THE JOURNAL OF BIOLOGICAL CHEMISTRY © 2004 by The American Society for Biochemistry and Molecular Biology, Inc.
Vol. 279, No. 17, Issue of April 23, pp. 17148 –17157, 2004 Printed in U.S.A.
Translational Induction of the Inhibitor of Apoptosis Protein HIAP2 during Endoplasmic Reticulum Stress Attenuates Cell Death and Is Mediated via an Inducible Internal Ribosome Entry Site Element* Received for publication, August 7, 2003, and in revised form, January 8, 2004 Published, JBC Papers in Press, February 11, 2004, DOI 10.1074/jbc.M308737200
Dinesh Warnakulasuriyarachchi, Sonia Cerquozzi, Herman H. Cheung, and Martin Holcı´k‡ From the Apoptosis Research Center, Children’s Hospital of Eastern Ontario, Department of Pediatrics, University of Ottawa, Ottawa, Ontario K1H 8L1, Canada
Prolonged endoplasmic reticulum (ER) stress leads to activation of caspases and cell death. The inhibitor of apoptosis (IAP) proteins are intrinsic inhibitors of apoptosis by virtue of inhibiting distinct caspases and are, therefore, critical regulators of cell death. Here we demonstrate that the expression of one member of the IAP family, HIAP2, is induced in response to ER stress and attenuates ER stress-induced cell death. The induction of HIAP2 is executed at the level of protein synthesis and is mediated by an inducible internal ribosome entry site (IRES) element. The triggering of ER stress results in caspase-mediated proteolytic processing of eukaryotic initiation factor p97/DAP5/NAT1, producing a fragment that specifically activates HIAP2 IRES. These data suggest an existence of a novel mechanism that regulates apoptotic response in ER stress.
Endoplasmic reticulum emerged recently as an intracellular organelle involved in the control of apoptosis. Endoplasmic reticulum is the site of protein folding and posttranslational modifications. As such, the ER is sensitive to perturbations in homeostasis triggered by different stimuli that results in the induction of a stress response known as unfolded protein response (UPR)1 (1). The UPR is characterized by reduction of global protein synthesis and specific up-regulation of stressresponse genes such as ATF4 or BiP. The role of UPR is to protect cell from the ER stress. Indeed, the failure to alleviate the stress leads to the activation of apoptotic pathway(s) and cell death. Recent evidence suggests that there exists an ERspecific apoptotic pathway in which caspase-12 is activated in response to ER stress (2). Activation of caspase-12 leads to the activation of downstream caspases-9 and -3, forming a caspase cascade analogous to the mitochondria-induced pathway (3). It is not clear, however, which caspase functions as an initiator caspase during the ER stress conditions in human cells since * This work was supported in part by Canadian Institutes of Health Research Operating Grant 43984. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ‡ A Canadian Institutes of Health Research New Investigator. To whom correspondence should be addressed: Apoptosis Research Center, Children’s Hospital of Eastern Ontario Research Institute, Rm. R310, 401 Smyth Rd., Ottawa, ON K1H 8L1, Canada. Tel.: 613-738-3207; Fax: 613-738-4833; E-mail:
[email protected]. 1 The abbreviations used are: UPR, unfolded protein response; ER, endoplasmic reticulum; IAP, inhibitor of apoptosis; uORF, upstream open reading frame; UTR, untranslated region; PARP, poly(ADP-ribose) polymerase; CAT, chloramphenicol acetyltransferase; ANOVA, analysis of variance; eIF, eukaryotic initiation factor; z-VAD-FMK, benzyloxycarbonyl-VAD-fluoromethyl ketone; IRES, internal ribosome entry sites; TRAF, tumor necrosis factor receptor-associated factor.
the functional caspase-12 is lost in humans (4). The inhibitor of apoptosis (IAP) family of proteins emerged a key regulators of apoptosis by virtue of their ability to directly bind and inhibit distinct caspases (5). Among the eight members of this family the X-linked inhibitor of apoptosis, XIAP, is the most potent inhibitor of both the initiator and effector caspases, although other members of the IAP family are thought to play important roles in a subset of apoptotic conditions (6, 7). HIAP2 (GenBankTM XM040717; also known as c-IAP1, MIHB, BIRC2) is a multifunctional protein that was initially identified through the interaction with the tumor necrosis factor-receptor complex proteins TRAF1 and TRAF2 (8). In addition to suppressing apoptosis induced by various triggers by directly binding to and inhibiting caspases-3 and -7 (9, 10), HIAP2 interacts with TRAFs and regulates the signaling through the tumor necrosis factor receptor (8). HIAP2 also exerts a pro-apoptotic activity; the C-terminal RING finger of HIAP2 possess a ubiquitin E3 ligase activity and sensitizes cells to apoptosis by directing ubiquitylation and removal of TRAF2 after the activation of tumor necrosis factor receptor (11). We have shown previously that the expression of HIAP2 is controlled at the level of translation (12). There exists a short regulatory upstream open reading frame (uORF) that is located in the 5⬘-UTR of HIAP2 and exerts an inhibitory effect on the translation of HIAP2. Here, we have investigated the expression of HIAP2 during ER stress. We demonstrate that HIAP2 is translationally up-regulated in response to ER stress induced by thapsigargin or tunicamycin. Surprisingly, the mechanism of this induction is independent of the regulatory uORF, and instead, it utilizes an inducible IRES element. The activity of the HIAP2 IRES is low in un-stressed cells but is enhanced after induction of ER stress by a caspase cleavage fragment of p97/ DAP5/NAT1. Induction of HIAP2 attenuates apoptotic response after ER stress since antisense-mediated down-regulation of HIAP2 enhanced sensitivity of HeLa cells to tunicamycin. EXPERIMENTAL PROCEDURES
Plasmids—Plasmids used in this study have been previously described as follows: pgal/CAT, pgal/XIAP/CAT, and pCAT (13); HIAP2(1–1222)CAT, HIAP2(1073–1222)CAT, HIAP2(1073–1222)(CUA1147)CAT, gal/HIAP2(1–1222)CAT, gal/HIAP2(1073–1222)CAT, and pgal (12); FLAG-p97 and FLAG-p86 (14), pcDNA3-HIAP2 (15). Bi-cistronic vectors containing various fragments of the HIAP2 5⬘-UTR were constructing by inserting the -galactosidase gene in front of the HIAP2 5⬘-UTR fragments in monocistronic constructs that were described previously (12). The plasmid hp-pgal/CAT was constructed by inserting the synthetic hairpin (ccg gcc ggc cgg ccg gcc ggc caa agg ccg gcc ggc cgg ccg gcc ggc tag cgg tac) in front of the -galactosidase reporter gene in plasmid pgal/CAT. In all plasmids the expression is directed from a cytomegalovirus promoter. All constructs were confirmed by nucleotide sequencing. Cell Culture and Transfection Reagents—Human embryonic kidney
17148
This paper is available on line at http://www.jbc.org
Induction of HIAP2 during ER Stress (HEK293T) and human cervical carcinoma (HeLa) cells were cultured in standard conditions in Dulbecco’s modified Eagle’s medium supplemented with 10% fetal calf serum, glutamine, and antibiotics as described previously (16). Transient DNA transfections were conducted using LipofectAMINE Plus reagent (Invitrogen) with the protocol provided by the manufacturer and 1 g of plasmid DNA per 6-well plate well. When mono-cistronic plasmids were used, 0.5 g of pgal reporter plasmid was included in each transfection to normalize for transfection efficiency. Co-transfection experiments were done as described above except that 1 g of bi-cistronic plasmid was combined with 1 g of indicated expression plasmid. For the endoplasmic stress experiments cells were first transfected with the indicated plasmids. 24 h after transfection the cells were treated with thapsigargin (5 M), tunicamycin (8.5 M), z-VAD-FMK (100 M) or Me2SO for 24 h and then collected for analysis as described below. Cell Viability Assay—HeLa cells were transiently transfected in 6-well plates with either pcDNA-HIAP2 or pgal plasmids as described above. For antisense experiments cells were infected with adenoviral vector encoding either LacZ or antisense HIAP2 cDNAs described previously (17) at a multiplicity of infection of 5. 24 h post-transfection/ infection the cells were treated with tunicamycin (8.5 M) or Me2SO for 24 h, and cell viability was determined by WST-1 (Roche Applied Science) for each treatment. All data are shown as an average ⫾ S.D. of three independent experiments performed in triplicate. Western blots were performed on parallel samples to determine the levels of HIAP2 expression. Immunoblot Analysis—Cells were harvested in ice-cold phosphatebuffered saline, lysed in radioimmune precipitation assay buffer (1% Nonidet P-40, 1% sodium deoxycholate, 0.1% SDS, 0.15 M NaCl, 0.01 M sodium phosphate, pH 7.2, 2 mM EDTA/0.1 mM phenylmethylsulfonyl fluoride) for 30 min at 4 °C followed by centrifugation at 14,000 ⫻ g for 10 min. Protein concentration was assayed by the protein assay kit (Bradford assay; Bio-Rad), and equal amounts of protein samples were separated by 10% SDS-PAGE. Samples were analyzed by Western blotting for expression of HIAP2, XIAP, BiP, actin, eIF4G, PARP, processed caspase 3, and FLAG-p97 using rabbit polyclonal anti-XIAP (Ægera), anti-eIF4G (18), anti-PARP (Cell Signaling), or mouse monoclonal anti-HIAP2 (R&D Laboratories), anti-GRP78/BiP (Transduction Laboratories), anti-FLAG M2 (Stratagene), anti-processed caspase 3 (Cell Signaling), and anti--actin (Sigma) antibodies, respectively, followed by secondary antibody (horseradish peroxidase-conjugated goat anti-rabbit or donkey anti-mouse IgG; Amersham Biosciences). Antibody complexes were detected using the ECL Plus system (Amersham Biosciences). -Galactosidase and CAT Analysis—Transiently transfected cells were harvested 24 h after transfection in the CAT enzyme-linked immunosorbent assay kit lysis buffer (Roche Applied Science), and cell extracts were prepared using the protocol provided by the manufacturer. -Galactosidase enzymatic activity in cell extracts was determined by the spectrophotometric assay using o-nitrophenyl--D-galactopyranoside (19), and the CAT levels were determined using the CAT enzyme-linked immunosorbent assay kit (Roche Applied Science) and the protocol provided by the manufacturer. The relative CAT activity and IRES activity were determined as a ratio of CAT/-galactosidase. All data are shown as an average ⫾ S.D. of three independent experiments performed in triplicate. In Vitro Transcription and Translation—Coupled in vitro transcription and translations (TNT Quick Coupled Transcription/Translation System, Promega) were performed under the conditions recommended by the manufacturer. Each reaction was programmed with 1 g of purified plasmid DNA of the indicated plasmids. In vitro translated proteins were labeled with L-[35S]methionine (Amersham Biosciences). Reactions were incubated at 30 °C for 90 min and analyzed on 12.5% SDS-PAGE. The intensities of the CAT band were determined on a Bio-Rad Model GS-670 imaging densitometer. Ribonuclease Protection Assay—Total RNA was isolated from untransfected or transiently transfected cells with the Trizol reagent (Invitrogen) and treated with 1 unit of DNase I. The 32P-labeled antisense RNA probes spanning 149 nucleotides of the HIAP2 5⬘-UTR (coordinates 1073–1222) or human glyceraldehyde-3-phosphate dehydrogenase (Pharmingen) were synthesized using the MaxiScript T7 transcription kit (Ambion) and gel-purified. Ribonuclease protection assay analysis was performed using the RPA III kit (Ambion) and the protocol provided by the manufacturer. Briefly, 10 g of total RNA was hybridized overnight at 42 °C with 5 ⫻ 104 cpm of high specific activity probe and then digested with RNase A/RNase T1 mix. The digested samples were ethanol-precipitated and separated on a denaturing polyacrylamide gel. The protected fragments were visualized by de-
17149
tection to x-ray film and quantified on a Bio-Rad Model GS-670 imaging densitometer. UV Cross-linking and Immunoprecipitation—The cellular extracts (S10) were prepared from p97- or p86-transfected 293 T cells as previously described (20). RNA-protein complexes for UV cross-linking and immunoprecipitation were prepared as described (21). The samples were irradiated in a 96-well dish on ice with a 254-nm UV light source at 400,000 J/cm2. UV-irradiated RNA-protein complexes were treated with RNase T1 (1 unit/l) for 10 min at room temperature, and heparin, SDS, and -mercaptoethanol were added to final concentrations of 5 mg/ml and 1 and 1%, respectively. The samples were boiled for 5 min and then diluted in the binding buffer (50 mM Tris-HCl, pH 8.0, 150 mM NaCl, 1 mM EDTA, 0.1% Triton X-100, 0.1 mM phenylmethylsulfonyl fluoride, 5 g/ml aprotinin, 5 g/ml leupeptin). The nonspecific protein binding was removed by incubating samples with 20 l of protein A ⫹ G-agarose (Calbiochem) for 1 h at 4 °C. The samples were then incubated with protein A ⫹ G-agarose precoated with 4 g of monoclonal anti-FLAG antibody for 2 h at 4 °C and washed 5 times in 600 l of the binding buffer. The immunoprecipitated RNA-protein complexes were released from the agarose beads by boiling in SDS-loading buffer (100 mM Tris, pH 6.8, 2.5% SDS, 10% glycerol, 0.025% -mercaptoethanol, 0.1% bromphenol blue), resolved on a 7.5% SDS-PAGE, and visualized by autoradiography. RESULTS
Induction of Endoplasmic Reticulum Stress Up-regulates HIAP2—We observed that the levels of IAP protein HIAP2 were increased in HEK293T (Fig. 1A) or HeLa cells (data not shown) treated with thapsigargin (an inhibitor of ER calcium pump) or tunicamycin (an inhibitor of protein glycosylation). The induction of HIAP2 paralleled the expression pattern of a known ER stress-inducible gene, BiP. In contrast, the levels of another member of the IAP family, XIAP, remained unchanged, suggesting that not all members of the IAP family are induced by ER stress. When we examined the RNA samples from treated and untreated cells we noted that the levels of HIAP2 mRNA did not change in response to drug treatments (Fig. 1B). Thus, the induction of HIAP2 protein after the ER stress could be due to translational up-regulation of HIAP2 protein. We have shown previously that the 5⬘-UTR of HIAP2 is refractory to translation because it contains a regulatory upstream open reading frame (12). We, therefore, wanted to determine if the HIAP2 5⬘-UTR or its uORF is involved in the regulation of HIAP2 expression in response to ER stress. Monocistronic CAT reporter plasmids in which the 5⬘-UTR of HIAP2 (with or without the mutation in the uORF-initiating CUG) is fused with the CAT gene (Fig. 1C) were transiently co-transfected with plasmid pgal into HEK293T cells, and the cells were treated with thapsigargin or tunicamycin. We observed that the translation of CAT from the 5⬘-UTR-containing construct, but not in the absence of the 5⬘-UTR, was enhanced in drug-treated cells (Fig. 1D). This observation paralleled the behavior of the endogenous HIAP2 in drug-treated cells, suggesting that the induction of HIAP2 after ER stress is due to the translation up-regulation that is mediated by the 5⬘-UTR. Surprisingly, the CUG 3 CUA mutation that prevents the translation of uORF (12) had no effect on the induction of CAT after ER stress. Therefore, although the 5⬘-UTR mediates induction of HIAP2 in response to ER stress, this induction is independent of the regulatory uORF. Triggering of the ER stress results in the repression of global protein synthesis. Indeed, both thapsigargin and tunicamycin inhibited the translation of -galactosidase reporter gene by ⬃30% (Fig. 1E). In contrast, the translation of the CAT reporter remained unchanged (Fig. 1E). This resistance to the ER-induced inhibition of translation was mediated by the HIAP2 5⬘-UTR since in the construct lacking the 5⬘-UTR the translation of CAT was inhibited to the same extent as that of -galactosidase (Fig. 1E). This suggests that the 5⬘-UTR of
17150
Induction of HIAP2 during ER Stress
FIG. 1. HIAP2 is translationally induced after ER stress, and this induction is independent of the uORF. A, Western blot (WB) analysis of endogenous HIAP2 and XIAP proteins in the HEK293T cell lysates treated with Me2SO (DMSO) (D), thapsigargin (Tg, 5 M), or tunicamycin (Tm, 8.5 M). Expression of HIAP2, but not XIAP, is induced after the induction of ER stress, which was confirmed by the induction of ER chaperone BiP. B, ribonuclease protection assay (RPA) analysis of HIAP2 mRNA in treated and untreated cells (same as A). GAPDH, glyceraldehyde-3-phosphate dehydrogenase. C, schematic diagram of the CAT reporter constructs harboring the 5⬘-UTR of HIAP2 with or without the uORF-initiating CUG codon. D, HEK293 T cells were transiently co-transfected with pgal and HIAP2 5⬘-UTR constructs indicated in C, and the effect of thapsigargin and tunicamycin on HIAP2 translation were determined by measuring the -galactosidase and CAT activities as described under “Experimental Procedures.” The relative CAT activity in Me2SO-treated cells was set as 100 (*, p ⬍ 0.05, compared with Me2SO, one-way ANOVA). E, relative activities of each reporter gene from the experiments described in D. The activity of each reporter gene in Me2SO-reated cells was set as 100 (*, p ⬍ 0.05, CAT compared with -galactosidase, one-way ANOVA).
HIAP2 allows translation of HIAP2 in response to ER stress by a mechanism that escapes inhibition of global protein synthesis. HIAP2 Is Translated by an Inducible IRES Element after ER Stress—An alternative mechanism of translation initiation that is often resistant to the changes in global proteins synthesis has been described for several cellular genes and is medi-
ated by IRES elements (for review, see Ref. 22). The data described thus far were consistent with the notion that ERinduced HIAP2 translational up-regulation could be mediated by internal initiation. However, when we tested the IRES activity of HIAP2 5⬘-UTR previously we failed to observe any IRES activity (12). We, therefore, hypothesized that the induction of ER stress may be necessary for the activation of the
Induction of HIAP2 during ER Stress
17151
FIG. 2. Translational induction of HIAP2 is mediated by an IRES element. A, HEK293T cells were transiently transfected with the indicated plasmids, and the relative IRES activity was determined in Me2SO (D)- or thapsigargin (Tg)-treated cells. The IRES activity in Me2SO-treated cells was set as 100 for each construct (*, p ⬍ 0.05, one-way ANOVA). B, the integrity of the transiently transfected bicistronic RNA -gal/HIAP2(1073–1222)/CAT in control and thapsigargin-treated cells was determined by ribonuclease protection assay. The position of the RNA probe is indicated above. C, indicated hairpin-containing plasmids were transfected into HEK293T cells, and the activity of each reporter construct was determined in control or thapsigargin-treated cells (AU, arbitrary units) (*, p ⬍ 0.05, one-way ANOVA). D, rabbit reticulocyte lysates were programmed with hairpin-containing bi-cistronic plasmids (as in C) and the in vitro translated products were separated on a 12.5% SDSelectrophoresis gel. The position of the molecular weight markers is indicated on the left; the position of the CAT product is indicated on the right.
17152
Induction of HIAP2 during ER Stress
FIG. 3. Activation of the HIAP2 IRES is caspase-dependent. A, transiently transfected HEK293T cells were pretreated with a pan-caspase inhibitor z-VAD-FMK or Me2SO (D) for 4 h, then treated with thapsigargin (Tg) or tunicamycin (Tm) for 24 h, and the relative IRES activity was determined as described under “Experimental Procedures.” The IRES activity in Me2SO-treated cells was set as 100 (*, p ⬍ 0.05, one-way ANOVA). B, the expression of eIF4G, p97/DAP5/NAT1, PARP, BiP, and actin in thapsigargin-, tunicamycin-, and Me2SO-treated cells was determined by Western blot analysis 24 h after drug treatment. The molecular masses of the indicated fragments are shown on the right; the identity of the detected proteins is shown on left.
HIAP2 IRES element. Indeed, when HEK293T or HeLa cells transfected with the bi-cistronic construct -gal/HIAP2(1073– 1222)/CAT were treated with thapsigargin or tunicamycin, we observed the induction of CAT translation similar to that observed with the monocistronic constructs (Fig. 2A and data not shown). Interestingly, the insertion of the full-length1.2-kilobase HIAP2 5⬘-UTR (plasmid -galactosidase/HIAP2(1–1222)/ CAT) resulted in higher IRES activity. This activation of translation was not due to an alternative splicing since no differences in the integrity or the amount of the bi-cistronic mRNA were observed in treated and untreated cells (Fig. 2B). In addition, reversing the orientation of the HIAP2 5⬘-UTR or the deletion of the cytomegalovirus promoter in the bi-cistronic plasmid prevented induction of CAT expression in thapsigargin-treated cells (data not shown). To confirm that HIAP2 5⬘-UTR harbors an IRES element we inserted a stable hairpin in front of the -galactosidase reporter gene that inhibits its translation by preventing ribosomal scanning. After the induction of ER stress by thapsigargin we observed that translation of CAT remained unchanged, although the levels of -galactosidase remained almost undetectable (Fig. 2C). The translation of CAT was due solely to the presence of the 5⬘-UTR since we could not detect any CAT levels in the construct lacking the 5⬘-UTR. In addition, when the rabbit reticulocyte lysates were programmed with the hairpin-containing bi-cistronic construct we detected a single CAT protein band of the expected size (26 kDa) in the construct containing the 5⬘-UTR but not with the control construct lacking the 5⬘-UTR (Fig. 2D). These data are consistent with the existence of an IRES element in the 5⬘-UTR of HIAP2. Translational Activation of HIAP2 IRES Is Dependent on Active Caspases—Prolonged ER stress leads to the activation of programmed cell death pathways and apoptosis. It was sug-
gested that, at least in rodent cells, caspase-12, which translocates to the outer membrane of the endoplasmic reticulum after ER stress, functions as the initiator caspase in response to stress stimuli and initiates the subsequent activation of other caspases (2, 3). In human cells the role of caspase-12 is likely replaced by an as yet unidentified caspase since the functional caspase-12 is lost (4). We, therefore, wished to test whether the activation of caspases is necessary for the induction of HIAP2 IRES. The activity of HIAP2 IRES reporter construct pgal/ HIAP2(1–1222)/CAT was tested in cells treated with thapsigargin or tunicamycin and in which caspase activation was blocked by treatment with pan-caspase inhibitor z-VAD-FMK (Fig. 3A). Indeed, we observed that in both thapsigargin- and tunicamycin-treated cells the pretreatment of cells with z-VAD-FMK inhibited induction of HIAP2 IRES. This suggested that caspase activation is necessary for the activation of HIAP2 IRES. Several laboratories have demonstrated that activation of caspases will result in the cleavage of several translation initiation factors (eIFs), resulting in the inhibition of protein synthesis (23). Among the eIFs, the caspase-mediated degradation of members of the eIF4G family (specifically eIF4GI and p97/DAP5/NAT1) was implicated in the regulation of cell survival during apoptosis (14, 23–25). We, therefore, examined the levels of eIF4GI and p97/DAP5/NAT1 during ER stress (Fig. 3B). In both thapsigargin and tunicamycin-treated cells we observed the disappearance of the full-length eIF4GI protein that was accompanied by the appearance of the 76-kDa M-FAG (middle fragment of apoptotic cleavage of eIF4G) fragment, which is characteristic of caspase-mediated degradation of eIF4GI during apoptosis (18). Because the available antibody to p97/DAP5/NAT1 does not recognize the cleavage fragment of p97/DAP5/NAT1 (data not shown), we transfected HEK293T
Induction of HIAP2 during ER Stress
17153
FIG. 4. Overexpression of p86 fragment of p97/DAP5/NAT1 specifically activates HIAP2 but not XIAP. A, HEK293T cells were co-transfected with HIAP2 or XIAP IRES containing bi-cistronic plasmids and p97- or p86-expressing pCI-FLAG vector, and the relative activity of IRES was determined 24 h post-transfection. The IRES activity in pCI-transfected cells was set as 100 for each IRES construct. The expression levels of p97 and p86 proteins were determined by Western blot analysis with anti-FLAG antibody (inset) (*, p ⬍ 0.05, one-way ANOVA). B, the levels of endogenous HIAP2 and XIAP proteins were determined in pCI-, p86-, or p97-transfected cells from A as described under “Experimental Procedures.” The identity of detected proteins is indicated on the left. C, to determine the effect of p97 and p86 on each reporter cistron, the raw data from A were normalized for total cellular protein and plotted (AU, arbitrary units) (*, p ⬍ 0.05, one-way ANOVA). D, rabbit reticulocyte lysates were programmed with the indicated hairpin-containing bi-cistronic plasmids, and the in vitro translated products were separated on a 12.5% SDS-PAGE gel. The position of the molecular weight markers is indicated on the left; the position of the CAT, p86, and p97 products is indicated on the right. The intensity of CAT, p97, and p86 bands was quantified from three independent gels as described under “Experimental Procedures,” and the expression of CAT relative to p86 or p97 expression was calculated and is indicated below the gel (average ⫾ S.D.). The relative expression of CAT in the first lane was set as 1.
cells with FLAG-tagged p97/DAP5/NAT1-expressing vector and examined the processing of p97/DAP5/NAT1 using the anti-FLAG antibody. Similarly to eIF4GI, the induction of endoplasmic reticulum stress by both drugs resulted in the cleavage of p97/DAP5/NAT1 and appearance of the 86-kDa caspasemediated fragment (p86) characterized previously (26). The proteolytic processing of eIF4GI and p97/DAP5/NAT1 paralleled the processing of PARP protein, a bona fide caspase substrate and was accompanied by the induction of BiP, which is indicative of ER stress (Fig. 3B). Activation of HIAP2 IRES and Induction of HIAP2 Protein Are Mediated by the p86 Fragment of p97/DAP5/NAT1—We and others have suggested previously that the caspase-mediated p86 fragment of p97/DAP5/NAT1 could function as a stress-specific mediator of IRES translation (14, 24 –26). We, therefore, wished to test whether the p86 fragment of p97/ DAP5/NAT1, which is produced during ER stress, is directly responsible for the activation of HIAP2 IRES. HEK293T cells
were co-transfected with HIAP2 IRES reporter construct pgal/HIAP2(1–1222)/CAT and either FLAG-p97- or FLAGp86-expressing vectors. Activation of HIAP2, but not of XIAP IRES, was observed in p86 but not in p97 co-transfected cells (Fig. 4A). Importantly, Western blot analysis of the protein lysates from transfected cells confirmed that overexpression of p86 but not p97 resulted in increased levels of HIAP2, but not XIAP, endogenous protein as well (Fig. 4B). In addition, overexpression of p86 fragment did not result in the activation of a construct with HIAP2 5⬘-UTR in the reverse orientation or a construct without an insert (data not shown). The activation of the HIAP2 IRES was also observed in the hairpin-containing reporter construct (Fig. 4C). In addition, the increase in IRES activity was due to the specific increase in the translation of CAT and not a decrease in the translation of -galactosidase (Fig. 4C). To further demonstrate the induction of HIAP2 IRES by p86 we utilized the in vitro transcription/translation system.
17154
Induction of HIAP2 during ER Stress
FIG. 5. Deletion analysis of HIAP2 IRES. A, DNA segments corresponding to the indicated regions of the HIAP2 5⬘UTR were inserted into the linker region of the pgal/CAT plasmid, and the IRES activity was determined in p86-cotrasfected HEK293T cells. The activity of 1.2kilobase HIAP2 UTR segment was set as 100. B, rabbit reticulocyte lysates were programmed with the indicated bi-cistronic plasmids, and the in vitro translated products were separated on a 12.5% SDS-PAGE gel. The position of the molecular weight markers is indicated on the left; the position of the CAT and -galactosidase products is indicated on the right. C, HEK293T cells were transfected with the indicated bi-cistronic plasmids, and the IRES activity was determined in Me2SO (DMSO)- or thapsigargin (Tg)treated cells. The IRES activity in Me2SO-treated cells was set as 100 (*, p ⬍ 0.05, one-way ANOVA).
Rabbit reticulocyte lysates were primed with the hairpin-containing HIAP2 IRES reporter construct and either the FLAGp97- or FLAG-p86-expressing plasmids. The addition of p86 fragment resulted in the activation of HIAP2 IRES (Fig. 4D, compare the second and third lanes). This activation was spe-
cifically mediated by the HIAP2 5⬘-UTR sequence since we did not observe any CAT protein product with the bi-cistronic construct that lacks the 5⬘-UTR insert (Fig. 4D, first lane). These data suggest that the caspase-mediated cleavage fragment of p97/DAP5/NAT1, p86, selectively enhances translation of
Induction of HIAP2 during ER Stress HIAP2 IRES element both in vivo and in vitro in a manner that is independent of the translation of the upstream cistron. Activation of HIAP2 IRES by p86 Correlates with the Activation of HIAP2 IRES by ER Stress—To further explore the link between HIAP2 IRES activity and p86 we sought to determine the region(s) within the HIAP2 5⬘-UTR that mediates HIAP2 IRES induction by p86. These mapping experiments were performed in HEK293T cells that were co-transfected with the FLAG-p86-expressing plasmid. Deletion analysis of the HIAP2 5⬘-UTR identified two regions of interest. The first, spanning the 5⬘ proximal 990 nucleotides of HIAP2 5⬘-UTR, appears to have a somewhat inhibitory function on IRES trans-
FIG. 6. Cleavage of p97 and the induction of HIAP2 occur at early stages of ER stress. The expression of HIAP2, p97/DAP5/NAT1, caspase 3 (cleaved form), PARP, BiP, and actin was determined by Western blot analysis at 3, 6, 12, and 18 h after exposure of cells to tunicamycin (Tm). The molecular weight of the indicated fragments is shown on the right; the identity of the detected proteins is shown on the left.
17155
lation since the deletion of this region resulted in the 1.5–2-fold enhancement of the HIAP2 IRES activity (Fig. 5A). This suggests that there may exist additional protein factors that coregulate HIAP2 IRES activity. The second region is located between the nucleotides 1073 and 1142. The deletion of this sequence completely abolished HIAP2 IRES activity both in transfected cells (Fig. 5A) and in reticulocyte lysate (Fig. 5B), suggesting that this region is essential for IRES activity and the p86-mediated induction of HIAP2 IRES. More importantly, this sequence is also necessary for the induction of HIAP2 IRES activity by ER stress since the construct lacking the 1073–1142 sequence failed to exhibit IRES activity in thapsigargin-treated cells (Fig. 5C). Similarly, deletion of the 990 –1222 region abolished IRES activity in thapsigargin-treated cells (Fig. 5C). Together, these results suggest that the sequence between the coordinates 1073 and 1142 is essential for the ER stress-induced HIAP2 IRES and that this induction is mediated by the caspase-mediated p86 cleavage fragment of p97/DAP5/NAT1. Our attempts to cross-link and immunoprecipitate the HIAP2 IRES RNA with either the p97 or p86 proteins failed, presumably because p97/p86 is not a direct RNA-binding protein (data not shown). If this is the case, then the activation of HIAP2 IRES brought about by p86 is mediated by another IRES binding protein. The efforts to identify the full ensemble of HIAP2 IRES-binding proteins are ongoing. Induction of HIAP2 Occurs at Initial Stages of ER Stress— Because the cleavage of p97 and subsequent induction of HIAP2 could be a result of nonspecific caspase processing at the final stages of ER-stress-induced cell death, we wished to examine the time course of p97 processing, caspase activation, and induction of HIAP2. We observed that processing of p97 correlated with the induction of HIAP2 and occurred 6 h after treatment of cells with tunicamycin and preceded the induction
FIG. 7. Changes in HIAP2 levels modulate ER stress-induced cell death. A, HeLa cells were transiently transfected with pcDNA3-HIAP2 or pgal plasmids and treated with tunicamycin (8.5 M) or Me2SO (DMSO), and cell viability was determined 24 h later by WST-1 assay. The data are shown as an average ⫾ S.D. of three independent experiments performed in triplicates. The cell viability in Me2SO-treated pgal-transfected cells was set as 100 (*, p ⬍ 0.05, compared with LacZ, one-way ANOVA). The levels of HIAP2 in transfected cells were determined by Western blot analysis (top). B, HeLa cells were infected with adenoviral (adeno) vectors encoding either LacZ or HIAP2 antisense (HIAP2as) cDNA and treated as in A. The levels of HIAP2 in infected cells were determined by Western blot analysis (top).
17156
Induction of HIAP2 during ER Stress
of BiP (Fig. 6). Interestingly, however, although we observed the cleavage of PARP at an earlier time point (3 h), the processing of caspase-3, the terminal executioner caspase, occurred only 12 h post-tunicamycin treatment. These data suggest that the processing of p97 and the induction of HIAP2 both occur at the early phase of ER stress response and are not a consequence of general degradation of translation machinery in the final stages of apoptosis. Changes in HIAP2 Levels Modulate ER Stress-induced Cell Death—We have shown that HIAP2 is increased after ER stress. We, therefore, wished to explore the physiological significance of this observation. Because prolonged ER stress leads to cell death and since HIAP2 is a potent inhibitor of cell death, we reasoned that increased expression of HIAP2 should prevent or delay ER stress-induced cell death. To test this hypothesis HeLa cells were transfected with either HIAP2- or -galactosidase-expressing vectors, and cell viability was determined 24 h after treatment with tunicamycin. We observed that overexpression of HIAP2, but not -galactosidase, attenuated tunicamycin-induced cell death (Fig. 7). Conversely, down-regulation of HIAP2 by antisense targeting enhanced cell sensitivity to tunicamycin. Taken together, these data indicate that HIAP2 levels are important in mediating the cellular response to ER stress. DISCUSSION
IRES elements are RNA regulatory motifs that have been originally discovered in the RNAs of picornaviruses. The RNA of these viruses is naturally uncapped and utilizes the IRES to recruit ribosomes directly to the vicinity of the initiation codon and facilitate translation (27). Recently, cellular IRES elements emerged as important regulators of translation of selected cellular mRNAs, particularly in conditions of cellular stress (22, 25). In contrast to the translation of a typical cellular mRNA that is capped, IRES translation does not require the presence of the m7G cap-binding protein eIF4E. Because the majority of the translational control is executed through the modulation of the initiation step of translation, IRES translation escapes many of the control mechanisms that govern cap-dependent translation (22). Indeed, of less than 50 cellular IRES elements identified to date, almost half are found in genes encoding oncogenes, growth factors, and proteins involved in the regulation of apoptosis, arguing that IRES-dependent translation is an important regulatory mechanism for the maintenance of cellular homeostasis (22, 25). It is becoming evident that IRES elements are controlled by different regulatory mechanisms. For instance, the IRES-mediated translation of the inhibitor of caspases XIAP is increased in response to serum starvation and low dose ␥ irradiation (13, 16) but not in etoposide (14) or ER stress-inducing drug-treated cells (Fig. 1A). IRES elements of some genes, such as cat-1, platelet-derived growth factor, c-myc, or vascular endothelial growth factor are activated by conditions that increase phosphorylation of eIF2␣ such as amino acid starvation, cell differentiation, or ER stress (28, 29). We demonstrated here that the translation of inhibitor of apoptosis protein HIAP2 is enhanced in response to ER stress and is mediated by an inducible IRES element. Interestingly, inhibition of phosphorylation of eIF2␣ had no effect on the activation of HIAP2 in thapsigargin- or tunicamycin-treated cells (data now shown), suggesting that HIAP2 IRES is refractory to the translational control mediated by eIF2␣. In contrast, HIAP2 IRES is activated in a caspase-dependent manner by a proteolytic cleavage fragment of p97/DAP5/NAT, p86. Interestingly, other apoptotic triggers such as serum starvation, hypoxia, etoposide, or tumor necrosis factor ␣ did not induce levels of endogenous HIAP2 protein nor did they enhance translation of HIAP2 IRES-driven
reporter construct (Ref. 12 and data not shown). These data suggest that in addition to caspase activation, an additional ER stress-specific event is required for the induction of HIAP2. Alternatively, the cleavage of p97 has to occur by ER-specific caspase to activate HIAP2 IRES. These possibilities are currently under investigation in our laboratory. Unfolded protein response evolved as a cellular-protecting mechanism. Both the reduction of global protein synthesis and the transcriptional induction of chaperones, the two arms of UPR, serve to relieve the cellular stress. Translation of HIAP2 is up-regulated during ER stress induced by thapsigargin or tunicamycin. In this respect HIAP2 could be considered a stress response gene, like ATF4 or BiP, which are classic UPR targets. Similarly to HIAP2, translation of BiP is also mediated by an IRES element (30). In contrast to HIAP2, however, the IRES of BiP is not activated after ER stress (28), suggesting that although both BiP and HIAP2 are translated by IRES, they are regulated by distinct molecular mechanisms, likely reflecting their diverse roles in UPR. Of note, induction of HIAP2 after ER stress precedes induction of BiP. The increased levels of BiP in UPR are thought to facilitate correct protein folding and maintain ER function (31). Although it was suggested that BiP could form a complex with caspases-7 and -12 and, thus, prevent their activation (32), caspase inhibition is not a primary function of BiP. In contrast, HIAP2 is a bona fide caspase inhibitor that blocks apoptosis induced by a variety of triggers (9, 10). The up-regulation of HIAP2 during ER stress could, therefore, serve to delay the progression of caspase cascade, attenuating apoptosis and allowing the execution of UPR. Indeed, we found that overexpression of HIAP2 attenuated tunicamycin-induced cell death, and conversely, antisense-mediated down-regulation of HIAP2 enhanced cell sensitivity to tunicamycin. The requirement of caspase activation for the up-regulation of HIAP2 may seem paradoxical since the activation of caspases indicates final stages of cellular death. However, the initial activation of caspases triggered by ER stress could be subsequently blocked by increased levels of HIAP2 and delay further caspase processing. Indeed, we observed that the processing of p97 and induction of HIAP2 occurred 6 h post-tunicamycin treatment, whereas caspase-3 cleavage occurred later, indicating that both events take place at an early phase of ER stress response. A similar situation was reported for granzyme B-induced apoptosis, where caspase-3, which is directly activated by granzyme-B, is subsequently kept inactive by another member of the IAP family, XIAP (33). It is interesting to note that XIAP is also translated by an IRES mechanism. Importantly, however, the levels of XIAP are not increased in response to ER stress. Furthermore, XIAP IRES is not activated by the p86 fragment of p97/DAP5/NAT1, indicating that the regulation of cellular antiapoptotic response to different triggers is executed by distinct molecular mechanisms. Acknowledgments—We thank the members of our laboratory for stimulating discussions and critical reading of this manuscript. We are grateful to Simon Morley for the gift of ␣-eIF4G antibody. REFERENCES 1. Ron, D. (2002) J. Clin. Invest. 110, 1383–1388 2. Nakagawa, T., Zhu, H., Morishima, N., Li, E., Xu, J., Yankner, B. A., and Yuan, J. (2000) Nature 403, 98 –103 3. Morishima, N., Nakanishi, K., Takenouchi, H., Shibata, T., and Yasuhiko, Y. (2002) J. Biol. Chem. 277, 34287–34294 4. Fischer, H., Koenig, U., Eckhart, L., and Tschachler, E. (2002) Biochem. Biophys. Res. Commun. 293, 722–726 5. Holcik, M., Gibson, H., and Korneluk, R. G. (2001) Apoptosis 6, 253–261 6. Holcik, M., and Korneluk, R. G. (2001) Nat. Rev. Mol. Cell Biol. 2, 550 –556 7. Salvesen, G. S., and Duckett, C. S. (2002) Nat. Rev. Mol. Cell Biol. 3, 401– 410 8. Rothe, M., Pan, M. G., Henzel, W. J., Ayres, T. M., and Goeddel, D. V. (1995) Cell 83, 1243–1252 9. Deveraux, Q. L., Roy, N., Stennicke, H. R., Van Arsdale, T., Zhou, Q., Srinivasula, S. M., Alnemri, E. S., Salvesen, G. S., and Reed, J. C. (1998) EMBO
Induction of HIAP2 during ER Stress J. 17, 2215–2223 10. Roy, N., Deveraux, Q. L., Takahashi, R., Salvesen, G. S., and Reed, J. C. (1997) EMBO J. 16, 6914 – 6925 11. Li, X., Yang, Y., and Ashwell, J. D. (2002) Nature 416, 345–347 12. Warnakulasuriyarachchi, D., Ungureanu, N. H., and Holcik, M. (2003) Cell Death Differ. 10, 899 –904 13. Holcik, M., Lefebvre, C. A., Yeh, C., Chow, T., and Korneluk, R. G. (1999) Nat. Cell Biol. 1, 190 –192 14. Nevins, T. A., Harder, Z. M., Korneluk, R. G., and Holcik, M. (2003) J. Biol. Chem. 278, 3572–3579 15. Liston, P., Roy, N., Tamai, K., Lefebvre, C., Baird, S., Cherton-Horvat, G., Farahani, R., McLean, M., Ikeda, J. E., MacKenzie, A., and Korneluk, R. G. (1996) Nature 379, 349 –353 16. Holcik, M., Yeh, C., Korneluk, R. G., and Chow, T. (2000) Oncogene 19, 4174 – 4177 17. Li, J., Feng, Q., Kim, J. M., Schneiderman, D., Liston, P., Li, M., Vanderhyden, B., Faught, W., Fung, M. F., Senterman, M., Korneluk, R. G., and Tsang, B. K. (2001) Endocrinology 142, 370 –380 18. Bushell, M., Poncet, D., Marissen, W. E., Flotow, H., Lloyd, R. E., Clemens, M. J., and Morley, S. J. (2000) Cell Death Differ. 7, 628 – 636 19. MacGregor, G. R., Nolan, G. P., Fiering, S., Roederer, M., and Herzenberg, L. A. (1991) in Methods in Molecular Biology (Murray, E. J., and Walker, J. M., eds) Vol. 7, pp. 217–235, Humana Press Inc., Clifton, NJ 20. Huez, I., Creancier, L., Audigier, S., Gensac, M. C., Prats, A. C., and Prats, H. (1998) Mol. Cell. Biol. 18, 6178 – 6190 21. Holcik, M., Gordon, B. W., and Korneluk, R. G. (2003) Mol. Cell. Biol. 23,
17157
280 –288 22. Hellen, C. U., and Sarnow, P. (2001) Genes Dev. 15, 1593–1612 23. Clemens, M. J., Bushell, M., Jeffrey, I. W., Pain, V. M., and Morley, S. J. (2000) Cell Death Differ. 7, 603– 615 24. Henis-Korenblit, S., Shani, G., Sines, T., Marash, L., Shohat, G., and Kimchi, A. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, 5400 –5405 25. Holcik, M., Sonenberg, N., and Korneluk, R. G. (2000) Trends Genet. 16, 469 – 473 26. Henis-Korenblit, S., Strumpf, N. L., Goldstaub, D., and Kimchi, A. (2000) Mol. Cell. Biol. 20, 496 –506 27. Jackson, R. J. (2000) in Translational Control of Gene Expression (Sonenberg, N., Hershey, J. W. B., and Mathews, M. B., eds) pp. 127–183, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY 28. Fernandez, J., Yaman, I., Sarnow, P., Snider, M. D., and Hatzoglou, M. (2002) J. Biol. Chem. 277, 19198 –19205 29. Gerlitz, G., Jagus, R., and Elroy-Stein, O. (2002) Eur. J. Biochem. 269, 2810 –2819 30. Macejak, D. G., and Sarnow, P. (1991) Nature 353, 90 –94 31. Kaufman, R. J. (1999) Genes Dev. 13, 1211–1233 32. Rao, R. V., Peel, A., Logvinova, A., del Rio, G., Hermel, E., Yokota, T., Goldsmith, P. C., Ellerby, L. M., Ellerby, H. M., and Bredesen, D. E. (2002) FEBS Lett. 514, 122–128 33. Goping, I. S., Barry, M., Liston, P., Sawchuk, T., Constantinescu, G., Michalak, K. M., Shostak, I., Roberts, D. L., Hunter, A. M., Korneluk, R., and Bleackley, R. C. (2003) Immunity 18, 355–365