Antisense Suppression of the Chloride Intracellular Channel Family Induces Apoptosis, Enhances Tumor Necrosis Factor A-Induced Apoptosis, and Inhibits Tumor Growth

Research Article

Antisense Suppression of the Chloride Intracellular Channel Family Induces Apoptosis, Enhances Tumor Necrosis Factor a-Induced Apoptosis, and Inhibits Tumor Growth

Kwang S. Suh, Michihiro Mutoh, Michael Gerdes, John M. Crutchley, Tomoko Mutoh, Lindsay E. Edwards, Rebecca A. Dumont, Pooja Sodha, Christina Cheng, Adam Glick, and Stuart H. Yuspa

Laboratory of Cellular Carcinogenesis and Tumor Promotion, National Cancer Institute, Bethesda, Maryland

Abstract CLICs are also found in a soluble form in the cytoplasm (2–5). mtCLIC/CLIC4 is a p53 and tumor necrosis factor a (TNFa) Crystallographic analysis of the structure of soluble CLIC1 regulated intracellular chloride channel protein that local- indicates homology to the glutathione transferase family of izes to cytoplasm and organelles and induces apoptosis proteins. It is hypothesized that soluble CLICs may become when overexpressed in several cell types of mouse and activated as anion channels or channel regulators when ‘‘auto- humanorigin.CLIC4iselevatedduringTNFa-induced inserted’’ into intracellular membranes (6). apoptosis in human osteosarcoma cell lines. In contrast, Among the CLIC family proteins, the biological functions of inhibition of NFKB results in an increase in TNFa-mediated CLIC4 have been most thoroughly studied. CLIC4 is expressed in apoptosis with a decrease in CLIC4 protein levels. Cell lines many cell types. In skin keratinocytes, CLIC4 was first localized to expressing an inducible CLIC4-antisense construct that also mitochondria and cytoplasm and later was localized specifically reduces the expression of several other chloride intracellular to the inner mitochondrial membrane by immunogold electron channel (CLIC) family proteins were established in the microscopy (7, 8). Other reports have localized CLIC4 in the trans- human osteosarcoma lines SaOS and U2OS cells and Golgi network in pancreatic cells, endoplasmic reticulum in rat a malignant derivative of the mouse squamous papilloma hippocampal HT-4 cells, and large dense core vesicles in line SP1. Reduction of CLIC family proteins by antisense neurosecretory cells (9–11). CLIC4 has also been associated with expression caused apoptosis in these cells. Moreover, CLIC4- the actin cytoskeleton in membrane ruffles and lamellipodia. À antisense induction increased TNFa-mediated apoptosis Electrophysiologic analysis suggests that CLIC4 has Cl selective in both the SaOS and U2OS derivative cell lines without channel activity (4, 10, 12, 13). CLIC4 is highly conserved in altering TNFa-induced NFKB activity. Reducing CLIC proteins different species with nearly 95% identity in amino acid sequence in tumor grafts of SP1 cells expressing a tetracycline-regulated indicating an important functional role in cellular physiology (8). CLIC4-antisense substantially inhibited tumor growth and CLIC4 associates with dynamin I, actin, tubulin, and 14:3:3 isoforms induced tumor apoptosis. Administration of TNFa i.p. mod- in neuronal cells, suggesting it may also play a role in cell signaling estly enhanced the antitumor effect of CLIC reduction (14). This is consistent with the recently reported induction of in vivo. These results suggest that CLIC proteins could serve CLIC4 in transforming growth factor-h and serum-activated as drug targets for cancer therapy, and reduction of CLIC human breast fibroblasts, where it was associated with trans- proteins could enhance the activity of other anticancer differentiation to myofibroblasts (15). drugs. (Cancer Res 2005; 65(2): 562-71) CLIC4 is a direct response gene for p53 transactivation, and the up-regulation of CLIC4 is strongly associated with p53-mediated Introduction apoptosis (7). CLIC4 overexpression induces apoptosis character- ized by changes in the intrinsic mitochondrial apoptotic pathway Chloride intracellular channel (CLIC) family of proteins (p64, CLIC1-5, and parchorin) is frequently localized to intracellular such as loss of mitochondrial membrane potential, cytochrome c organelles in multiple cell types. The putative chloride ion gating release, and caspase activation (7). CLIC4 also translocates to the activity of some members of this family suggests that CLIC nucleus in cells induced to undergo apoptosis by a variety of proteins function to regulate organellar volume, ionic homeosta- stress inducers (16), and nuclear-targeted CLIC4 is strongly sis, and pH (1). CLIC1 to CLIC5 are similar in size and highly proapoptotic even when the mitochondrial death pathway is inhibited by genetic deletion of Apaf1 (16). Exposure to tumor homologous, whereas p64 and parchorin have distinct NH2- terminal domains but share strong sequence similarity to the necrosis factor a (TNFa) also increases CLIC4 transcripts and other family members in the COOH terminus (CLIC module; protein and causes CLIC4 to translocate to the nucleus ref. 2). Common to all members is a hydrophobic region in the independent of p53 (8, 16). CLIC module consistent with a transmembrane domain, although TNFa induces apoptosis in some cell types and is in clinical trials for the treatment of certain cancers (17). The interaction of TNFa with its receptor can activate a death pathway through caspase-8 and caspase-3 leading to a cytochrome c–independent Note: K.S. Suh and M. Mutoh contributed equally to this work. apoptotic response (18). However, TNFa can simultaneously Requests for reprints: Stuart H. Yuspa, Cell Carcinogenesis and Tumor induce an antiapoptotic response through its activation of the Promotion, Room 3B25, MSC 4255, 37 Convent Drive, Bethesda, MD 20892-4255. K Phone: 301-496-2162; Fax: 301-496-8709; E-mail: [email protected]. downstream transcription factor NF B and subsequent induction I2005 American Association for Cancer Research. of inhibitors of apoptosis and other NFKB response genes to

Cancer Res 2005; 65: (2). January 15, 2005 562 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 2005 American Association for Cancer Research. CLIC4 Antisense Induces Apoptosis and Inhibits Tumor Growth blunt the apoptotic response (19). In experimental settings, this proapoptotic response. We considered this an important under- can be overcome by inhibiting the TNFa-mediated nuclear taking because CLIC4 could be a collateral target in biological translocation of NFKB using the mutant form of the NFKB approaches to cancer therapy with TNFa. cytoplasmic binding partner IKB (20, 21). The mutant IKB (IKBsr) cannot be phosphorylated and degraded and thus does not dissociate from NFKB to allow nuclear translocation and Materials and Methods DNA binding. Whereas this has been an effective tool to Cell Culture. Tet-On U2OS cell lines were purchased from Clontech understand the antiapoptotic activity of NFKB, this antiapoptotic (Palo Alto, CA). Both Tet-On U2OS, p53 Tet-On SaOS cell line (26), and their pathway could compromise the clinical effectiveness of TNFa as derivatives were maintained in DMEM/10% fetal bovine serum. SP1 an antitumor agent (22). keratinocytes and its derivatives were maintained as described previously The death receptor pathway together with inhibition of NFKB (7). Recombinant human and murine TNFa were obtained from is considered the major route through which TNFa induces Calbiochem (San Diego, CA). apoptosis in experimental settings, but other pathways, such as Immunoblot Analysis. Cells were lysed into 100 AL M-Per (Pierce, p53 and mitogen-activated protein kinases, have also been Rockford, IL), and 30 Ag of proteins were resolved by SDS-PAGE and implicated in TNFa-mediated apoptosis (23). These pathways transferred onto polyvinylidene difluoride membranes (Millipore, Bedford, K may contribute to cell killing by TNFa independently of NF B MA). Antibodies against the COOH terminus of CLIC4 (8) were used at a (24, 25). Because expression and nuclear translocation of 1:10,000 dilution. The following antibodies were also used: rabbit polyclonal proapoptotic CLIC4 are induced by TNFa, we embarked on a anti-NFKB, Bax, Bcl-2, and anti-IKBa antibodies were from Santa Cruz study to determine where CLIC4 might fit into the TNFa Biotechnology (Santa Cruz, CA); anti-h-actin mouse monoclonal antibody

Figure 1. IKBsr or CLIC4-antisense expression does not inhibit TNFa-induced nuclear translocation of CLIC4. p53 Tet-On SaOS cells (A, B, and D) and Tet-On U2Os cells (A and C) were pretreated with (A) null adenovirus (N), (B) Tet-Off adenovirus (T) to induce antisense expression, (C) IKBsr adenovirus (I), or in (D) combination of IKBsr and Tet-Off adenoviruses (I+T) with 10 MOI for 17 hours, exposed to 25 ng/mL TNFa for 30 minutes (ÀTNF or +TNF), and fixed in 2% paraformaldehyde. NFKB, IKB, and CLIC4 were examined by immunofluorescence using confocal microscopy. Insets in A-D, confocal images correspond to the antibody designated in parenthesis. CLIC4 immunostaining for B and D: overexposed confocal image in order to detect CLIC4 staining when CLIC4 antisense is expressed.

www.aacrjournals.org 563 Cancer Res 2005; 65: (2). January 15, 2005

Figure 2. IKBsr enhances TNFa-mediated apoptosis and reduces CLIC4 protein independent of CLIC4 promoter activity. p53 Tet-On SaOS cells (A, C, and E) and Tet-On U2OS cells (B, D, and F) were untreated or pretreated with 50 MOI of null Ad or IKBsr adenovirus for 24 hours prior to 25 ng/mL TNFa. Annexin V fluorescence-activated cell sorting analysis was done 24 hours after TNFa (A and B). Columns, mean of three independent experiments; bars, FSE. CLIC4 protein levels were determined by Western blot analysis in p53 Tet-On SaOS cells (C) and Tet-On U2OS cells (D). Numbers between CLIC4 and actin blots (C) and (D): CLIC4 band intensity relative to the corresponding actin band as determined by densitometry using ImageMaster software. The cells were pretreated with IKBsr adenovirus for 24 hours before the treatment with TNFa for 6 hours. CLIC4 promoter transcriptional activity was determined by luciferase reporter gene assay (E and F). Columns, mean (n = 3); bars, FSD. *, P < 0.01, when compared with TNFa alone.

was from Chemicon International, Inc. (Temecula, CA); peroxidase- Generation of Bovine Keratin 5–Driven Tet-Off CLIC4-Antisense conjugated secondary antibodies were obtained from Amersham Bioscien- Squamous Papilloma-1 Cell Line (Bk5/tTA SP1). To generate the pBk5/ ces (Piscataway, NJ). Monospecific polyclonal antibodies recognizing CLIC1 tTA transactivator plasmid, pBK5 (ref. 28; obtained from David Bol, MD and CLIC5 as well as recombinant CLIC1, CLIC4, and CLIC5 proteins were Anderson, TX) was digested with SnaB1 and ligated with BamHI/EcoRI– generous gifts from Dr. Mark Berryman (College of Osteopathic Medicine, digested tTA coding sequence from the plasmid pUHD15-1 (29). Ohio University; ref. 27). Immunoblots were developed with SuperSignal Recombinant constructs with correct orientation were identified by NotI chemiluminescent substrates (Pierce). Total Image software (Amersham and SalI digestion. SP1 cells were cotransfected with 10 Ag of BK5/tTA Pharmacia, Sunnyvale, CA) was used to determine densitometry for protein DNA and 2 Ag of the Hygromycin marker plasmids using lipofectamine bands. (Invitrogen, San Diego, CA) for 6 hours. Hygromycin-resistant colonies Immunofluorescent Microscopy. Cells at the density of 6 Â 104 per were ring cloned and tested for induced regulation of a transfected Tet-On well were seeded in 60-mm dish or 8-chamber slides and incubated in luciferase construct. the presence or absence of TNFa for 30 minutes with null, IKBsr or Tet- Generation of Tetracycline-Inducible CLIC4-Antisense (Tet-On or Off adenovirus for the specified times. Immunostaining was done as Tet-Off CLIC4-Antisense) Cell Lines. The cloning of CLIC4 and the described previously (16), and antibody dilutions were based on construction of the sense CLIC4 plasmids have been described elsewhere suggestions from the manufacturers. The stained cells were detected by (8). The recombinant plasmid was digested using NotI, and the resulting Zeiss510 confocal microscope, and LSM browser was used for cropping sense and antisense orientations of CLIC4 were subcloned into the NotI images. site of the expression vector pTRE2pur Vector (Clontech). To allow NFKB-DNA Binding Activity Assay. The activity of p65 binding to tetracycline regulation, f8 Â 105 Tet-On U2OS cells and p53 Tet-On SaOS oligonucleotides containing a NFKB consensus binding site was measured cells were transfected with 0.4 Ag of CLIC4-antisense pTRE2pur plasmid by using TransAM NFKB p65 Transcription Factor Assay Kits according to using Effectene Transfection Reagent (Qiagen, Chatsworth, CA). To the manufacturer’s instructions (Active Motif, Inc., Carlsbad, CA). generate Tet-inducible CLIC4-antisense SP1 cell lines, 1 Â 107 Bk5/tTA Adenoviruses. IKBsr (generous gift from Dr. Dennis, Guttridge of the SP1 cells were transfected with 1 Ag of CLIC4-antisense pTRE2pur University of North Carolina), Tet-On and Tet-Off adenoviruses (Clontech, plasmid. Stable transfectants were selected by limiting dilution or cloning- Palo Alto, CA) were amplified in the Molecular Biology Core Facility of ring method in a medium containing 1 Ag/mL puromycin, and the National Cancer Institute-Frederick Cancer Research and Development selected clones were analyzed for inducible suppression of CLIC4 by Center. Adenoviruses were purified over two CsCl gradients, dialyzed with a immunoblotting. These p53 Tet-On SaOS cell, Tet-On U2OS, and SP1 cell TGM buffer [Tris-HCl (pH 7.5), 1 mmol/L MgCl2, and 10% glycerol], and derivatives are designated as CLIC4AS-SaOS, CLIC4AS-U2OS, and CLI- stored in aliquots at À70jC. Viral titer was determined by plaque assay, and C4AS-SP1 cells, respectively. The presence of the CLIC4-antisense region the adenoviral vector without a recombinant insert was used as a viral was confirmed by genomic PCR analysis. The CLIC4AS-SaOS cells or control (Null virus). CLIC4AS-U2OS cells were treated with doxycycline (800 ng/mL) or infected

Cancer Res 2005; 65: (2). January 15, 2005 564 www.aacrjournals.org

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 2005 American Association for Cancer Research. CLIC4 Antisense Induces Apoptosis and Inhibits Tumor Growth with Tet-Off adenovirus without doxycycline for 17 hours to induce CLIC4- This region was subcloned into pGlow-Topo (Invitrogen), cut with NheI, antisense before treatment with test agents. The Tet-Off adenovirus and subcloned into a NheI-linearized pGL3 (Promega, San Luis Obispo, CA) method was used to avoid doxycycline in experiments comparing CLIC4- vector. Osteosarcoma cells were transfected for 17 hours with either the antisense and the IKBsr adenovirus. 1 Ag/mL of CLIC4 reporter or NFKB reporter plasmid (31; a gift from Immunohistochemistry and Tissue BrdU and Terminal Deoxynu- Dr. Zheng-Gang Liu, National Cancer Institute) using effectene transfection cleotidyl Transferase (Tdt)–Mediated Nick End Labeling Staining. For reagent (Qiagen) before treatment with IKBsr and/or TNFa, and luciferase tissue BrdU staining, mice were i.p. injected with 20 mg/mL of BrdU (Roche, activity was measured by using the luciferase reporter assay kit (Clontech). Nutley, NJ) 1 hour before harvesting tumor tissues, and the proliferative cells Results were normalized to the total protein content. from the harvested tissues were detected by 5-bromo-2V-deoxy-uridine Labeling and Detection Kit II (Roche). For tissue terminal deoxynucleotidyl Apoptosis Assays transferase (Tdt)–mediated nick end labeling (TUNEL) staining, Apoptag Annexin V and TUNEL Fluorescence-Activated Cell Sorting Analysis. (Intergen, Burlington, MA) was used as described by the manufacturer Control or treated cells (n =2Â 105) were analyzed by allophycocyanin- except metal-enhanced 3,3V-diaminobenzidine chromogenic substrate was conjugated recombinant human Annexin V (Caltag, Burlingame, CA) used. Stained slides mounted and analyzed with bright-field microscopy and In situ Cell Death assay (Roche) as described by the manufacturer. using Leitz-DMRB (Leica, Bannockburn, IL) and OpenLab (Improvision, Analysis was done after 10,000 counting events. Data acquisition Lexington, MA) software. and analysis were done using Cell Quest software. The apoptosis data Xenograft and Tumor Development. CLIC4AS-SP1 cells were grafted as shown are one of representative results of at least three independent a skin graft on the back of nude mice, and tumor size was measured experiments. as described previously (30). One group of mice was given doxycycline- Statistical Analysis. Comparisons of experimental data were analyzed containing feed (200 Ag/kg w/w, Bioserve, Laurel, MD) starting 1 week by a two-tailed Student’s t test. P < 0.05 was considered to indicate a before grafting, whereas another group received standard mouse feed statistically significant difference. (Purina) to allow expression of the CLIC4-antisense. In one experiment, all recipient mice received doxycycline diet 1 week before grafting and subgroups were changed to control diets at 0, 1, and 3 weeks following the Results application of tumor grafts. In the second experiment, recombinant TNFa Suppressing NFKB Enhances TNFa-Mediated Apoptosis but (1 Ag per injection per mouse) was given by i.p. injection twice per week Reduces CLIC4. Human osteosarcoma (p53 Tet-On SaOS and Tet- starting at week 5 after grafting, which then continued for four more weeks On U2OS) cells were treated with TNFa to activate NFKB and/or before termination. with IKBsr (the dominant-negative mutant of IKB) adenovirus to Construction of Human CLIC4 Promoter Reporter and Luciferase K K Assays. The cloning of the 3.5-kb human CLIC4 promoter region was inhibit NF B activity selectively. NF B translocated to the described elsewhere (7). Primer sets containing NheI sites (5VNheI- nucleus within 30 minutes of TNFa treatment in both gtgtaccatgagctgtcctctgagccagg-3V and 5VNheI-ctgtgtttcaggctctgagc- osteosarcoma cells that are infected with empty (null) adenovirus tagcccttgg-3V) were used to PCR amplify 2.5 kb of the human (Fig. 1A), and IKB fluorescence was noticeably reduced after TNFa CLIC4 promoter in pGEM-T Easy (7) excluding the repeated segments. treatment (Fig. 1A). TNFa treatment also caused translocation

Figure 3. Reduction of CLIC4 by antisense expression enhances TNFa-mediated apoptosis. Osteosarcoma cell lines with stably integrated tetracycline regulated CLIC4-antisense vector were treated with TNFa and Tet-Off Ad and examined for apoptosis (A and B) and CLIC4 protein levels (C and D). Numbers between CLIC4 and actin blots (C) and (D): CLIC4 band intensity relative to the corresponding actin band as determined by densitometry using ImageMaster software. Cells were pretreated with Tet-Off Ad for 17 hours before 25 ng/mL TNFa treatment. Apoptosis was examined by flow cytometric analysis for Annexin V binding after treatment with TNFa for 24 hours in CLIC4AS-SaOS cells (A) and for 48 hours in CLIC4AS-U2OS cells (B). Columns, of three independent experiments mean; bars, FSE. Western blotting was done for CLIC4 after treatment with TNFa for 24 hours in both cell lines. *, P < 0.01, when compared with TNFa alone.

www.aacrjournals.org 565 Cancer Res 2005; 65: (2). January 15, 2005

Figure 4. IKBsr, but not CLIC4-antisense, inhibits NFKB-DNA binding and transcriptional activity. CLIC4AS-SaOS cells (A and C) and CLIC4AS-U2OS cells (B and D) were pretreated with Tet-Off, IKBsr, or Null Ad for 24 hours before TNFa treatment, and then NFKB binding and reporter gene assays were done. For the binding assay, both cell types (A and B) were treated with 25 ng/mL TNFa for 1 hour, and cell lysates were used for ELISA assay. For reporter gene assay, both cell types (C and D) were transiently transfected with NFKB-luciferase plasmid for 17 hours before the indicated MOI of adenoviral infection. Both cell types were treated with 25 ng/mL TNFa for 10 hours, and cell lysates were used for luciferase and protein assays. Columns, mean (n = 3) of one of two independent experiments; bars, FSD. *, P < 0.01, when compared with TNFa alone.

of CLIC4 to the nucleus as reported previously (16). Expression of associated with condensed chromatin and nuclear fragmentation CLIC4-antisense (indicated by T in Fig. 1B) did not prevent nuclear as detected by Hoechst staining (data not shown). Immunoblots of translocation of CLIC4 or NFKB in response to TNFa. Introduction the duplicate experiments showed that TNFa alone increased the of IKBsr (Fig. 1C), or the combination of both CLIC4-antisense and level of CLIC4 protein in both cell types (Fig. 3C and D). The IKBsr (Fig. 1D) did not prevent TNFa-mediated translocation of combination of TNFa and antisense expression enhanced apopto- CLIC4 but did prevent NFKB translocation to the nucleus, sis in both cell types, and antisense expression prevented the suggesting CLIC4 and NFKB nuclear trafficking are mediated by increase in TNFa-mediated CLIC4 protein expression in both cell independent mechanisms. IKBsr prevented translocation of NFKB types (Fig. 3B and D). Apoptosis measured in this setting was at multiplicity of infection (MOI) of 50 in both cell types (Fig. 1C), dependent on the adenoviral MOI used, suggesting that TNFa and indicating that IKBsr adenovirus was effective in blocking NFKB CLIC4-antisense expression synergize to enhance apoptosis. In activation in these cells. CLIC4AS-U20S cells, apoptosis induced by TNFa progressed more For both cell lines, flow cytometry data showed a small slowly (48 hours) than in CLIC4AS-SaOS cells, and more increase in Annexin-positive apoptotic cells after treatment with adenovirus (5 MOI of Tet-Off adenovirus) was needed to achieve 25 ng/mL TNFa (Fig. 2A and B). This increase in apoptosis is an apoptotic response in 65% of treated cells (Fig. 3B). independent of p53 because the p53 Tet-On SaOS cells lack p53 CLIC4-antisense induced by the Tet-Off adenovirus was not in the absence of doxycycline. TNFa also induced CLIC4 protein influencing NFKB nuclear translocation (Fig. 1B), but as an (Fig. 2C and D) and caused a modest increase in CLIC4 promoter independent approach, NFKB binding activity assays were done activity (Fig. 2E and F) in both cell lines. Combined treatment of after TNFa treatment of CLIC4AS-SaOS and CLIC4AS-U2OS cells these cells with TNFa plus IKBsr caused a substantial increase in (Fig. 4). NFKB in the nucleus as detected by binding to its apoptotic cells but decreased CLIC4 protein levels (Fig. 2A-D). consensus DNA binding site and NFKB transcriptional function as CLIC4 promoter activity was not significantly changed from TNFa measured by luciferase reporter activity were not significantly alone with the addition of IKBsr (Fig. 2E and F). affected by the Tet-Off adenovirus, but were completely inhibited Expression of Stably Integrated Conditional CLIC4- by the IKBsr adenovirus (Fig. 4A-D). In both cell lines, NFKB Anti-sense Enhances TNFa-Mediated Apoptosis. To determine nuclear activity was enhanced by TNFa. if reduction of CLIC4 enhanced TNFa-induced apoptosis as seen CLIC4-Antisense Reduces the Expression of Other CLIC for inhibition of NFKB activation, CLIC4AS-SaOS and CLIC4AS- Proteins. The extensive sequence homology among the CLIC U2OS cells were infected with the Tet-Off adenovirus to induce the family proteins (32) prompted us to examine if our reverse antisense vector in the presence or absence of TNFa. Treatment orientation full-length CLIC4-antisense could also reduce the with TNFa or Tet-Off adenovirus alone caused apoptosis in expression of other CLIC proteins. Antibodies to CLIC1, CLIC4, CLIC4AS-SaOS cells, and TNFa alone induced apoptosis in and CLIC5 detected single bands on immunoblots of the respective CLIC4AS-U2OS cells (Fig. 3A and B). The Annexin V binding was recombinant proteins, although there was slight cross-reactivity of

Cancer Res 2005; 65: (2). January 15, 2005 566 www.aacrjournals.org

COOH terminus CLIC4 antibody with CLIC5-recombinant protein CLIC4 when doxycycline is withdrawn or the dose is reduced (Fig. (Fig. 5A). However, CLIC5 protein is in extremely low abundance in 7A and B). CLIC4AS-SP1 cells were then grafted to nude mice that the cell lines used in these studies and requires ultra sensitive had been primed with doxycycline diet for 1 week or maintained on reagents for detection (see legend of Fig. 5B); thus, essentially each a standard diet. Tumor growth was markedly retarded on mice antibody is monospecific for cell lysates. Upon induction of CLIC4- receiving standard diet (Fig. 7C). Withdrawal of doxycycline from antisense with addition of doxycycline in the cell culture medium, the diet at 3 weeks when tumors were well established produced the levels of all three CLIC proteins decreased indicating that tumors of intermediate size compared with control mice main- CLIC4-antisense has the broader capability of reducing expression tained on doxycycline and immediate doxycycline withdrawal, of multiple CLIC family members. To determine if CLIC4 antisense suggesting CLIC4-antisense expression affected growth even in had a more general influence on reducing cellular protein well-established tumors. Three independent grafting experiments expression, the blot shown in Fig. 5B was reprobed for Bax, another showed identical results. All tumors were poorly differentiated proapoptotic protein associated with mitochondria and Bcl-2, an squamous cell carcinomas independent of expression of the CLIC4- antiapoptotic protein. In neither case was the protein level reduced antisense. Analysis of CLIC4 protein in a series of tumors excised by CLIC4-antisense expression (Fig. 5C). from the experimental groups indicated that tumors expressing the Reduction of CLIC Proteins Is as Effective as Inhibition of antisense had significantly reduced CLIC4 levels (Fig. 8A and B), NFKB for Enhancing TNFa-Induced Apoptosis in Osteosarcoma whereas the parental tumor and the CLIC4AS-SP1 tumors from Cell Lines. To compare the effects of inhibiting NFKB activity and mice maintained on doxycycline had CLIC4 levels similar to normal suppressing CLIC4 expression on TNFa-mediated apoptosis, we skin keratinocytes. Immunostaining of tumor sections revealed that treated CLIC4AS-SaOS cells with either IKBsr or Tet-Off adenovirus CLIC4 was relatively abundant in the cytoplasm of tumors not together with TNFa. After incubation with 25 ng/mL TNFa for 12 or expressing the antisense but substantially reduced in tumors in 24 hours in combination with IKBsr, 32% and 55% of the cells which doxycycline was withdrawn (Fig. 9A). Furthermore, prolifer- underwent apoptosis, whereas a 30% and 66% apoptotic response ation in the antisense expressing tumors was reduced by 65% when was measured for Tet-Off adenovirus and TNFa treatment at these assayed by BrdU staining (Fig. 9B), although a substantial number of time points (Fig. 6A). Thus, reduction of CLIC4 (together with other proliferating cells were detected. The number of apoptotic cells CLIC family members) is as effective for enhancing TNFa-mediated detected by TUNEL staining was 3-fold higher in antisense apoptosis as inhibiting NFKB in this cell line. In TNFa dose- expressing tumors than in tumors in which antisense was response studies, reduction of CLIC4 was more effective than suppressed by doxycycline (Fig. 9C). Together, these changes are suppression of NFKB(Fig.6B) in CLIC4AS-SaOS cells. In CLIC4AS- consistent with a suppressed tumor expansion, reduced prolifera- U2OS cells, both NFKB inhibition and CLIC reduction enhanced tion and increased apoptosis when CLIC proteins are reduced. TNFa apoptosis at all TNFa doses tested, and suppression of NFKB To determine if the addition of TNFa would further reduce may be more effective in these cells (Fig. 6C). tumor size, mice carrying tumor grafts on doxycycline or standard CLIC4-Antisense Expression Inhibits Tumor Growth In vivo. diets were treated with recombinant mTNFa after appearance of The reduction in viability associated with CLIC4-antisense expres- tumors at 5 weeks after grafting (Fig. 10A). Whereas TNFa and sion in osteosarcoma tumor cell lines prompted us to test the CLIC4-antisense both had a substantial antitumor effect as a single possibility that reduction in CLIC family proteins might be agent, combination therapy using TNFa and antisense modestly inhibitory to tumor growth in vivo. We created a mouse tumor improved the tumor response for at least 3 weeks after TNFa was cell line CLIC4AS-SP1 that expresses abundant CLIC4 in the given (Fig. 10A). CLIC4-antisense alone had a more powerful tumor presence of doxycycline, but substantially reduced the level of inhibitory influence than TNFa as a single therapeutic agent in vivo

Figure 5. CLIC4-antisense expression down-regulates other CLIC family members. A, bacterially expressed recombinant CLIC1, CLIC4, and CLIC5 were resolved by SDS-PAGE, and the recombinant proteins were detected by silver staining. Solid black arrows, recombinant CLIC proteins. A parallel gel was used to probe the membrane with CLIC1, CLIC4, and CLIC5 antibodies sequentially after stripping. B, CLIC4AS-U2OS cells (AS) and the parental U2OS (Par) cells were treated with doxycycline for 2 days and analyzed by immunoblotting using CLIC monospecific antibodies. Because of the low abundance of CLIC5 protein in these cells, the highly sensitive enhanced chemiluminescence (Dura, Pierce) reagent and longer exposure time were used to detect CLIC 5, whereas the more conventional enhanced chemiluminescence (Pico, Pierce) was used to detect CLIC1 and CLIC4. C, immunoblot from B was reprobed for Bax and Bcl-2. www.aacrjournals.org 567 Cancer Res 2005; 65: (2). January 15, 2005

apoptosis (7). Based on the results of the CLIC4 promoter-reporter construct, CLIC4 does seem to be a direct transcriptional target for TNFa independent of NFKB stimulation. However, the up- regulation of CLIC4 is not required for TNFa-mediated apoptosis, and in contrast, reduction of CLIC4 enhances TNFa-mediated apoptosis. This conclusion is tempered somewhat since our antisense approach involved reduction of several CLIC family proteins whereas not affecting other cellular proteins that we have tested. Thus, until more selective reagents for individual CLIC family proteins are available, we conclude that reductions in CLIC proteins together can enhance TNFa-mediated apoptosis independent of p53. A critical amount of CLIC proteins is needed to maintain cell survival. If the CLIC family proteins are required to maintain ionic balance, pH, and volume in cellular organelles, one would expect precise regulation of protein levels to be required. The inhibition of NFKB in concert with TNFa treatment reduces CLIC4 protein without altering CLIC4 transcription. This suggests that NFKB- regulated factors may contribute to the stability of CLIC proteins, and this could be a component of NFKB antiapoptotic activity. This possibility will require additional studies because the processing of CLIC proteins has not yet been explored.

Figure 6. Relative activity of IKBsr and CLIC4-antisense for enhancing TNFa- mediated apoptosis. Apoptosis was induced in CLIC4AS-SaOS (A and B) and CLIC4AS-U2OS cells (C) by TNFa in the presence of Null, IKBsr, or Tet-Off Ad and examined by flow cytometric analysis of Annexin V binding. Cells were pretreated with IKBsr or Tet-Off Ad for 17 hours before TNFa treatment. Apoptosis was induced with indicated dose of TNFa for 12 and 24 hours in CLIC4AS-SaOS cells (A), 24 hours in B and 48 hours in CLIC4AS-U2OS cells (C). Columns, mean of three independent experiments; bars, FSD. in this model. In contrast, the apoptotic response of CLIC4AS-SP1 cells to combined antisense expression and TNFa treatment in vitro Figure 7. CLIC4-antisense expression retards tumor growth in vivo. Inducible was synergistic (Fig. 10B), similar to that seen with the human CLIC4AS-SP1 keratinocytes were cultured in medium without doxycycline for the osteosarcoma tumor cell lines. indicated times (A) or in medium with various amounts (Ag/mL) of doxycycline for 2 days (B) before harvest and analyzed for CLIC4 levels by immunoblotting. (C) two million CLIC4AS-SP1 cells were mixed with 5 million BALB/c dermal Discussion fibroblasts, and the cell mixture was grafted onto the backs of the nude mice being fed either a diet containing doxycycline (ÀAS, solid triangle on black line), control This study was designed to determine if induction of CLIC4 by diet (+AS, solid circle on dotted line) or changed from doxycycline to control diet at 3 weeks (+AS at 3 weeks, solid square on gray line; n = 10 per group). Tumor size TNFa was an essential component of the TNFa-mediated apop- was measured weekly for 7 weeks and presented as mean volume F SE. totic response, as we have previously shown for p53-mediated Similar results were obtained from three independent grafting experiments.

Cancer Res 2005; 65: (2). January 15, 2005 568 www.aacrjournals.org

induces apoptosis in fibroblasts, whereas down-modulation induces apoptosis in hematopoietic cells (33, 34). The physiologic significance of CLIC4 induction and nuclear translocation in TNFa-treated cells is still unclear. It is possible that induction and translocation of CLIC4 might contribute to another physiologic function of TNFa such as differentiation or growth inhibition. The development of the Caenorhabditis elegans excretory canal requires the function of EXC-4, the worm orthologue of mammalian CLIC proteins (35), suggesting CLIC proteins may also have a role in development or tissue remodeling.

Figure 8. Tumors expressing CLIC4-antisense have reduced levels of CLIC4. Normal mouse skin (N), parental Tet-Off SP1 cells (P), and individual tumors from CLIC4AS-SP1 (solid black bar) grafts with or without antisense expression were dissected, homogenized, and protein lysates were assayed for CLIC4 levels by immunoblotting and further quantified by densitometry (A and B) using Image Master software.

We considered that suppressing CLIC4 by antisense expression could have an influence on TNFa-induced NFKB activation. However, suppressing CLIC4 did not inhibit TNFa-induced translocation of NFKB to the nucleus, DNA binding activity of NFKB, or NFKB-mediated transcriptional activation, indicating that inhibition of NFKB activation is not one of the mechanisms by which CLIC4 suppression sensitizes cells to TNFa-mediated apoptosis. Reduction in CLIC proteins is as potent as inhibition of NFKB function for enhancing TNFa-mediated apoptosis in human osteosarcoma cells. In CLIC4AS-SP1 mouse cells, the combined influence of TNFa and reduction in CLIC proteins seems to be synergistic for apoptosis. Whereas our use of antisense methods to study reduction in CLIC4 results in a broader reduction in CLIC family proteins, CLIC4 itself is perhaps the most ubiquitous and abundantly Figure 9. Expression of CLIC4-antisense in tumors inhibits cell proliferation and expressed family member and likely to be central to the results increases apoptosis. Tissue sections from the tumors with (+AS) or without (ÀAS) reported here. Nevertheless, the development of selective CLIC4 CLIC4-antisense expression were immunostained for CLIC4 and examined by light microscopy (A). BrdU was injected i.p. 1 hour before sacrifice, and BrdU modulating agents will be essential to sorting out specific action incorporation in tumors was detected by anti-BrdU antibody and immunohisto- of individual family members. A previous report (7) and current chemistry (B). BrdU positive cells (+BrdU) were counted from eight random fields results consistently indicate that an increase or reduction of CLIC4 from +AS and ÀAS tumor sections and plotted (solid black bar). Serial sections from B wereTUNEL stained (C),andpositive cells werequantifiedasdescribed inB impairs cell viability through an apoptotic pathway. This may be and plotted. Sections without secondary antibody were used as the negative control unusual but is not unique. For example, C-MYC up-regulation (c), and the DNase-treated (+DNase) sections were used as the positive control. www.aacrjournals.org 569 Cancer Res 2005; 65: (2). January 15, 2005

not as potent as the proapoptotic response in vitro; although there was a modest combined effect in vivo. Because only one regimen of TNFa administration was tested, additional modification of the protocol may produce a more pronounced result. This preclinical model suggests that reduction of CLIC levels, physiologically translated to reduced chloride channel activity, could enhance biological or cytotoxic drug–mediated antitumor therapy, but further testing is required to confirm this possibility. Ion channels and transporters have been among the most successful therapeutic targets (i.e., Ca2+ channel blocker for heart disease and Na+ transport/exchanger inhibitors for diuretics) for pharmaceuticals. Previous experimental studies indicate that a chloride channel blocker (i.e., tamoxifen), calcium channel inhibitor (i.e., verapamil) or sodium/hydrogen exchanger inhibitor (i.e., amiloride) enhance the effectiveness of cancer chemotherapy by interfering with the activity of multidrug resistance proteins that influence the ability of chemotherapeutic agents to accumulate in cancer cells in sufficient concentrations. From a therapeutic point of view, ion transport inhibitors such as intracellular chloride channel inhibitors can serve as modifiers of drug traffic across the plasma membrane. Recent reports suggest that multiple classes of intracellular chloride channel proteins are central to cell viability and thus should be considered as potential therapeutic targets in cancer. Voltage-gated chloride channels CLC-2, CLC-3, and CLC-5 channels are expressed at high levels in acute patient biopsies from low- and high-grade malignant gliomas (36). Glycine inhibits the growth of B16 melanoma tumors in vivo through a glycine-gated ClÀ channel in endothelial cells that blocks the effect of vascular endothelial growth factor on blood vessel growth (37), suggesting

Figure 10. Combined CLIC4-antisense and TNFa enhances CLIC4AS-SP1 that chloride channels may influence angiogenesis. CLIC4 expres- regression in vivo and is synergistic for apoptosis in vitro. CLIC4AS-SP1 cells sion is increased in myofibroblasts that form the stroma in breast were grafted as described in Fig. 7 and treated with recombinant mTNFa cancers, participating in the regulation of the tumor microenviron- by twice weekly intraperitoneal injections (1 Ag per injection per mouse) starting at 5 weeks postgrafting when tumors were established (A). Mice were ment (15). Calcium-activated chloride channels, CLCA1 and CLCA2, fed either a diet containing doxycycline with (+TNF/ÀAS, solid square with dotted are significantly down-regulated in f80% of colorectal carcinomas line) or without TNFa injection (PBS/ÀAS, solid diamond on black line), and >90% of highly proliferating tumor cells, suggesting a tumor and control diet with (TNF/+AS, solid triangle on dotted line) or without TNFa injection (PBS/+AS, solid circle on black line; n = 10 per group). suppressor activity (38, 39). Reconstitution of CLCA2 expression in Solid black arrow, start of TNFa administration. Tumor size was measured highly malignant cell lines reduced Matrigel invasion in vitro and weekly for 9 weeks and presented as mean volume F SE. Similar results prevented the growth or metastasis of tumor cells transplanted s.c. were obtained from three independent grafting experiments. CLIC4AS-SP1 cells in vitro were treated with (+) or without (À) TNFa (25 A/mL) in nude mice (40, 41). It is anticipated that reduction of CLIC levels at 0 hour (white bar), 24 hours (gray bar), and 48 hours (black bar) could provide an advantage in increasing the sensitivity of tumor in the presence (ÀAS) or absence (+AS) of doxycycline to control cells to drug-mediated anticancer therapy, and further clarification CLIC4-antisense expression (B). TUNEL-positive cells were detected of CLIC functions will be needed for a development of small by fluorescence-activated cell sorting analysis. molecules to modulate CLIC activities. These molecular characteristics of intracellular chloride channels suggest that chloride The human CLIC4 gene is mapped to chromosome 1p36.11, gradients and flux influence a variety of important cellular a region that is frequently altered in cancers, and previous studies controls for proliferation, migration, adhesion, and viability that have shown aberrations at this locus in lymphomas, leukemias, are involved in cancer pathogenesis, and CLIC proteins should be invasive ductal and lobular breast carcinomas, metastasizing considered as potential molecular targets for cancer. squamous cell carcinomas, and lung cancer (comparative genomic hybridization and loss of heterozygosity database, National Center Acknowledgments for Biotechnology Information). The generation of human tumor cell lines where we can conditionally regulate CLIC expression has Received 7/26/2004; revised 10/4/2004; accepted 11/11/2004. Grant support: NIH JSPS Research Fellow in Biomedical and Behavioral Research, provided evidence that CLIC proteins could be useful targets for 2003 to 2005 (M. Mutoh). tumor therapy. This prediction is further supported by the in vivo The costs of publication of this article were defrayed in part by the payment of page grafting model of a mouse squamous cell tumor line where charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. conditional expression of CLIC4-antisense reduces CLIC4, inhibits We thank Drs. Karen Vousden and Kevin Ryan of the National Cancer Institute tumor proliferation and expansion of tumor size, and increases for supplying the SaOS cell lines; Dr. Zheng-Gang Liu for supplying the NFKB- tumor cell apoptosis. These results indicate that a pan-CLIC luciferase plasmid DNA; Dr. Dennis Guttridge of the University of North Carolina for supplying the IKBsr adenovirus; Dr. Narayan Bhat of the Science Applications knockdown could serve as a novel anticancer therapy. In vitro International Co., Inc., Frederick, National Cancer Institute, Molecular Biology, Gene studies suggest that lowering CLIC levels could significantly Expression Laboratory for adenoviral amplification; Barbara Taylor of the CCR FACS K Core Facility; Dr. Mark Berryman for the generous contribution of antibodies against enhance the apoptotic activity of TNFa, independent of NF Bor CLIC1 and CLIC5 and recombinant CLIC proteins; and Bettie Sugar for the excellent 53 pathways. The antitumor response to combined treatment was editorial assistance.

Cancer Res 2005; 65: (2). January 15, 2005 570 www.aacrjournals.org

References 15. Ronnov-Jessen L, Villadsen R, Edwards JC, Petersen type-specific expression of a lacZ gene in the adult and OW. Differential expression of a chloride intracellular during development. Differentiation 1994;58:53–64. 1. Jentsch TJ. Chloride channels: a molecular perspec- channel gene, CLIC4, in transforming growth factor- 29. GossenM,BujardH.Tightcontrolofgene tive. Curr Opin Neurobiol 1996;6:303–10. h1-mediated conversion of fibroblasts to myofibro- expression in mammalian cells by tetracycline-respon- 2. Jentsch TJ, Stein V, Weinreich F, Zdebik AA. blasts. Am J Pathol 2002;161:471–80. sive promoters. Proc Natl Acad Sci U S A 1996;89: Molecular structure and physiological function of 16. Suh KS, Mutoh M, Nagashima K, et al. The 5547–51. chloride channels. Physiol Rev 2002;82:503–68. organellular chloride channel protein CLIC4/mtCLIC 30. Dlugosz AA, Glick AB, Tennenbaum T, Weinberg WC, À 3. Edwards JC. A novel p64-related Cl channel: translocates to the nucleus in response to cellular Yuspa SH. Isolation and utilization of epidermal subcellular distribution and nephron segment-specific stress and accelerates apoptosis. J Biol Chem 2004;279: keratinocytyes for oncogene research. In: Vogt PK, expression. Am J Physiol 1999;276:F398–408. 4632–41. Verma IM, editors. Methods in enzymology, vol. 254. 4. Nishizawa T, Nagao T, Iwatsubo T, Forte JG, 17. Lejeune FJ. Clinical use of TNF revisited: improving New York: Academic Press; 1995. p. 3–20. Urushidani T. Molecular cloning and characterization penetration of anti-cancer agents by increasing 31. Hur GM, Lewis J, Yang Q, et al. The death of a novel chloride intracellular channel-related vascular permeability. J Clin Invest 2002;110:433–5. domain kinase RIP has an essential role in DNA protein, parchorin, expressed in water-secreting cells. 18. Baud V, Karin M. Signal transduction by tumor damage-induced NF-K B activation. Genes Dev 2003; J Biol Chem 2000;275:11164–73. necrosis factor and its relatives. Trends Cell Biol 2001; 17:873–82. 5. Valenzuela SM, Martin DK, Por SB, et al. Molecular 11:372–7. 32. Ashley RH. Challenging accepted ion channel cloning and expression of a chloride ion channel of 19. Wang CY, Mayo MW, Korneluk RG, Goeddel DV, biology: p64 and the CLIC family of putative cell nuclei. J Biol Chem 1997;272:12575–82. Baldwin AS Jr. NF-KB antiapoptosis: induction of intracellular anion channel proteins (review). Mol 6. Harrop SJ, DeMaere MZ, Fairlie WD, et al. Crystal TRAF1 and TRAF2 and c-IAP1 and c-IAP2 to suppress Membr Biol 2003;20:1–11. structure of a soluble form of the intracellular chloride caspase-8 activation. Science 1998;281:1680–3. 33. Evan GI, Wyllie AH, Gilbert CS, et al. Induction of ion channel CLIC1 (NCC27) at 1.4-A resolution. J Biol 20. Brown K, Gerstberger S, Carlson L, Franzoso G, apoptosis in fibroblasts by c-myc protein. Cell 1992;69: Chem 2001;276:44993–5000. Siebenlist U. Control of IKB-a proteolysis by site- 119–28. 7. Fernandez-Salas E, Suh KS, Speransky VV, et al. specific, signal-induced phosphorylation. Science 1995; 34. Kimura S, Maekawa T, Hirakawa K, Murakami A, mtCLIC/CLIC4, an organellular chloride channel 267:1485–8. Abe T. Alterations of c-myc expression by antisense protein, is increased by DNA damage and participates 21. Whiteside ST, Ernst MK, LeBail O, Laurent-Winter C, oligodeoxynucleotides enhance the induction of apo- in the apoptotic response to p53. Mol Cell Biol Rice N, Israel A. N- and C-terminal sequences control ptosis in HL-60 cells. Cancer Res 1995;55:1379–84. 2002;22:3610–20. degradation of MAD3/IKBa in response to inducers of 35. Berry KL, Bulow HE, Hall DH, Hobert O. A C. 8. Fernandez-Salas E, Sagar M, Cheng C, Yuspa SH, NF-KB activity. Mol Cell Biol 1995;15:5339–45. elegans CLIC-like protein required for intracellular Weinberg WC. p53 and tumor necrosis factor a 22. Wang CY, Mayo MW, Baldwin AS Jr. TNF-a and tube formation and maintenance. Science 2003; regulate the expression of a mitochondrial chloride cancer therapy-induced apoptosis: potentiation by 302:2134–7. channel protein. J Biol Chem 1999;274:36488–97. inhibition of NF-KB. Science 1996;274:784–7. 36. Olsen ML, Schade S, Lyons SA, Amaral MD, 9. Chuang JZ, Milner TA, Zhu M, Sung CH. A 29 kDa 23. Ichijo H, Nishida E, Irie K, et al. Induction of Sontheimer H. Expression of voltage-gated chloride intracellular chloride channel p64H1 is associated with apoptosis by ASK1, a mammalian MAPKKK that channels in human glioma cells. J Neurosci 2003;23: large dense-core vesicles in rat hippocampal neurons. activates SAPK/JNK and p38 signaling pathways. 5572–82. J Neurosci 1999;19:2919–28. Science 1997;275:90–4. 37. Yamashina S, Konno A, Wheeler MD, Rusyn I, Rusyn 10. Duncan RR, Westwood PK, Boyd A, Ashley RH. Rat 24. Bartke T, Siegmund D, Peters N, et al. p53 EV, Cox AD, Thurman RG. Endothelial cells contain a brain p64H1, expression of a new member of the p64 upregulates cFLIP, inhibits transcription of NF-KB- glycine-gated chloride channel. Nutr Cancer 2001;40: chloride channel protein family in endoplasmic regulated genes and induces caspase-8-indepen- 197–204. reticulum. J Biol Chem 1997;272:23880–6. dent cell death in DLD-1 cells. Oncogene 2001;20: 38. Bustin SA, Li SR, Phillips S, Dorudi S. Expres- 11. Edwards JC, Tulk B, Schlesinger PH. Functional 571–80. sion of HLA class II in colorectal cancer: evidence expression of p64, an intracellular chloride channel 25. Tang F, Tang G, Xiang J, Dai Q, Rosner MR, Lin A. for enhanced immunogenicity of microsatellite- protein. J Membr Biol 1998;163:119–27. The absence of NF-KB-mediated inhibition of c-Jun N- instability-positive tumours. Tumour Biol 2001;22: 12. Qian Z, Okuhara D, Abe MK, Rosner MR. Molecular terminal kinase activation contributes to tumor 294–8. cloning and characterization of a mitogen-activated necrosis factor a-induced apoptosis. Mol Cell Biol 39. Bustin SA, Li SR, Dorudi S. Expression of the Ca2+- protein kinase-associated intracellular chloride chan- 2002;22:8571–9. activated chloride channel genes CLCA1 and CLCA2 is nel. J Biol Chem 1999;274:1621–7. 26. Ryan KM, Ernst MK, Rice NR, Vousden KH. Role of downregulated in human colorectal cancer. DNA Cell 13. Tulk BM, Kapadia S, Edwards JC. CLIC1 inserts from NF-KB in p53-mediated programmed cell death. Biol 2001;20:331–8. the aqueous phase into phospholipid membranes, Nature 2000;404:892–7. 40. Pauli BU, Abdel-Ghany M, Cheng HC, Gruber AD, where it functions as an anion channel. Am J Physiol 27. Berryman M, Bretscher A. Identification of a novel Archibald HA, Elble RC. Molecular characteristics and Cell Physiol 2002;282:C1103–12. member of the chloride intracellular channel gene functional diversity of CLCA family members. Clin Exp 14. Suginta W, Karoulias N, Aitken A, Ashley RH. family (CLIC5) that associates with the actin cytoskel- Pharmacol Physiol 2000;27:901–5. Chloride intracellular channel protein CLIC4 (p64H1) eton of placental microvilli. Mol Biol Cell 2000;11: 41. Elble RC, Pauli BU. Tumor suppression by a binds directly to brain dynamin I in a complex 1509–21. proapoptotic calcium-activated chloride channel in containing actin, tubulin and 14-3-3 isoforms. 28. Ramirez A, Bravo A, Jorcano JL, Vidal M. Sequences mammary epithelium. J Biol Chem 2001;276: Biochem J 2001;359:55–64. 5V of the bovine keratin 5 gene direct tissue- and cell- 40510–7.

www.aacrjournals.org 571 Cancer Res 2005; 65: (2). January 15, 2005

Downloaded from cancerres.aacrjournals.org on October 1, 2021. © 2005 American Association for Cancer Research. Antisense Suppression of the Chloride Intracellular Channel Family Induces Apoptosis, Enhances Tumor Necrosis Factor α-Induced Apoptosis, and Inhibits Tumor Growth

Kwang S. Suh, Michihiro Mutoh, Michael Gerdes, et al.

Cancer Res 2005;65:562-571.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/65/2/562

Cited articles This article cites 38 articles, 21 of which you can access for free at: http://cancerres.aacrjournals.org/content/65/2/562.full#ref-list-1

Citing articles This article has been cited by 7 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/65/2/562.full#related-urls

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/65/2/562. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.