasmid as indicated. pG13-luciferase plasmid [31] was a generous gift from Dr. Hong-duk Youn (Seoul National University, Korea). Transfections were carried out using Nucleofector Solution Kit R (Amaxa) according to the manufacturer’s instructions. Lysates were prepared and assayed for luciferase activities according to the manufacturer’s instructions (Promega). Luciferase values were normalized for transfection efficiency by measuring b-galactosidase activity.

3. Results 3.1. Ischemia induces nuclear export of Daxx To simulate ischemic injury, we used 2-deoxy-D -glucose and sodium cyanide to induce chemical hypoxia (CH) [24]. In resting H9c2 cells, Daxx was visible as distinct nuclear speckles (Fig. 1a). However, after 2 h of CH, Daxx was detected mainly in the cytoplasm of H9c2 cells (Fig. 1a). Daxx expression was increased in a time-dependent manner after CH (Fig. 1b). Also to simulate induction, H9c2 cells were incubated in serum-free, glucose-free DMEM in an anaerobic chamber [32]. Translocation of Daxx into the cytoplasm was also observed under these conditions (Fig. 1c). After 4 h in an anaerobic chamber, Daxx was detected mainly in the cytoplasm. Therefore, two different experimental models show that, in response to ischemic injury, Daxx is exported into the cytoplasm of H9c2 cells. Export of Daxx from the nucleus to the cytoplasm was also observed in L6 cells (Fig. 2d), HeLa cells (Supplemental Figs. 1 and 2), Chinese hamster lung fibroblast PS120 cells (data not shown) and neonatal primary rat cardiac myocytes (Fig. 2e). 3.2. Ischemic stress induces nuclear export of some of the PML-NB proteins Of the known nuclear bodies [33], PML-NBs are the best characterized [34,35]. Daxx has been reported to be a component of PML-NBs; hence we characterized the effects of ischemic stress on subcellular trafficking of the PML-NB constituents. In resting cells, PML was detected in the nuclei of H9c2 cells and shown to co-localize with Daxx (Fig. 2a). p53 and SUMO1 were also co-localized with Daxx in H9c2 cells (Fig. 2b, c). Under CH conditions, PML remains in the nucleus and has a speckle distribution, whereas Daxx, SUMO1, and p53 are exported to the cytoplasm. In order to


investigate the changes in Daxx subcellular localization during cell differentiation, neonatal primary rat cardiac myocytes and two myoblast cell lines from embryonal heart (H9c2 cells) and skeletal muscle (L6 cells) were selected in this study. In response to ischemic stress, Daxx was exported to the cytoplasm but PML was not in rat L6 myoblasts (Fig. 2d) and neonatal primary rat cardiac myocytes (Fig. 2e). There were no significant differences in Daxx–PML association among cells of different developmental stages, for example, L6 (myoblast), H9c2 (cardiac myocyte), and neonatal primary rat cardiac myocytes. In HeLa and L6 cells, not all Daxx was exported from the nucleus in response to CH, demonstrating that the extent of Daxx nuclear export differs with cell type (Fig. 2d and Supplemental Fig. 2). To further validate our confocal observations, we performed immunoprecipitation with the anti-PML monoclonal antibody 5E10 using lysates of H9c2 cells (Fig. 2f). Daxx and p53 were detected in the precipitates of untreated cells but not in those of CH-treated cells. These immunoprecipitation data correlate well with the confocal microscopy data regarding the redistribution of PML-NB constituents. 3.3. The nuclear presence of Daxx is essential for tethering PML-NB constituents Daxx can interact with SUMO1, p53, and PML [36–39]; therefore we questioned whether Daxx influences the subcellular redistribution of SUMO1, p53, and PML. An RNA inter-

847

ference approach was used to address this issue [40]. H9c2 cells were transfected with rat Daxx siRNA (si-rDaxx), rat PML siRNA (si-rPML), or rat p53 siRNA (si-rp53), as described in Section 2. Our results indicated that siRNA efficiently reduced expression of target proteins in H9c2 cells after 72 h, whereas control scrambled siRNA (scRNA) did not (Fig. 3a). Therefore, all depletion experiments were performed 72 h after transfection. As expected, PML-NB components were visible as distinct nuclear speckles in scRNA transfected cells (Fig. 3b, scRNA). Silencing of Daxx and PML caused the release of p53 and SUMO1 from PML-NBs (Fig. 3b, middle two columns). SUMO1 was exported from PML-NBs and a faint dispersed pattern was visible throughout the nucleus and in the nuclear periphery, but not in the cytoplasm. Consistent with previous observations [10,18,25,38], depletion of PML did not affect the Daxx distribution pattern, and depletion of Daxx did not affect the PML nuclear structure (Fig. 3b and Supplemental Fig. 3). These observations indicate that the nuclear presence of Daxx is required for tethering PML-NB speckle constituents together. Our data show that Daxx persists as intact nuclear speckles independently of PML and p53. However, depletion of nuclear Daxx is as detrimental as depletion of PML in assembling PML-NB constituents and causes dissociation of PML-NB constituents. Collectively, our data suggest that, like PML, Daxx might be a PML-NB scaffold protein.

Fig. 3. The nuclear presence of Daxx determines the integrity of the PML-NBs. (a) H9c2 cells were electroporated with scrambled RNA (scRNA), sirDaxx, si-rPML or si-rp53 and cell lysates were prepared after 24, 48 and 72 h. Western-blot analysis of whole-cell lysates shows target proteins were down-regulated at the indicated time points after introduction of siRNA. The relative target protein level after normalization with b-tubulin is shown at the bottom of the panel. (b) Transfected cells were immunostained with anti-PML (5E10), rabbit anti-Daxx, rabbit anti-p53, or anti-SUMO1 monoclonal antibodies. Representative immunofluorescence images were obtained 72 h after transfection.

848

3.4. Subcellular localization of Daxx regulates cellular sensitivity to CH The physiological relevance of the subcellular localization of Daxx was studied using leptomycin B (LMB) and Daxx mutants that localize to different regions of the cell. LMB has


been shown to directly bind to CRM1 and disrupt the interaction between CRM1 and the nuclear export sequence of Daxx [15,41]. In H9c2 cells, Daxx (W621A) and Daxx (S667A) mutants localize to the cytoplasm and nuclei, respectively (Fig. 4a), and CH conditions did not alter the subcellular local-

Fig. 4. Cytoplasmic Daxx mutants promote CH-induced cell death. (a) H9c2 cells were transfected with pFLAG–CMV-2, pFLAG–CMV-2/Daxx (W621A), or pFLAG–CMV-2/Daxx (S667A). Transfected cells were stained with the rabbit anti-Daxx antibody (upper row) and a mouse FLAG antibody (middle and bottom rows). Bound primary antibodies were labeled with an FITC-conjugated anti-rabbit secondary antibody (upper row) or an FITC-conjugated anti-mouse secondary antibody (middle and bottom rows). Cells were treated with LMB (20 nM) for 4 h before being subjected to CH for 1 h. Nuclei were visualized with PI. (b) Transfection and LMB treatment were performed as described in (a) but cells were subjected to CH for 2 h and the pEGFP-C2 vector was used as a co-transfection marker. Dead cells were counted using DAPI staining. The percentage of GFPpositive cells with condensed or fragmented nuclei was determined by counting at least 300 cells from three independent experiments in H9c2 (left panel) and HeLa cells (right panel). (c) Transfection and CH were performed as described in (b). The percentage of dead cells was determined by FACS analysis. The mean percentage of cell death was shown as means ± S.E.M. from three independent experiments. MOCK, vector-transfected cells; WB, Western blot.


izations of these mutants. However, under CH conditions, LMB treatment prevented nuclear export of the transfected Daxx mutants as well as endogenous Daxx, but did not to prevent dispersion of Daxx to the nucleoplasm. The effects of Daxx subcellular localization on CH-induced cell death were assessed using the DAPI assay. CH-induced cell death was also suppressed by LMB treatment and by transfection with either wild-type (WT) Daxx or the Daxx (S667A) mutant, but was increased after transfection of H9c2 cells with the Daxx (W621A) (Fig. 4b, left panel). The positive correlation between CH-induced cell death and the presence of cytoplasmic Daxx was further confirmed in HeLa cells (Fig. 4b, right panel). H9c2 cells were transfected with Daxx mutants and flow cytometry was used to determine the percentage of dead GFP-gated cells (Fig. 4c). Similar protection against and sensitization to CH were observed in cells transfected with Daxx (S667A) and with Daxx (W621A), respectively. 3.5. Cytoplasmic Daxx mediates CH-induced cell death through JNK activation CH induces JNK activation and cell death [30]. Daxx can mediate JNK activation in response to stresses such as UV light, TGFb and TNFa [8,42,43]. To determine whether subcellular localization of Daxx affects JNK activation in response to CH, we analyzed JNK activation in H9c2 cells. H9c2 cells were transfected with Daxx mutants, as shown in Fig. 5. In the absence of CH, transfection of Daxx (WT) and Daxx (S667A) did not activate JNK (Fig. 5a). Transfection of Daxx (WT) and Daxx (S667A) suppressed JNK activation in response to CH when compared with MOCK transfection of vector only. However, transfection with Daxx (W621A) was

849

found to activate JNK in the presence or absence of CH, indicating that cytoplasmic Daxx enhances JNK activation. Next, we examined whether, under CH conditions, inhibition of JNK signaling affects the subcellular localization of Daxx. Under CH conditions, nuclear export of endogenous Daxx was blocked by incubation of H9c2 cells with 10 lM SP600125, which is an inhibitor of JNK (Supplemental Fig. 4, top row). Therefore, JNK activation seems to be required for nuclear export of Daxx. Previously, we have shown that ASK1–SEK1–JNK1–HIPK1 signaling pathway is required for Daxx nuclear export [15]. By contrast, the subcellular localizations of Daxx (W621A) and Daxx (S667A) were not affected by SP600125 treatment (Supplemental Fig. 4, middle and bottom rows). SP600125 did not prevent the nuclear export of Daxx (W621A), indicating that Daxx (W621A) was already in the correct conformation for nuclear export [15]. The finding that SP600125 administration did not induce cell death in Daxx (W621A)-transfected cells indicates that the presence of Daxx in the cytoplasm is not sufficient to cause cell death. By contrast, inhibition of JNK activation prevented ischemic cell death regardless of the subcellular localizations of Daxx in H9c2 cells (Fig. 5b). Together, these data indicate that JNK plays an essential role in ischemic cell death via cytoplasmic Daxx, and it is likely that cytoplasmic Daxx activates JNK by ASK1 oligomerization [14,15]. 3.6. Daxx depletion induces JNK activation and p53 activation We next analyzed the silencing effect of Daxx. JNK activation was enhanced in si-rDaxx-transfected cells regardless of CH (Fig. 6a). Recent studies have revealed that Ras-associated domain family 1C protein (RASSF1C) activates JNK in the

Fig. 5. JNK is required for ischemic cell death. (a) H9c2 cells were transfected with 2 lg plasmid DNA cloned into pFLAG–CMV-2 as indicated. Forty-eight hours after transfection, H9c2 cells were subjected to CH for 1 h, and cell lysates were immunoprecipitated with JNK antibody and assayed for kinase activity in vitro. Transfected cells were incubated with SP600125 (10 lM) for 30 min before and during CH. The experiments were repeated at least three times and representative images are shown. Top row, phosphorylation of GST-c-Jun; middle row, WCLs were immunoprecipitated and immunoblotted with JNK antibody; bottom row, WCLs were immunoblotted with FLAG antibody. (b) H9c2 and HeLa cells were cotransfected with pEGFP-C2 and FLAG-conjugated Daxx constructs. Dead cells were counted by DAPI staining assay as described above. Assessment of cell death was determined from three independent experiments in H9c2 (left panel) and HeLa cells (right panel).

850


Daxx affects p53-mediated transcription. As shown in Fig. 6b, cells transfected with si-rDaxx enhanced p53-mediated transcription. And Daxx depletion increased cell death significantly; cell deaths were 23.7 ± 1.3% for scRNA and 36.3 ± 2.1% for si-rDaxx, respectively, upon induction of CH when determined by DAPI analysis (Fig. 6c). p53 depletion and PML depletion, respectively, protected cells from CH-induced cell death. Our data were consistent with previous studies showing the protective effect of p53 depletion [46–48]. Collectively, these data might suggest that Daxx depletion sensitizes H9c2 cells to CH via JNK activation and p53 activation.

4. Discussion

Fig. 6. Daxx depletion activates JNK and p53-mediated transcriptional activity. (a) H9c2 cells were transfected with 2 lg of each RNA oligonucleotide as indicated. Seventy-two hours after transfection, CH was applied to H9c2 cells for 1 h and an in vitro kinase assay was performed. (b) The pG13-luc was transfected with different combinations as indicated into H9c2 cells. Luciferase assays were performed 72 h after transfection. (c) H9c2 cells were transiently cotransfected with pEGFP-C2 and RNA oligonucleotides (scRNA, si-rDaxx, si-rp53 and si-rPML). After 72 h of transfection, 2 h of CH was applied or not and the percentages of cell death were calculated using DAPI staining. Ctrl, untransfected cells.

cytoplasm [44]. Since nuclear Daxx sequesters RASSF1C to PML-NB and inhibits JNK activation, Daxx depletion would release RASSF1C to the cytoplasm. Therefore, JNK activation upon Daxx depletion under our experimental condition is not an unexpected consequence. Previous studies have shown that Daxx acts as a negative regulator of p53 [39,45]. Therefore, we examined whether

We have studied Daxx function in ischemic stress and, through the use of localization-specific Daxx mutants, shown that Daxx has multiple functions depending on its subcellular distribution. Nuclear Daxx in PML-NBs is required for tethering the constituents of PML-NBs, as well as protecting H9c2 cells from ischemic death. Cytoplasmic Daxx promotes cell death through JNK activation. Collectively, this study shows that Daxx has different effects in CH-induced cell death according to its subcellular localization. Among the nuclear speckles reported to date, PML-NB is the best characterized and more than 30 proteins have been reported to localize to PML-NBs [12,49,50]. To date, PML has been the focus of studies on PML-NB structure, as it is an important scaffold component. However, the present study provides evidence that Daxx is also essential for PML-NB integrity. This requirement for Daxx was validated by Daxx depletion experiments, which showed that Daxx depletion induced dissociation of PML-NB constituents as effectively as PML depletion. Moreover, PML depletion failed to disrupt Daxx speckles, indicating that Daxx localization is independent of PML localization, which is contrary to the previous reports [38,51]. Based on the fact that both Daxx and PML are required for localization of p53 and SUMO1 to PML-NBs, it is possible that Daxx and PML cooperate in maintaining PML-NB integrity. Our data show that JNK activation plays an important role in ischemia-induced cell death in H9c2 cells. Considering that Daxx facilitates ASK1 oligomerization, the increase in JNK activation by cytoplasmic Daxx is not surprising. However, suppression of improper JNK activation in resting cells seems to be an equally important function of Daxx. RASSF1C is associated with Daxx in PML-NBs in resting cells. If Daxx disappears from PML-NBs, RASSF1C is released from the nucleus and activates JNK [44]. Therefore, sequestration of RASSF1C by Daxx to the PML-NBs would prevent erroneous JNK activation. It is interesting that nuclear Daxx suppresses JNK activation, whereas cytoplasmic Daxx promotes JNK activation. Blocking JNK activation prevented Daxx nuclear export. And blocking Daxx nuclear export inhibited JNK activation. Our results show the presence of reciprocal regulation between nuclear export of Daxx and JNK activation suggesting the presence of a positive feedback loop. In summary, the results of the present study demonstrate protection against ischemic cell death by nuclear Daxx. Nuclear Daxx seems to prevent improper activation of JNK


in resting cells. It is tempting to speculate that nuclear Daxx may sequester other potentially death-promoting proteins in addition to p53 and RASSF1C to PML-NBs. Identification of additional constituents of the PML-NBs would facilitate our understanding of these protective mechanisms. Considering the reduction in ischemia-induced cell mortality when Daxx export to the cytoplasm is blocked, antagonistic compounds that block Daxx trafficking could potentially serve as therapeutic agents for the treatment of ischemic diseases.

851

[14] [15] [16] [17]

Acknowledgements: We thank Dr. K.S. Kwon (Korea Research Institute of Bioscience and Biotechnology) for helpful comments and critical reading of the manuscript. This research was supported by a grant (CBM2-B211-001-1-0-0) from the Center for Biological Modulators of the 21st century Frontier R&D Program, the Ministry of Science and Technology, Korea. [18]

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.febslet.2007. 01.055.

[19]

[20]

References [1] Kiriakidou, M., Driscoll, D.A., Lopez-Guisa, J.M. and Strauss III, J.F. (1997) Cloning and expression of primate Daxx cDNAs and mapping of the human gene to chromosome 6p21.3 in the MHC region. DNA Cell Biol. 16, 1289–1298. [2] Yang, X., Khosravi-Far, R., Chang, H.Y. and Baltimore, D. (1997) Daxx, a novel Fas-binding protein that activates JNK and apoptosis. Cell 89, 1067–1076. [3] Chang, H.Y., Nishitoh, H., Yang, X., Ichijo, H. and Baltimore, D. (1998) Activation of apoptosis signal-regulating kinase 1 (ASK1) by the adapter protein Daxx. Science 281, 1860–1863. [4] Charette, S.J. and Landry, J. (2000) The interaction of HSP27 with Daxx identifies a potential regulatory role of HSP27 in Fasinduced apoptosis. Ann. NY Acad. Sci. 926, 126–131. [5] Junn, E., Taniguchi, H., Jeong, B.S., Zhao, X., Ichijo, H. and Mouradian, M.M. (2005) Interaction of DJ-1 with Daxx inhibits apoptosis signal-regulating kinase 1 activity and cell death. Proc. Natl. Acad. Sci. USA 102, 9691–9696. [6] Khelifi, A.F., D’Alcontres, M.S. and Salomoni, P. (2005) Daxx is required for stress-induced cell death and JNK activation. Cell Death Differ. 12, 724–733. [7] Ko, Y.G., Kang, Y.S., Park, H., Seol, W., Kim, J., Kim, T., Park, H.S., Choi, E.J. and Kim, S. (2001) Apoptosis signal-regulating kinase 1 controls the proapoptotic function of death-associated protein (Daxx) in the cytoplasm. J. Biol. Chem. 276, 39103–39106. [8] Perlman, R., Schiemann, W.P., Brooks, M.W., Lodish, H.F. and Weinberg, R.A. (2001) TGF-beta-induced apoptosis is mediated by the adapter protein Daxx that facilitates JNK activation. Nat. Cell Biol. 3, 708–714. [9] Pluta, A.F., Earnshaw, W.C. and Goldberg, I.G. (1998) Interphase-specific association of intrinsic centromere protein CENPC with HDaxx, a death domain-binding protein implicated in Fas-mediated cell death. J. Cell Sci. 111, 2029–2041. [10] Ishov, A.M., Sotnikov, A.G., Negorev, D., Vladimirova, O.V., Neff, N., Kamitani, T., Yeh, E.T., Strauss III, J.F. and Maul, G.G. (1999) PML is critical for ND10 formation and recruits the PML-interacting protein daxx to this nuclear structure when modified by SUMO-1. J. Cell Biol. 147, 221–234. [11] Torii, S., Egan, D.A., Evans, R.A. and Reed, J.C. (1999) Human Daxx regulates Fas-induced apoptosis from nuclear PML oncogenic domains (PODs). EMBO J. 18, 6037–6049. [12] Salomoni, P. and Khelifi, A.F. (2006) Daxx: death or survival protein? Trends Cell Biol. 16, 97–104. [13] Song, J.J. and Lee, Y.J. (2003) Catalase, but not MnSOD, inhibits glucose deprivation-activated ASK1–MEK–MAPK signal transduction pathway and prevents relocalization of Daxx, hydrogen

[21]

[22]

[23] [24]

[25] [26] [27]

[28]

[29] [30]

[31] [32] [33]

peroxide as a major second messenger of metabolic oxidative stress. J. Cell Biochem. 90, 304–314. Song, J.J. and Lee, Y.J. (2003) Role of the ASK1–SEK1–JNK1– HIPK1 signal in Daxx trafficking and ASK1 oligomerization. J. Biol. Chem. 278, 47245–47252. Song, J.J. and Lee, Y.J. (2004) Tryptophan 621 and serine 667 residues of Daxx regulate its nuclear export during glucose deprivation. J. Biol. Chem. 279, 30573–30578. Gongora, R., Stephan, R.P., Zhang, Z. and Cooper, M.D. (2001) An essential role for Daxx in the inhibition of B lymphopoiesis by type I interferons. Immunity 14, 727–737. Shimoda, K., Kamesaki, K., Numata, A., Aoki, K., Matsuda, T., Oritani, K., Tamiya, S., Kato, K., Takase, K., Imamura, R., Yamamoto, T., Miyamoto, T., Nagafuji, K., Gondo, H., Nagafuchi, S., Nakayama, K. and Harada, M. (2002) Cutting edge: tyk2 is required for the induction and nuclear translocation of Daxx which regulates IFN-alpha-induced suppression of B lymphocyte formation. J. Immunol. 169, 4707–4711. Li, H., Leo, C., Zhu, J., Wu, X., O’Neil, J., Park, E.J. and Chen, J.D. (2000) Sequestration and inhibition of Daxx-mediated transcriptional repression by PML. Mol. Cell. Biol. 20, 1784– 1796. Muratani, M., Gerlich, D., Janicki, S.M., Gebhard, M., Eils, R. and Spector, D.L. (2002) Metabolic-energy-dependent movement of PML bodies within the mammalian cell nucleus. Nat. Cell Biol. 4, 106–110. Dellaire, G., Eskiw, C.H., Dehghani, H., Ching, R.W. and BazettJones, D.P. (2006) Mitotic accumulations of PML protein contribute to the re-establishment of PML nuclear bodies in G1. J. Cell Sci. 119, 1034–1042. Everett, R.D., Earnshaw, W.C., Pluta, A.F., Sternsdorf, T., Ainsztein, A.M., Carmena, M., Ruchaud, S., Hsu, W.L. and Orr, A. (1999) A dynamic connection between centromeres and ND10 proteins. J. Cell Sci. 112, 3443–3454. Luciani, J.J., Depetris, D., Usson, Y., Metzler-Guillemain, C., Mignon-Ravix, C., Mitchell, M.J., Megarbane, A., Sarda, P., Sirma, H., Moncla, A., Feunteun, J. and Mattei, M.G. (2006) PML nuclear bodies are highly organised DNA–protein structures with a function in heterochromatin remodelling at the G2 phase. J. Cell Sci. 119, 2518–2531. Baty, C.J. and Sherry, B. (1993) Cytopathogenic effect in cardiac myocytes but not in cardiac fibroblasts is correlated with reovirusinduced acute myocarditis. J. Virol. 67, 6295–6298. Gottlieb, R.A., Gruol, D.L., Zhu, J.Y. and Engler, R.L. (1996) Preconditioning rabbit cardiomyocytes, role of pH, vacuolar proton ATPase, and apoptosis. J. Clin. Invest. 97, 2391– 2398. Chen, L.Y. and Chen, J.D. (2003) Daxx silencing sensitizes cells to multiple apoptotic pathways. Mol. Cell. Biol. 23, 7108–7121. Michaelson, J.S. and Leder, P. (2003) RNAi reveals antiapoptotic and transcriptionally repressive activities of DAXX. J. Cell Sci. 116, 345–352. Zhang, L., Fogg, D.K. and Waisman, D.M. (2004) RNA interference-mediated silencing of the S100A10 gene attenuates plasmin generation and invasiveness of Colo 222 colorectal cancer cells. J. Biol. Chem. 279, 2053–2062. Stuurman, N., Floore, A., Colen, A., de Jong, L. and van Driel, R. (1992) Stabilization of the nuclear matrix by disulfide bridges, identification of matrix polypeptides that form disulfides. Exp. Cell Res. 200, 285–294. Pollenz, R.S., Chen, T.L., Trivinos-Lagos, L. and Chisholm, R.L. (1992) The dictyostelium essential light chain is required for myosin function. Cell 69, 951–962. Jung, Y.S., Jung, Y.S., Kim, M.Y. and Kim, E. (2004) Identification of caspase-independent PKCe-JNK/p38 MAPK signaling module in response to metabolic inhibition in H9c2 cells. Jpn. J. Physiol. 54, 23–29. Roe, J.S., Kim, H., Lee, S.M., Cho, E.J. and Youn, H.D. (2006) p53 stabilization and transactivation by a von hippel-Lindau protein. Mol. Cell 22, 395–405. Zhu, L. and Goldstein, B. (2002) Use of an anaerobic chamber environment for the assay of endogenous cellular protein-tyrosine phosphatase activities. Biol. Proc. Online 4, 1–9. Roth, M.B. (1995) Spheres, coiled bodies and nuclear bodies. Curr. Opin. Cell Biol. 7, 325–328.

852 [34] Spector, D.L. (2001) Nuclear domains. J. Cell Sci. 114, 2891– 2893. [35] Maul, G.G., Negorev, D., Bell, P. and Ishov, A.M. (2000) Review: properties and assembly mechanisms of ND10, PML bodies, or PODs. J. Struct. Biol. 129, 278–287. [36] Jang, M.S., Ryu, S.W. and Kim, E. (2002) Modification of Daxx by small ubiquitin-related modifier-1. Biochem. Biophys. Res. Commun. 295, 495–500. [37] Salomoni, P. and Pandolfi, P.P. (2002) The role of PML in tumor suppression. Cell 108, 165–170. [38] Zhong, S., Salomoni, P., Ronchetti, S., Guo, A., Ruggero, D. and Pandolfi, P.P. (2000) Promyelocytic leukemia protein (PML) and Daxx participate in a novel nuclear pathway for apoptosis. J. Exp. Med. 191, 631–640. [39] Zhao, L.Y., Liu, J., Sidhu, G.S., Niu, Y., Liu, Y., Wang, R. and Liao, D. (2004) Negative regulation of p53 functions by Daxx and the involvement of MDM2. J. Biol. Chem. 279, 50566–50579. [40] Elbashir, S.M., Harborth, J., Lendeckel, W., Yalcin, A., Weber, K. and Tuschl, T. (2001) Duplexes of 21-nucleotide RNAs mediate RNA interference in cultured mammalian cells. Nature 411, 494–498. [41] Kudo, N., Matsumori, N., Taoka, H., Fujiwara, D., Schreiner, E.P., Wolff, B., Yoshida, M. and Horinouchi, S. (1999) Leptomycin B inactivates CRM1/exportin 1 by covalent modification at a cysteine residue in the central conserved region. Proc. Natl. Acad. Sci. USA 96, 9112–9117. [42] Davis, R.J. (2000) Signal transduction by the JNK group of MAP kinases. Cell 103, 239–252. [43] Wu, S., Loke, H.N. and Rehemtulla, A. (2002) Ultraviolet radiation-induced apoptosis is mediated by Daxx. Neoplasia 4, 486–492.

Y.-S. Jung et al. / FEBS Letters 581 (2007) 843–852 [44] Kitagawa, D., Kajiho, H., Negishi, T., Ura, S., Watanabe, T., Wada, T., Ichijo, H., Katada, T. and Nishina, H. (2006) Release of RASSF1C from the nucleus by Daxx degradation links DNA damage and SAPK/JNK activation. EMBO J. 25, 3286–3297. [45] Tang, J., Qu, L.K., Zhang, J., Wang, W., Michaelson, J.S., Degenhardt, Y.Y., El-Deiry, W.S. and Yang, X. (2006) Critical role for Daxx in regulating Mdm2. Nat. Cell Biol. 8, 855– 862. [46] Seth, R., Yang, C., Kaushal, V., Shah, S.V. and Kaushal, G.P. (2005) p53-dependent caspase-2 activation in mitochondrial release of apoptosis-inducing factor and its role in renal tubular epithelial cell injury. J. Biol. Chem. 280, 31230–31239. [47] Zhao, R., Xiang, N., Domann, F.E. and Zhong, W. (2006) Expression of p53 enhances selenite-induced superoxide production and apoptosis in human prostate cancer cells. Cancer Res. 66, 2296–2304. [48] Buganim, Y., Kalo, E., Brosh, R., Besserglick, H., Nachmany, I., Rais, Y., Stambolsky, P., Tang, X., Milyavsky, M., Shat, I., Kalis, M., Goldfinger, N. and Rotter, V. (2006) Mutant p53 protects cells from 12-O-tetradecanoylphorbol-13-acetate-induced death by attenuating activating transcription factor 3 induction. Cancer Res. 66, 10750–10759. [49] Borden, K.L. (2002) Pondering the promyelocytic leukemia protein (PML) puzzle: possible functions for PML nuclear bodies. Mol. Cell. Biol. 22, 5259–5269. [50] Takahashi, Y., Lallemand-Breitenbach, V., Zhu, J. and de The, H. (2004) PML nuclear bodies and apoptosis. Oncogene 23, 2819– 2824. [51] Ching, R.W., Dellaire, G., Eskiw, C.H. and Bazett-Jones, D.P. (2005) PML bodies: a meeting place for genomic loci? J. Cell Sci. 118, 847–854.