Importin β1 Protein-mediated Nuclear Localization of Death Receptor 5 (DR5) Limits DR5/Tumor Necrosis Factor (TNF)-related Apoptosis-inducing Ligand (TRAIL)-induced Cell Death of Human Tumor Cells*

Background: Nuclear localization of DR5 was observed in TRAIL-resistant tumor cells in human. Results: TRAIL-resistant tumor cells were sensitized to TRAIL by knockdown of importin β1. Conclusion: Importin β1-mediated nuclear translocation of DR5 limits DR5/TRAIL-induced cell death of human tumor cells. Significance: This provides a novel strategy to improve the efficiency of recombinant TRAIL and anti-DR5 antibodies in cancer therapy. Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)/death receptor 5 (DR5)-mediated cell death plays an important role in the elimination of tumor cells and transformed cells. Recently, recombinant TRAIL and agonistic anti-DR5 monoclonal antibodies have been developed and applied to cancer therapy. However, depending on the type of cancer, the sensitivity to TRAIL has been reportedly different, and some tumor cells are resistant to TRAIL-mediated apoptosis. Using confocal microscopy, we found that large amounts of DR5 were localized in the nucleus in HeLa and HepG2 cells. Moreover, these tumor cells were resistant to TRAIL, whereas DU145 cells, which do not have nuclear DR5, were highly sensitive to TRAIL. By means of immunoprecipitation and Western blot analysis, we found that DR5 and importin β1 were physically associated, suggesting that the nuclear DR5 was transported through the nuclear import pathway mediated by importin β1. Two functional nuclear localization signals were identified in DR5, the mutation of which abrogated the nuclear localization of DR5 in HeLa cells. Moreover, the nuclear transport of DR5 was also prevented by the knockdown of importin β1 using siRNA, resulting in the up-regulation of DR5 expression on the cell surface and an increased sensitivity of HeLa and HepG2 cells to TRAIL. Taken together, our findings suggest that the importin β1-mediated nuclear localization of DR5 limits the DR5/TRAIL-induced cell death of human tumor cells and thus can be a novel target to improve cancer therapy with recombinant TRAIL and anti-DR5 antibodies.


Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL)/death receptor 5 (DR5)-mediated cell death plays an important role in the elimination of tumor cells and transformed cells. Recently, recombinant TRAIL and agonistic anti-DR5 monoclonal antibodies have been developed and applied to
cancer therapy. However, depending on the type of cancer, the sensitivity to TRAIL has been reportedly different, and some tumor cells are resistant to TRAIL-mediated apoptosis. Using confocal microscopy, we found that large amounts of DR5 were localized in the nucleus in HeLa and HepG2 cells. Moreover, these tumor cells were resistant to TRAIL, whereas DU145 cells, which do not have nuclear DR5, were highly sensitive to TRAIL. By means of immunoprecipitation and Western blot analysis, we found that DR5 and importin ␤1 were physically associated, suggesting that the nuclear DR5 was transported through the nuclear import pathway mediated by importin ␤1. Two functional nuclear localization signals were identified in DR5, the mutation of which abrogated the nuclear localization of DR5 in HeLa cells. Moreover, the nuclear transport of DR5 was also prevented by the knockdown of importin ␤1 using siRNA, resulting in the up-regulation of DR5 expression on the cell surface and an increased sensitivity of HeLa and HepG2 cells to TRAIL. Taken together, our findings suggest that the importin ␤1-mediated nuclear localization of DR5 limits the DR5/ TRAIL-induced cell death of human tumor cells and thus can be a novel target to improve cancer therapy with recombinant TRAIL and anti-DR5 antibodies.
Tumor necrosis factor (TNF)-related apoptosis-inducing ligand (TRAIL), 2 a member of the TNF family, is a type II transmembrane protein expressed on the cell surface of natural killer cells, T lymphocytes, macrophages, and dendritic cells, and its mRNA is induced by type I interferons (1)(2)(3)(4). Death receptor 5 (DR5), one of the TRAIL receptors and a member of the TNF receptor superfamily, is a 48-kDa type I transmembrane protein consisting of 440 amino acids and has a death domain (DD) in its cytoplasmic region like death receptor 4 (DR4), Fas, and TNF receptor 1 (1,(5)(6)(7)(8)(9). Similar to Fas and DR4, DR5 is generally expressed in both normal and malignant cells (5). TRAIL triggers apoptosis through DR4 and/or DR5 in humans or through DR5 in mice and is an important molecule in tumor suppression through its ability to induce apoptosis in tumor cells and transformed cells without obvious damage to normal cells (1,4,6). The TRAIL-induced apoptosis signaling pathway has been extensively characterized (1,6,7). The ligation of homotrimeric TRAIL to DR5 results in the trimerization of DR5, the clustering of the DR5 intracellular DD, and the recruitment of Fas-associated death domain (FADD), leading to the formation of death-inducing signaling complex (1). Death-inducing signaling complex moves into a membrane compartment enriched in lipid rafts and linked to the cytoskeleton, therefore initiating signal transduction through intracellular signaling machinery involving the activation of caspases (1,2). Procaspase-8 is dimerized by its interaction with Fasassociated death domain, and it is activated through interchain cleavage at two cleavage sites. Furthermore, the activated caspase-8 cleaves the procaspase-3 dimer at one cleavage site, and finally, the now active caspase-3 dimer executes apoptosis (10). In some cell types, when the activation of caspase-8 is not sufficient, amplification is required and involves the activation of other caspases. In the current model, the intracellular signal transduction is regulated by complex mechanisms, and the resistance to TRAIL-induced cell death can occur at several points in the signal pathways (1,11). However, in the course of inducing cell death in tumor cells, DR5 expression on the plasma membrane is required because the TRAIL/DR5-mediated cell death is initiated by the ligation of DR5 with TRAIL on the cell surface.
TRAIL and DR5 have a great potential for cancer therapy. For instance, we have reported potent antitumor effects of an agonistic mAb against DR5 for the eradication of solid tumors in mice (12)(13)(14)(15). Recently, clinical trials with recombinant TRAIL (dulanermin) and an agonistic anti-DR5 mAb (lexatumumab) have been initiated against solid tumors, hematological tumors, non-Hodgkin lymphoma, and non-small cell lung cancer (7,16). Some strategies to increase the cell surface expression of DR5 on tumor cells are expected to improve the efficacy of these agents.
In this study, we found that large amounts of DR5 were localized in the nucleus in some tumor cell lines. We identified two functional nuclear localization signal (NLS) sequences in DR5 recognized by importin (17,18), the mutation of which abrogated the nuclear import of DR5. Moreover, the knockdown of importin ␤1 resulted in the accumulation of DR5 on the cell surface and increased TRAIL/DR5-mediated apoptosis in tumor cells. These findings may be translatable to the TRAIL and anti-DR5 mAb therapies by the tumor-specific targeting of importin ␤1 using siRNA or a small molecule inhibitor of importin-mediated nuclear transport.
Cell Culture and Transfection of siRNA-A human hepatocellular carcinoma cell line (HepG2), a human cervical adenocarcinoma cell line (HeLa), and a human prostate carcinoma cell line (DU145) were obtained from ATCC and maintained in complete Dulbecco's modified Eagle's medium (Nissui Pharmaceutical) supplemented with 10% heat-inactivated fetal bovine serum, 2 mM L-glutamine, and 50 M 2-mercaptoethanol. For importin ␤1 knockdown, 30 nM importin ␤1 siRNA or control siRNA was transfected into HeLa cells using Lipofectamine RNAiMAX (Invitrogen) according to the manufacturer's instructions. Cells were incubated for 24 -48 h at 37°C and then collected for subsequent analysis.
Measurement of TRAIL Sensitivity-Cells (1 ϫ 10 4 in a volume of 100 l of culture medium/well) were seeded into a flat bottomed 96-well plate (Corning) and incubated with various concentrations of rTRAIL for 24 h at 37°C. Cell viability was determined by the 2-(4-iodophenyl)-3-(4-nitrophenyl)-5-(2,4disulfophenyl)[ 2 H]tetrazolium monosodium salt (WST) assay using Cell Counting Kit-8 (Dojindo) according to the manufacturer's instructions, and the percentage of cell death was calculated as described previously (19). To measure TRAIL sensitivity after importin ␤1 knockdown, rTRAIL or vehicle alone was added to the culture medium at 24 h post-transfection with siRNA.
Flow Cytometry-To examine the cell surface expression of DR5, 2 ϫ 10 5 cells were incubated with 0.5 g of mouse antihuman DR5 mAb or isotype-matched mouse IgG1 for 30 min at 4°C followed by PE-conjugated goat anti-mouse IgG F(abЈ) 2 . To detect the total cellular (namely cell surface and intracellular) expression of DR5, cells were treated with 70% ethanol in PBS for 30 min on ice and washed with PBS before the staining. The cells were then analyzed on a FACScan TM (BD Biosciences), and the data were processed using the Cell Quest program. The net mean fluorescence intensity was calculated as described previously (20).
Immunostaining and Confocal Microscopy-After removal of the culture medium, tumor cells incubated for 24 -48 h at 37°C on a poly-L-lysine-coated 4-well chamber slide (Nalge Nunc) were rinsed in PBS and fixed with 8% paraformaldehyde in 100 mM phosphate buffer for 30 min at 4°C. After the removal of the sidewall from the chamber slide, each chamber was circled with a Dako pen, and the cells were treated with a permeabilization buffer (Takara) according to manufacturer's instructions. The cells were incubated with appropriate concentrations of the primary antibodies or the isotype-matched control Igs overnight at 4°C for the detection of DR5 or for 1 h at 37°C for the detection of other proteins. Importin ␤1 was detected using Alexa Fluor 488-conjugated goat anti-rabbit IgG, whereas DR5 was identified by staining with biotin-conjugated goat anti-mouse IgG and Alexa Fluor 594-conjugated streptavidin. Endogenous biotin was blocked using an avidinbiotin blocking kit (Vector Laboratories) according to the manufacturer's specifications. Cell nuclei were counterstained with Vectashield mounting medium for fluorescence with DAPI (Vector Laboratories) and viewed with a confocal microscope (LSM510, Carl Zeiss).
Cell Fractionation and Western Blot Analysis-Cells were lysed in NE-PER nuclear and cytoplasmic extraction reagents (Thermo Scientific) following the manufacturer's instructions. To prepare total cell lysate for the detection of importin ␤1, sample buffer containing SDS was added directly to the cells, which were sonicated briefly. For the detection of caspases, the cells were lysed in radioimmune precipitation assay buffer (50 mM Tris-Cl (pH 8.0), 150 mM NaCl, 1% Nonidet P-40, 0.5% deoxycholate, 0.1% SDS, 25 mM ␤-glycerophosphate, 1 mM sodium orthovanadate, 1 mM sodium fluoride, 1 mM PMSF, 1 mg/ml aprotinin, and 1 mg/ml leupeptin). The cytoplasmic, nuclear, and total lysate fractions of the cells were subjected to 7.5, 10, or 16% SDS-PAGE under reducing conditions and transferred onto PVDF membrane (Millipore). The membranes were analyzed by immunoblotting with appropriate primary antibodies followed by HRP-conjugated secondary antibodies. The membranes were developed using an enhanced chemiluminescence (ECL) Plus Western blotting detection kit (GE Healthcare) or Super Signal West Dura Extended Duration Substrate (Thermo Scientific) according to the manufacturers' instructions and were analyzed by LAS4000 (GE Healthcare). Immunoprecipitation and Western blotting were performed as described previously (21) on samples prepared from 5 ϫ 10 7 cells. For the detection of DR5, SDS-PAGE and Western blotting were performed in non-reducing conditions, and the membranes were probed using biotin-conjugated goat anti-human DR5 pAb followed by HRP-conjugated streptavidin.
Plasmid Construction and Transfection-The pEGFP-N1 vector was obtained from Clontech. The cDNA encoding the DD-deleted DR5 containing the wild-type NLS sequences was amplified by PCR using pMKITNeo-DR5 (3) as a template with the following primers: sense, 5Ј-AATCTCGAGATGGAACA-ACGGGGACAG-3Ј; and antisense, 5Ј-CGGAATTCGGGAG-TCAAAGGGCACCAAGT-3Ј. This cDNA was designated wild-type NLS (WT-NLS). To eliminate the putative NLS in the DR5 sequence, R 15 K 16 R 17 was mutated to AAA by using the oligonucleotide sequences 5Ј-TCGGGGGCCGCGGCAGCG-CACGGCCCA-3Ј and its reverse strand. This mutant was designated M1-NLS. A second NLS R 322 R 323 R 324 sequence was mutated to AAA by using the oligonucleotide sequences 5Ј-AGGTCTCAGGCGGCGGCGCTGCTGGTTCCA-3Ј and its reverse strand. This mutant was designated M2-NLS. PCR products were subcloned into the XhoI and EcoRI sites of the pEGFP-N1 vector, and the constructs were verified by DNA sequencing with an ABI PRISM 3100 sequence detection system (Applied Biosystems). One day before transfection, 2 ϫ 10 5 cells were seeded on a 6-well plate and then transfected with 4 g of plasmid DNA using Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. Following transfection, the cells were cultured in selection medium containing 2 mg/ml G418, and GFP-positive cells were sorted using a FACSCalibur TM (BD Biosciences).

Intracellular Localization of DR5 in Human Tumor Cells-
The detection of DR5 in tumor cell lines was achieved using flow cytometry. As shown in Fig. 1A, HepG2 and HeLa cells expressed a moderate level of DR5, whereas DU145 cells expressed a high level of DR5 on their surface. After treatment with 70% ethanol, which permeabilized the plasma membrane and the nuclear membrane, an increase in DR5 was detected in HepG2 and HeLa cells but not in DU145 cells, suggesting that HepG2 and HeLa cells expressed intracellular DR5. To determine the intracellular localization of DR5 in these cells, we performed multicolor staining with anti-DR5 mAb, DAPI for nuclear staining, and WGA for staining the Golgi apparatus (22). As observed by confocal microscopy (Fig. 1B), a large amount of DR5 was detected in the nuclei of both HepG2 and HeLa cells. In contrast, the DR5 in DU145 cells was mostly localized to the plasma membrane and the cytoplasm but was barely detectable in the nucleus (Fig. 1B). To confirm the presence of DR5 in the nucleus of HeLa cells, we performed Western blot analysis of cytoplasmic and nuclear fractions extracted from HeLa cells. As shown in Fig. 1C, DR5 was clearly detected in both the cytoplasmic and nuclear fractions of HeLa cells. DR4 is another death receptor that is engaged by TRAIL like DR5. We also investigated the expression and localization of DR4 in tumor cell lines. As shown in supplemental Fig. 1A, DU145 cells expressed a significant level of DR4 on the cell surface, but HepG2 and HeLa cells expressed only a low level of DR4. After the permeabilization of the plasma membrane and the nuclear membrane, the DR4 levels did not increase in any of the cell lines. Using confocal microscopy, DR4 was barely detectable on the cell surface or in the cytoplasm of HepG2 and HeLa cells, whereas DR4 was localized on the cell surface and in the cytoplasm of DU145 cells (supplemental Fig. 2B). Nuclear DR4 was not detectable in any of the cell lines.
Correlation between TRAIL Resistance and Nuclear Localization of DR5 in Tumor Cells-We next investigated the sensitivity of these cell lines to TRAIL. As shown in Fig. 2A, DU145 cells were highly sensitive to TRAIL, and 25 ng/ml rTRAIL induced 80% cell death in 24 h. In contrast, the same doses of rTRAIL had no effect on cell death in HepG2 and HeLa cells. We confirmed that concentrations up to 625 ng/ml rTRAIL did not cause significant cell death in these cells even after 48 h (data not shown). These results suggest that tumor cells containing nuclear DR5 are resistant to TRAIL. To further substantiate the TRAIL-induced apoptosis, we performed Western blot analysis to estimate the cleavage of caspase-8 and caspase-3, which are required for the generation of active initiator and executioner caspases that induce apoptosis. As shown in Fig. 2B, in DU145 cells, procaspase-8 was completely cleaved, a significant amount of procaspase-3 was cleaved following a 2-h incubation with 125 ng/ml rTRAIL, and new bands corresponding to cleaved caspase-8 and caspase-3 were detected. After a 24-h incubation with rTRAIL, procaspase-8 was cleaved completely, and only a small amount of procaspase-3 remained. In contrast to DU145 cells, only small amounts of cleaved procaspase-8 and cleaved procaspase-3 were detected in HeLa cells that were incubated with rTRAIL for 2 h. After a 24-h incubation, there was no difference in the expression levels of procaspase-8 and procaspase-3 in the presence or absence of rTRAIL in HeLa cells. Collectively, these data indicate that tumor cells containing nuclear DR5 are resistant to TRAIL-induced apoptosis, and tumor cells having no nuclear DR5 are responsive to TRAILinduced apoptosis.
Association of DR5 with Importin ␤1-The transport of proteins into the nucleus is generally mediated by importins via the recognition of NLS sequences. Analysis of the primary structure of human DR5 revealed two NLS-like sequences, RKR starting at amino acid 15 and RRR starting at amino acid 322 (Fig. 3A, hDR5). Therefore, we investigated the association of DR5 with importin ␤1, which is essential for the transport of various proteins into the nucleus regardless of the presence of importin ␣ (18). Two-color immunostaining for DR5 and importin ␤1 showed that DR5 colocalized with importin ␤1 in the nucleus and in the perinuclear region of the cytoplasm in HeLa cells (Fig. 3B). We next performed Western blot analysis to detect importin ␤1 in anti-DR5 immunoprecipitates from HeLa cells and DU145 cells. As shown in Fig. 3C, importin ␤1 was co-precipitated with DR5 from the HeLa cell lysate but not from the DU145 cell lysate. We also detected DR5 in anti-importin ␤1 immunoprecipitates from the HeLa cell lysates (Fig.  3D). The results suggest that DR5 was transported into the nucleus through its physical interaction with importin ␤1 in HeLa cells.
Identification of NLS within DR5-To identify the functional NLS in human DR5, we generated two NLS mutants, M1-NLS and M2-NLS, by replacing basic amino acids with AAA (Fig.   3A). In these mutants and in the WT-NLS constructs, the death domain was replaced with GFP to avoid selective death of the transfected cells and to monitor the expression of the DR5-GFP fusion proteins directly (Fig. 3A). After transfection of the constructs in pEGFP-N1 vector into HeLa cells, stable transfectants were generated. As estimated by cell surface staining of intact cells with anti-DR5 mAb, the cell surface expression levels of M1-NLS and M2-NLS correlated with the expression level of the introduced DR5-GFP (Fig. 3E). In contrast, cell surface DR5 levels did not increase in the majority of the cells that expressed DR5-GFP with WT-NLS (Fig. 3, E and F). When compared with intact cells, the permeabilization of the plasma membrane and nuclear membrane with 70% ethanol increased anti-DR5 mAb staining in the WT-NLS cells but not in the M1-NLS or M2-NLS cells (Fig. 3, F and G). Moreover, confocal microscopy showed that the expression of DR5-GFP with WT-NLS was mainly localized in the nucleus, but the expression of DR5-GFP with the mutant M1-NLS or M2-NLS was predominantly localized in the plasma membrane and the perinuclear region corresponding to the ER and Golgi apparatus (Fig. 3H). These results indicate that both of the NLS-like sequences in human DR5 are required for the nuclear import.
Knockdown of Importin ␤1 Abrogates Nuclear Transport of DR5-We next examined whether the localization of DR5 was affected by the knockdown of importin ␤1. When HeLa cells were transfected with importin ␤1 siRNA, importin ␤1 protein levels decreased substantially as estimated by Western blotting (Fig. 4A). Compared with the control siRNA-treated cells, which showed a prominent nuclear localization of DR5, confocal microscopy of the importin ␤1 siRNA-treated cells demonstrated the distinct localization of DR5 in the WGA ϩ Golgi apparatus, the cytoplasm, and at the cell surface but not in the nucleus (Fig. 4B). Similar results were obtained with the translocation to STAT6, which is translocated from the cytoplasm to the nucleus through the importin ␤1 pathway after the stimulation with IL-4 (23), whereas histone H3, which is known to be imported to the nucleus by transportin, importin ␤, importin 5, and importin 7 (24), was localized in the nucleus following importin ␤1 knockdown (Fig. 4B). Moreover, knockdown of importin ␤1 reduced the level of DR5 protein in the nuclear fraction and increased the DR5 level in the cytoplasmic fraction as compared with the control siRNA-treated cells (Fig. 4C). We also investigated whether importin ␤1 regulates the expression and localization of DR4 as well as DR5. By flow cytometric and confocal microscopic analyses, the expression level and localization of DR4 were not affected by importin ␤1 knockdown in HepG2, HeLa, or DU145 cells (supplemental Fig. 1, C, D, and E).
The results indicate that the nuclear transport is unique to DR5 and is mediated by importin ␤1.
Surface Expression of DR5 Is Increased by Knockdown of Importin ␤1-We investigated whether the cell surface expression of DR5 increases following the knockdown of importin ␤1. Flow cytometric analysis of intact or permeabilized cells after anti-DR5 mAb staining showed that the cell surface expression levels of DR5 on intact cells increased significantly following transfection of importin ␤1 siRNA compared with control siRNA, whereas the total cellular DR5 expression levels in the permeabilized cells were not affected (Fig. 4, D and E).  DECEMBER 16, 2011 • VOLUME 286 • NUMBER 50

JOURNAL OF BIOLOGICAL CHEMISTRY 43387
Knockdown of Importin ␤1 Increases TRAIL Sensitivity of TRAIL-resistant Cells-Finally, we examined whether the knockdown of importin ␤1 increases the sensitivity of tumor cells to TRAIL. Phase-contrast microscopy showed that massive cell death was induced in the importin ␤1 siRNA-treated HeLa cells but not in the control siRNA-treated cells after a 24-h treatment with 25 ng/ml rTRAIL (Fig. 5A). WST assays showed that the sensitivity of HeLa cells to either 25 or 125 ng/ml rTRAIL increased following the transfection of importin ␤1 siRNA compared with control siRNA (Fig. 5B). We also examined the effect of importin ␤1 knockdown on TRAIL sensitivity in HepG2 and DU145 cells. As shown in supplemental Fig. 2, A and B, TRAIL sensitivity of DU145 cells was not affected by the knockdown of importin ␤1. In contrast, the TRAIL sensitivity of HepG2 cells was greatly increased by the importin ␤1 knockdown. To confirm the induction of apoptosis by TRAIL following importin ␤1 knockdown, we performed Western blot analysis to estimate cleavage of caspase-8 and caspase-3 in whole cell lysates from HeLa cells. After the knockdown of importin ␤1, a large portion of the procaspase-8 and procaspase-3 proteins was cleaved, and the cleaved forms were clearly detected following a 4-h incubation with 125 ng/ml rTRAIL (Fig. 5C). These data indicate that the TRAIL-induced caspase-8 and caspase-3 activation was enhanced by the importin ␤1 knockdown. Because DR4 expression was not affected by importin ␤1 knockdown (supplemental Fig. 1C), it is possible that the increase in TRAIL sensitivity of HeLa and HepG2 cells was caused by the augmentation of DR5-mediated apoptosis. To confirm this possibility, we finally examined whether the addition of anti-DR5 mAb inhibits the increase in TRAIL sensitivity of HeLa cells. As expected, the increased sensitivity of importin ␤1 siRNA-treated cells to TRAIL was completely abolished by the addition of an anti-DR5 blocking antibody (Fig. 5D). These data indicate that the knockdown of importin ␤1 can sensitize HeLa cells to TRAIL in a DR5-dependent manner.

DISCUSSION
In this study, we found by means of confocal microscopy that DR5 was localized in the nucleus in HeLa and HepG2 cells. Previous reports have shown that DR5 localizes predominantly in the plasma membrane and in the trans-Golgi network in human melanoma cells (25), melanocytes, endothelial cells, and fibroblasts (26). Recently, DR5 was detected in the cytoplasm in the majority of tissue samples obtained from patients suffering from non-small cell lung carcinoma, and it was detected in the nucleus in 27% of the same samples (27). Our present observation of the nuclear localization of DR5 in HepG2 and HeLa cells is consistent with these clinical observations. It has also been shown that TRAIL and DR5 are internalized by clathrin-dependent and -independent endocytosis (1,28,29), and the internalized DR5 is transported to the lysosome (30). However, our study showed that DR5 did not colocalize with cathepsin D ϩ lysosomes in HeLa cells, HepG2 cells, or DU145 cells that lack nuclear DR5 (data not shown). Moreover, the nuclear translocation of DR5 was not observed even after preincubation with rTRAIL (data not shown). Therefore, pathways other than internalization pathways might transport the nuclear DR5. Protein transport into the nucleus is generally mediated by the importin system, which recognizes the NLS in the target proteins (17,18). We identified two functional NLS sequences in DR5, the mutation of which abrogated the nuclear transport of DR5 and confirmed the physical association of DR5 with importin ␤1. One NLS is located near the N terminus, corresponding to the signal peptide, and the other NLS is located in the cytoplasmic region between the transmembrane region and the DD. Transmembrane and secreted proteins are generally translocated across the ER membrane barrier through the sequential actions of the signal recognition particle and its receptor, which is associated with translocation in the ER membrane (31). The signal peptide is a 15-20-residue hydrophobic N-terminal signal sequence that is recognized by the signal recognition particle; however, no obvious sequence homologies exist among signal peptides (32). Therefore, the signal peptidase does not recognize a specific sequence in the cleavage site, and not every transmembrane or secreted protein is necessarily cleaved by signal peptidase (e.g. fibroblast growth factor (FGF)-9 and FGF-16 (33,34)). The molecular weights of membrane-bound DR5 and nuclear DR5 were similar in our Western blot analysis, suggesting that the signal peptide sequence in DR5 is not targeted by a signal peptidase. This implies that the N-terminal NLS of DR5 is functional. In a previous study of the TNF receptor family, CD40 was localized not only in the plasma membrane and cytoplasm but also in the nucleus by the transport through the NLS-importin pathway (35). To our knowledge, our present study is the first to demonstrate the NLSmediated nuclear transport of a death receptor. Because the ER-associated degradation system removes full-length transmembrane proteins from the ER into the cytoplasm (36), DR5 synthesized in the ER membrane might be released into the cytoplasm where it is recognized by importin and transported into the nucleus. In this scenario, DR5 is not translocated to the Golgi apparatus nor is it modified with glycans. Although DR5 does not contain any potential N-linked glycosylation sites (5), the modification of death receptors by O-glycosylation is very important in cancer cells (1,37). Previously, it has been reported that the O-glycosylation of DR4 and DR5 in cancer cells modulates the sensitivity of cells to TRAIL by promoting ligand-induced receptor clustering and subsequent caspase-8 activation (37). Given that protein modification by O-linked glycosylation can occur in the nucleus (38), DR5 that is localized in the nucleus could be modified by O-glycosylation. How-ever, it is unlikely that nuclear DR5 would be transported to the plasma membrane. At present, it remains unclear why a portion of DR5 is translocated to the plasma membrane and another portion of DR5 is transported to the nucleus. Further studies are needed to determine how the localization of DR5 is regulated.
Tumor cells expressing DR5 in the nucleus, such as HeLa and HepG2 cells, exhibit a high resistance to TRAIL compared with the tumor cells that lack nuclear DR5, such as DU145 cells. The nuclear localization of DR5 is thought to be advantageous for tumor cells to escape from TRAIL/DR5-mediated cell death. Similar results were obtained in three independent experiments. C, 1 day before transfection, HeLa cells (1 ϫ 10 5 ) were seeded onto a 6-well plate and transfected with importin ␤1 siRNA or control siRNA. After a 24-h incubation, cells were treated with or without 125 ng/ml rTRAIL for 4 or 24 h. Whole cell extracts were subjected to Western blot for caspase-8, caspase-3, importin ␤1, and tubulin. D, cells were also incubated with anti-DR5 mAb or isotype-matched control mouse IgG1 for 30 min at 37°C before the addition of 125 ng/ml rTRAIL. After a 24-h incubation, cell death was analyzed by WST assay. Data are the mean Ϯ S.D. of triplicates. **, p Ͻ 0.01 by Student's t test. Similar results were obtained in three independent experiments. Therefore, one could hypothesize that the nuclear DR5 contributes to the survival of tumor cells. The knockdown of importin ␤1 resulted in a significant sensitization of HeLa and HepG2 cells to TRAIL/DR5-mediated cell death by enhancing the cell surface expression of DR5. Thus, the nuclear localization of DR5 can limit the DR5/TRAIL-induced cell death of tumor cells. Kau et al. (39) pointed out the mislocalization from the cytoplasm to the nucleus of some proteins in cancers (e.g. nuclear factor-B and ␤-catenin) and reported that protein mislocalization resulted in cancer cell intervention. Our observations correlate with these in vivo findings. Moreover, it has been shown that nuclear CD40 binds to and stimulates the B lymphocyte stimulator/B cell activation factor promoter and regulates the growth and survival of B lymphocytes (35). In this context, future studies will determine whether nuclear DR5 has a similar function in tumor cell survival. The regulation of sensitivity and resistance to TRAIL-induced cell death has been investigated in many reports. For example, it has been suggested that the relative expression of death receptors and decoy receptors that lack a DD may regulate TRAIL sensitivity (40,41). Other factors such as TRAIL-induced nuclear factor-B activation and death inhibitors including cellular FADD-like interleukin-1␤ converting enzyme inhibitory protein and X-linked inhibitor of apoptosis protein have also been implicated in the differential sensitivity to TRAIL (1,2,11,42). On the other hand, many cancer therapeutic drugs have been shown to enhance the TRAIL sensitivity of tumor cells by upregulating DR4 or DR5 expression (6,7). We have demonstrated that the knockdown of importin ␤1 efficiently sensitized TRAIL-resistant tumor cells to rTRAIL by increasing the cell surface expression of DR5. siRNA and shRNA have been reported to be effective in silencing the gene in vivo, and phase I-III clinical studies have already been initiated against various diseases including cancers (43). Therefore, tumor-specific delivery of importin ␤1 siRNA or a recently identified small molecule inhibitor of importin ␤-mediated nuclear import (44) may be useful in applications that will improve the antitumor effects of recombinant TRAIL (dulanermin) and anti-DR5 mAb (lexatumumab) for the therapeutic treatment of human cancers.