Death Receptor 4 (DR4) Efficiently Kills Breast Cancer Cells Irrespective Of

Home , Death receptor 4, KIT (gene)

Death receptor 4 (DR4) efficiently kills breast cancer cells irrespective of their sensitivity to tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) Irene Kazhdan,1 and Robert A Marciniak1,2 1Department of Medicine, Division of Medical Oncology, University of Texas Health Science Center at San Antonio, Texas, USA; and 2South Texas Veterans Health Care System, Texas, USA.

Breast cancer cells are generally resistant to induction of apoptosis by treatment with tumor necrosis factor-related apoptosis- inducing ligand (TRAIL). In this study, we demonstrate that both TRAIL-sensitive and TRAIL-resistant breast cancer cell lines can be efficiently killed by overexpression of the TRAIL receptor, death receptor 4 (DR4). The extent of cell death depended on the strength of the promoter driving DR4 expression. When driven by the strong CMV promoter, expression of DR4 killed over 90% of cells in five out of six cell lines tested in the absence of exogenous TRAIL. When driven by the relatively weak tumor-specific hTERT promoter, DR4 was less effective alone, but sensitized cells to killing by TRAIL. The extent of TRAIL sensitization depended on the magnitude of hTERT promoter activity. MCF-7 cells were relatively resistant to the action of DR4. We compared expression of the genes involved in transduction and execution of the death receptor-initiated apoptotic stimuli between MCF-7 and DR4-sensitive cell lines. We confirmed that in the panel of cell lines, MCF-7 was the only line deficient in expression of caspase 3. Bcl-2 and FLIP proteins, implicated in suppression of TRAIL-induced apoptosis, were expressed at a higher level. Cancer Gene Therapy (2004) 11, 691–698. doi:10.1038/sj.cgt.7700747 Published online 3 September 2004 Keywords: breast cancer; death receptor 4; tumor targeting; hTERT promoter

reatment with tumor necrosis factor-related apopto- balance of the apoptotic signal, transmitted by death Tsis-inducing ligand (TRAIL) induces programmed receptors, and prosurvival action of inhibitors of the cell death in a wide range of transformed cells both in extrinsic pathway of apoptosis. Therefore, we hypothe- vitro and in vivo without producing significant toxicity to sized that overexpression of a death receptor will tip this normal cells.1–3 This unique property makes TRAIL an balance towards programmed cell death and will thus attractive candidate for cancer therapy. However, a sensitize breast cancer cells to TRAIL. In this report, we significant proportion of breast cancer cell lines is show that expression of DR4 from a strong CMV resistant to TRAIL-induced apoptosis.4,5 promoter efficiently kills breast cancer cells even in the Resistance to TRAIL is likely mediated by a number of absence of exogenous TRAIL. When targeted to cancer mechanisms. Binding of TRAIL to its functional recep- cells by a relatively weak hTERT promoter, DR4 alone tors, DR4 and DR5, can be decreased due to TRAIL induces apoptosis less efficiently, but sensitized cells to the sequestering by nonfunctional decoy receptors that do not effects of soluble TRAIL. transmit an apoptotic stimulus.6–9 Mutational loss of expression of DR4 and DR5, allelic loss or methylation of the genes has also been observed in several neoplasms including metastatic breast cancers.10–14 Death receptor Materials and methods signaling can be counteracted at multiple levels by Cell lines and reagents enhanced expression and/or activity of inhibitory mole- Human breast cancer cell lines MCF-7, CAMA-1, HCC cules, including FLIP,15–18 bcl-2 and bcl-XL,19–21 mem- 1937, MDA-MB-231, T-47D and AU 565 were obtained bers of IAP family,22 Akt23,24 and NFkB.4,25 The from American Type Culture Collection (ATCC, Mana- resultant sensitivity to TRAIL is thus determined by a ssas, VA). Cells were maintained in RPMI 1640 (T-47D, AU-562 and HCC 1937), a-MEM (MCF-7 and CAMA-1) or Leibovitz’s-15 (MDA-MB-231) medium with 10% Received October 11, 2003. Address correspondence and reprint requests to: Dr Irene Kazhdan, FBS, penicillin (100 U/ml) and streptomycin (100 mg/ml) MD, PhD, Division of Medical Oncology, Department of Medicine, at 371C. Except for MDA-MB-231, all cell lines were University of Texas Health Science, Center at San Antonio, 7703 grown in humidified atmosphere with 5% CO2. Reagents Floyd Curl Drive, San Antonio, TX 78229, USA. used in the study were obtained from the following E-mail: [email protected] companies: FuGENE 6 Transfection Reagent — Roche DR4 efficiently kills breast cancer cells I Kazhdan and RA Marciniak 692 Applied Science (Indianapolis, IN); Dual-Luciferase reporter gene at a ration of 1:1. When indicated, soluble Reporter Assay System — Promega (Madison, WI); TRAIL was added to cultures at the time of transfection. Luminescent b-gal detection kit — BD Bioscience b-Galactosidase activity was tested with a luminescent b- Clontech (Palo Alto, CA); soluble TRAIL (residues gal detection kit 36 hours later. Relative cell survival was 114–281 of the soluble domain of human TRAIL, measured as a percent of b-galactosidase reporter activity expressed in Escherichia coli.) — BIOMOL Research in DR4 vs. vector-transfected cultures. Laboratories, Inc. (Plymouth Meeting, PA); CellTiter 96 Non-Radioactive Cell Proliferation Assay (MTS assay) MTS cytotoxicity assay — Promega (Madison, WI); ApoAlertTM Annexin V-PE Cells were plated 2 Â 104/well in triplicate wells and Apoptosis kit — BD Bioscience Clontech (Palo Alto, incubated overnight. Serial dilutions of soluble TRAIL in CA); hApo Multi-Probe Template Sets — BD Bioscience Hank’s balanced salt solution (HBSS) were added to the Clontech (Palo Alto, CA); Antibodies to DR4, DR5, wells and cells were incubated for additional 36 hours. FLIP and TRAIL — QED Bioscience Inc. (San Diego, Cytotoxicity was assessed using CellTiter 96 Non-Radio- CA); antibodies to caspase 3, bcl-2 and actin — Santa active Cell Proliferation Assay and medium background Cruz (San Diego, CA); donkey anti-rabbit and anti- was subtracted. Results are presented as percent of HBSS- mouse IgG peroxidase-labeled antibody and ECLTM treated control. Western Blotting Detection Reagents —Amersham Bio- sciences (Little Chalfont, Buckinghamshire, UK). Annexin V determination Plasmids and cloning In total, 1 Â 106 cells were plated per 60 mm dish and cultured overnight. Cells were then left untreated or pRL-TK and pSV40-b-galactosidase were purchased treated with soluble TRAIL (100 ng/ml) and incubated from Promega. DR4 cDNA in pCMV-Sport6 vector for additional 36 hours. Cells were then trypsinized and was obtained from ATCC Mammalian Gene Collection washed with 1 Â Binding Buffer. For each condition, half (BC012866). An EcoRI/ApoI DR4-containing fragment of the cell suspension was left untreated and another half was subcloned into EcoRI sites of pcDNA3.1 and IRES2- was incubated for 15 minute at room temperature in the EGFP to produce CMV-DR4 and IRES-DR4, respec- dark with Annexin V. Immediately after staining, cells tively. pp76, containing luciferase gene under control of were analyzed by flow cytometry. hTERT promoterÀ1375 relative to transcription start site in pGL3 Basic vector, was kindly provided by Dr Kyo. It was cut on MluI/PstI, blunt-ended and reclosed to delete Western blotting XbaI site. Luciferase activity of the resultant construct, Exponentially growing cells were lysed in lysis buffer tested in our panel of breast cancer cell lines, did not differ (62.5 mM Tris-HCL, pH 6.8, 2% SDS, 1 mM phenyl- from that of pp76. The XbaI/HindIII fragment containing methylsulfonyl fluoride (PMSF). Cell lysates were soni- luciferase gene was than replaced by XbaI/HindIII cated and protein concentrations were determined using fragment containing DR4, isolated from CMV-DR4, to Bio-Rad protein assay reagent. In all, 25 mg of whole-cell produce hTERT-DR4. To produce CMV promoter- protein lysate from each sample was diluted in 25 mlof guided luciferase, luciferase gene was cut out of pGL3- SDS loading buffer (4% SDS, 2% glycerol, 0.01% Basic on HindIII/XbaI, cloned into HindIII/XbaI site of bromphenol blue and 125 mM Tris-HCL,pH 6.8) incu- pcDNA3.1 and called CMV-Luc. To incorporate lucifer- bated 5 minutes at 981C, 5 minutes on ice and spun down ase gene into IRES-EGFP2 vector, pGL3-Basic was cut at maximal speed in a microcentrifuge at room tempera- on XbaI, blunt-ended and cut on NheI. The resultant ture. Proteins were resolved on SDS polyacrylamide gel luciferase-containing fragment was cloned into the SmaI/ and transferred to PVDF membrane. Membranes were NheI site of IRES2-EGFP vector to produce IRES-Luc. blocked for 1 hour in 5% nonfat dry milk, 10 mM Tris- HCl pH 7.5, 150 mM NaCl, 0.1% Tween 20 and Transfections and reporter assays incubated for 2 hours with primary antibodies diluted in blocking buffer. Membranes were then washed four times Transient transfections were performed with FuGENE 6 with washing buffer (10 mM Tris-HCl pH 7.5, 150 mM transfection reagent as per the manufacturer’s protocol. NaCl, 0.1% Tween 20) and incubated for 1 hour with Briefly, 5 Â 104 cells/well were plated in 24-well plates and horseradish peroxidase-labeled secondary antibody in incubated overnight. For promoter activity studies, blocking buffer. hTERT- or CMV-firefly luciferase promoter construct was cotransfected with pRL-TK (Promega) at a ratio of 1:1 into duplicate wells. Cells were harvested 36 hours RNase protection assay later and dual luciferase assay was performed using RNase protection assays were performed with hApo Promega Dual-Luciferase Reporter Assay System as per Multi-Probe Template Sets (BD Biosciences) according to the manufacturer’s protocol. Results are presented as a the manufacturer’s protocol. Briefly, 5 mg of total RNA, ratio of Firefly/Renilla luciferase activity. For cell killing collected from each of the exponentially growing cell experiments, 5 Â 104 cells in duplicate wells were cotrans- lines, was hybridized overnight to 32P-labeled riboprobes, fected with a DR4-containing construct or a correspond- produced on the templates included in the sets. Unhy- ing vector together with the pSV40-b-galactosidase bridized probes were digested by RNaseA. Probes,

Cancer Gene Therapy DR4 efficiently kills breast cancer cells I Kazhdan and RA Marciniak 693 protected by hybridization to cellular RNA, were resolved ‘‘killer gene’’ in transfected cells results in cell death and on 5% sequencing gel and detected by autoradiography. thus prevents accumulation of the b-galactosidase repor- Expression of the apoptosis-related genes was normalized ter protein. Survival is measured as reduction of b- to that of housekeeping genes (GAPDH). galactosidase activity in DR4-transfected compared to vector-transfected cultures (Fig 1). Introduction of CMV- DR4 decreased cell viability below 5% of the vector- Results transfected cells in five out of six cell lines in the absence of exogenous TRAIL (Fig 1a). In the relatively resistant CMVpromoter/enhancer-guided DR4 efficiently kills MCF-7 cell line, cell survival ranged between 16 to 21%. breast cancer cells Addition of TRAIL further decreased cell survival to To evaluate the potential of DR4 overexpression in below 1% in AU-562, CAMA-1 and HCC 1937 cells, but treating breast cancer, we assembled a panel of six breast had no effect in MCF-7, MDA-MB-231 and T47D cell cancer cell lines. These cell lines differ in expression of, or lines (Fig 1b). harbor mutations in, genes that are considered relevant As an alternative method to estimate the effect of DR4 for the prognosis and management of the disease: expression on cell survival, DR4 or luciferase (as a estrogen receptor (ER), Her2/neu and p53. Thus, these control) was cloned into vector pIRES2-eGFP (IRES- cell lines partially reflect the genetic heterogeneity present DR4 and IRES-Luc, respectively). Thus, the reporter and in primary breast cancer. When known, other significant the ‘‘killer’’ gene did not have to be cotransfected; they alterations (c-myc amplification, BRCA1 mutation) were were present on the same construct and translated from also included. The following cell lines were used: MCF-7 the same bicistronic mRNA. Cell killing was manifested (ER positive, Her2/neu negative, p53 wild type); MDA- by the decrease in GFP-positive cells in IRES-DR4- MB-231 (ER negative, Her2/neu negative, p53 mutant); transfected cultures as compared with cultures transfected T-47D (ER positive, Her2/neu amplified, p53 mutant); with IRES-Luc. To confirm that overexpression of DR4 HCC 1937 (ER negative, Her2/neu negative, p53 mutant, killed cells via induction of apoptosis, transfected cell BRCA1 mutation); AU 565 (ER negative, Her2/neu population was also incubated with AnnexinV-PE. As amplified and overexpressed, p53 mutant, c-myc ampli- demonstrated in Table 1, when the CAMA-1 cell line is fied); CAMA-1 (ER positive, Her2 positive, p53 status transiently transfected with the IRES-Luc construct, unknown, c-myc amplified). 19.7% of cells expressed GFP 48 hours after transfection. To assess the killing efficiency of DR4, we employed In IRES-DR4-transfected populations, the proportion of two approaches. In the first, all six breast cancer cell lines GFP-positive cells was significantly decreased. This were cotransfected with a b-galactosidase reporter gene decrease in viable, GFP-positive cells correlated with the together with DR4 under the control of the CMV increase in apoptotic, Annexin V-positive cells in the promoter/enhancer (CMV-DR4) or empty vector. This transfected population. We did not see a significant approach is based on the observation that expression of a number of GFP þ AnnexinV þ cells. This is most likely

Figure 1 Killing efficiency of CMV-DR4. (a) Cells were plated in duplicates and cotransfected at a ratio of 1:1 with a b-galactosidase reporter plasmid and pcDNA3.1 vector (Control) or CMV-DR4. b-Galactosidase activity was determined 36 hours later. Cell survival was measured as a percent of b-galactosidase activity in CMV-DR4-transfected cultures relative to vector-transfected control cultures. The data represent an average of at least three independent experiments. (b) Cells were transfected as described above and either left untreated, or treated with 100 ng of TRAIL at the time of transfection. b-Galactosidase activity was measured 36 hours post-transfection. Relative cell survival was determined as described above.

Cancer Gene Therapy DR4 efficiently kills breast cancer cells I Kazhdan and RA Marciniak 694 Table 1 Cell death induced by overexpression of DR4 in CAMA-1 high, although significantly (120–450 times) lower than cells that of the CMV promoter/enhancer (Fig 3a, b). Decrease in Increase in We next assessed the killing efficiency of hTERT GFP+cells Annexin promoter-guided DR4 (hTERT-DR4). In all cell lines GFP due to DR4 Annexin V+cells examined, hTERT-DR4 alone was significantly less +cells expression V+cells due to DR4 effective than CMV-DR4. Among three cell lines with Construct (%) (%) (%) expression (%) relatively high hTERT promoter activity, two (CAMA-1 and AU 565) demonstrated an average of 36 and 63% cell Untransfected 0 11.8 survival, respectively (Fig 4a). Addition of TRAIL further IRES-Luc 19.7 19.7 decreased survival of CAMA-1 cells to an average of 7% IRES-DR4 3.5 16.2 37.5 17.8 and AU 565 cells to an average of 36%. MCF-7 cells, CAMA-1 cells were transfected with IRES2-EGFP constructs. which were relatively resistant to high levels of DR4 IRES-Luc, in which luciferase gene served as a surrogate ‘‘killer expressed from a CMV promoter, did not show sig- gene’’, were used as controls to IRES-DR4. At 36 hours after nificant cell death when transfected with hTERT-DR4, transfection, cells were harvested and evaluated for the fractions with or without addition of exogenous TRAIL. of GFP and Annexin V-positive cells by flow cytometry. The In cell lines with relatively low hTERT promoter results of a representative experiment are presented. activity, hTERT-DR4 did not cause significant decrease in cell survival when used alone (Fig 4b, stripped bar). due to the rapid cell death induced by transfected ‘‘killer When hTERT-DR4 was used in combination with gene’’, which does not provide sufficient time for GFP to TRAIL (white bar), it did not increase cell death over be transcribed, translated, accumulate and undergo that achieved by TRAIL alone (black bar). oxidation to the ‘‘mature’’ fluorescent form. To demonstrate that increased expression of DR4 was capable of killing TRAIL-resistant cells, we tested sensitivity of the cell lines to soluble TRAIL. Cells were Expression of genes involved in TRAIL-induced treated with increasing concentrations of recombinant apoptosis in breast cancer cells TRAIL or PBS control for 48 hours and cell viability was The data presented indicates that the killing efficiency of tested by MTS assay. Only two cell lines, MDA-MB-231 hTERT-DR4 depends on two factors: the activity of and CAMA-1, showed a significant TRAIL-dependent hTERT promoter and the resistance to DR4-induced reduction of viability (Fig 2a). In MDA-MB-231 cells, apoptosis. To begin assessment of the mechanisms that TRAIL had a biphasic effect, decreasing cell viability at may contribute to resistance to DR4, we compared higher doses while increasing cell proliferation and/or expression of genes involved in TRAIL-mediated apop- survival at lower concentrations. This prosurvival effect tosis in the relatively DR4-resistant MCF-7 cells and the may be explained by the observation that low doses of five DR4-sensitive cell lines. TRAIL are sufficient to activate of NF-kB, but insuffi- The levels of DR4 and DR5 proteins in MCF-7 cells cient to induce apoptosis.26,27 Further increase of TRAIL were not significantly lower than in other cell lines under concentrations did not enhance cell death in any of the study (Fig 5a). MCF-7 cells did not express TRAIL, but cell lines tested (data not shown). To confirm results this was also true for three out of five sensitive cell lines. obtained in MTS test, cells were incubated with 100 ng of MCF-7 cells did express elevated levels of two splice TRAIL for 36 hours, stained for Annexin V and analyzed variants of the apoptosis-inhibiting protein FLIP — by flow cytometry (Fig 2b). The results were in general FLIPL and FLIPs. agreement with the data obtained by MTS test, although MCF-7 cells are deficient in caspase 3.28 This deficiency MCF-7 and HCC 1937 cells, completely resistant to could contribute to their relative inability to undergo TRAIL-induced apoptosis by MTS, both showed ap- apoptosis in response to DR4. However, this deficiency proximately 17% increase in Annexin V staining. Thus, might not be restricted to MCF-7 cell line, as recent although only two breast cancer cell lines were sensitive to evaluation of caspase 3 mRNA and protein in primary soluble TRAIL, all of them were efficiently destroyed by breast tumors demonstrated lack of expression in 75% of DR4. the samples.29 To determine whether the absence of caspase 3 expression in MCF-7 cells was unique, we evaluated expression of caspases in our panel of breast Killing efficiency of hTERT-guided DR4 in breast cancer cancer cell lines by Western blotting. No caspase 3 protein lines was detected in MCF-7 cells, while all other cell lines The efficiency of DR4 in killing cancer cells irrespective to expressed readily detectable level of protein (Fig 5b). Of their sensitivity to TRAIL suggested that DR4 might be note, MCF-7 cells express caspase 7, a member of the equally toxic to normal cells. To minimize potential caspase 3 subfamily, which could substitute for caspase 3 toxicity, we targeted expression of this gene to telomerase- in execution of apoptosis (Fig 5b). We also performed positive cancer cells. First, we examined hTERT promo- screening of expression of other caspase family members ter activity in breast cancer cells and compared it to the by RNase protection assay. There was no difference in activity of CMV (Fig 3). In three cell lines, CAMA-1, AU expression of caspases 2, 4, 8 and 10A between MCF-7 565 and MCF-7, hTERT promoter activity was relatively and the DR4-sensitive cell lines. Caspase 6 expression was

Cancer Gene Therapy DR4 efficiently kills breast cancer cells I Kazhdan and RA Marciniak 695

Figure 2 Effect of TRAIL on breast cancer cell lines. (a) Breast cancer cell lines were plated in triplicate wells and treated with increasing concentrations of TRAIL; PBS-treated cells served as control. Cytotoxicity was determined by MTS test 36 hours later. The data represent an average of three independent experiments. (b) Cells were left untreated or treated with TRAIL at a concentration of 100 ng/ml. After 36 hours, cells were harvested, stained with PE-labeled Annexin V and analyzed by flow cytometry. Estimated percentages of GFP-positive cells are indicated. lower in MCF-7 cells, but the significance of this finding is expression of bcl 2 was further confirmed by Western uncertain (data not shown). blotting (Fig 5b). As expression of Bcl-XL was specifically Other mediators of apoptosis that have implicated in implicated in TRAIL resistance,21 we confirmed the resistance to TRAIL include members of the bcl 2 and absence of correlation of Bcl-XL expression and DR4 IAP families. Given a large number of genes in each resistance by Western blot (Fig 5b). family, we initially screened expression of 22 bcl 2 and IAP family members by RNAse protection.19–21 No correlation of DR4 resistance with mRNA levels was identified for bcl-2 family members Bcl-XL, bclw and Discussion mcl-1, as well as for IAP family members XIAP, survivin, c-IAP 1 and 2 and NIAP (data not shown). bcl 2 mRNA The unique ability of TRAIL to selectively kill tumor cells was expressed at a higher level in MCF-7 as compared to of diverse origin has attracted much attention over the the DR4-sensitive cell lines (data not shown). Over- past years.1,2 However, several studies demonstrated that

Cancer Gene Therapy DR4 efficiently kills breast cancer cells I Kazhdan and RA Marciniak 696

Figure 4 Killing efficiency of hTERT-DR4 in breast cancer cell lines with relatively high hTERT promoter activity (a) and absence of toxicity in cells with low hTERT promoter activity (b). Cells were transfected with b-galactosidase together with hTERT-DR4 or hTERT-Luciferase and incubated for 36 hours with or without 100 ng of TRAIL. Cell survival was measured as a percent of b-galactosidase activity in hTERT-DR4-transfected cultures relative to vector-transfected cultures.

TRAIL, which then interacts with DR4; (2) high intracellular concentrations of DR4 lead to its ligand- independent oligomerization. As only two cell lines Figure 3 Activity of hTERT promoter in breast cancer cell lines. (a) expressed detectable levels of TRAIL on Western blot, Evaluation of hTERT promoter activity. Cells in duplicate wells were the first explanation is unlikely to be true. A role of cotransfected with phTERT and pRL-TK plasmids at a ratio of 1:1. ligand-independent receptor oligomerization at high Dual luciferase assay was performed 36 hours later. The data are concentrations is also suggested by the fact that cytotoxi- presented as a ratio of hTERT-guided firefly luciferase activity to Renilla luciferase activity and represents an average of at least three city of CMV-guided DR4 was not significantly enhanced independent experiments. (b) hTERT promoter activity is signifi- by addition of exogenous TRAIL. Ligand-independent cantly lower than that of CMV promoter in breast cancer cells. Cells oligomerization was previously demonstrated for another 30,31 were transfected with pRL-TK together with either phTERT or CMV- death receptor, Fas. Luc construct (in which firefly luciferase is guided by CMV promoter) For cancer gene therapy applications, strong cytotoxi- as described above. The data is presented as a ratio of hTERT- city, exhibited by DR4 overexpression, required targeting. guided firefly luciferase activity to RL activity. For this purpose, we used the promoter of the hTERT gene, which has been shown to be selectively active in a significant proportion of breast cancer cell lines are cancer cells,32–35 to drive DR4 expression. The main resistant to soluble TRAIL. In this study, we attempted to disadvantage of the majority of the currently available overcome TRAIL resistance by overexpression of the tumor-specific promoters, including hTERT, is relative TRAIL receptor, DR4. We hypothesized that overexpres- low promoter strength. Indeed, among breast cancer cell sion of a functional death receptor will enhance the death lines tested, only three (CAMA-1, AU 565 and MCF-7) stimulus sufficiently to counterbalance the activity of had comparatively high hTERT promoter activity. In the prosurvival pathways and thus will tip the balance cell lines with high hTERT promoter activity, the level of towards apoptosis. expression of the Luciferase reporter driven by the Indeed, introduction of DR4 under strong CMV hTERT promoter was still two orders of magnitude promoter efficiently killed all breast cancer cell lines lower than that of CMV promoter. In two of these cell irrespective to their sensitivity to soluble TRAIL. MCF-7 lines (CAMA-1 and AU 565), hTERT-DR4 induced cell cells were somewhat more resistant than the other cell death. In accord with the promoter activity studies, the lines (average relative cell survival in CMV-DR4-expres- degree of cell killing by hTERT-DR4 in these cell lines sing cells was 18.3 vs. 0.75–4.4% survival obtained with was lower than that induced by CMV-DR4. However, other cells). The ability of DR4 to kill cells in the absence when cells were treated with a combination of hTERT- of external TRAIL could be explained by two mechan- DR4 and TRAIL, a significant increase in cell killing was isms: (1) cell lines under study themselves produce observed. In both cases, the combined effect of hTERT-

Cancer Gene Therapy DR4 efficiently kills breast cancer cells I Kazhdan and RA Marciniak 697 DR4-resistance genes. Expression of Bcl-XL, another bcl- 2 family member implicated in TRAIL resistance,19,21 did not show correlation with resistance to DR4. There was also no correlation between DR4 resistance and expression of caspases, IAP and other bcl-2 family members. In our panel of cell lines, MCF-7 was the only line that lacked expression of caspase 3. It has been demonstrated that, despite the presence of functional effector caspase 7, apoptotic response to some stimuli, including staurosporine, VP-1638 and photodynamic therapy,39 was delayed or diminished in this cell line. Reconstitution of caspase 3 expression enhanced apoptosis induced by staurosporine and chemotherapeutic agents.29,38 These data suggest that caspase 3 deficiency could at least partially account for resistance of MCF-7 cells to DR4. On the other hand, apoptosis induced by overexpression of the proapoptotic bcl-2 family member Bak was not impaired in MCF-7 cells as compared to the caspase 3-expressing HeLa cells.40

Acknowledgments

We thank Dr Satoru Kyo for kindly supplying us with the hTERT promoter constructs. This work was supported Figure 5 Expression of genes implicated in TRAIL-induced apopto- by the NIH 5 K12 CA01723-10 Physician Scientist sis. Total protein extracts were prepared from exponentially growing Training Grant and an Institutional Research Grant cells. In all, 25 mg of whole-cell protein lysate from each sample was from American Cancer Society. subjected to Western blot analysis as described in Materials and methods. Equal loading is demonstrated by actin protein levels. Names of cell lines are indicated above and target proteins are indicated to the right of the blot. (a) Expression of Death Receptors 4 References and 5, FLIP and TRAIL in breast cancer cell lines. (b) Caspases 3 and 7, bcl-XL and bcl-2 protein levels in breast cancer cell lines were 1. Held J, Schulze-Osthoff K. Potential and caveats of TRAIL determined by Western blotting. in cancer therapy. Drug Resist Update. 2001;4:243–252. 2. Nagane M, Huang HJ, Cavenee WK. The potential of TRAIL for cancer chemotherapy. Apoptosis. 2001;6: DR4 and TRAIL was at least additive. Thus, combina- 191–197. tion of two nontoxic modalities, hTERT-DR4 and 3. Van Ophoven A, Ng CP, Patel B, Bonavida B, Belldegrun TRAIL, efficiently eliminates breast cancer cells with A. Tumor necrosis factor-related apoptosis-inducing ligand high hTERT promoter activity. (TRAIL) for treatment of prostate cancer: first results and In cells with low hTERT promoter activity (MDA-MB- review of the literature. Prostate Cancer Prostatic Dis. 231, HCC 1937 and T-47D, Figure 4b), introduction of 1999;2:227–233. hTERT-DR4 alone did not cause significant cell death 4. Keane MM, Rubinstein Y, Cuello M, et al. Inhibition of (striped bar). When hTERT-DR4 was combined with NF-kappaB activity enhances TRAIL mediated apoptosis in TRAIL (white bar), it did not increase toxicity over breast cancer cell lines. Breast Cancer Res Treat. 2000;64:211–219. TRAIL alone in these cell lines (black bar). 5. Keane MM, Ettenberg SA, Nau MM, Russell EK, MCF-7 cells were not affected by either hTERT-DR4 Lipkowitz S. Chemotherapy augments TRAIL-induced alone or by the combination of hTERT-DR4 with apoptosis in breast cell lines. Cancer Res. 1999;59:734–741. TRAIL, despite comparatively high hTERT promoter 6. Ashkenazi A, Dixit VM. Apoptosis control by death and activity. These cells were also relatively resistant to CMV- decoy receptors. Curr Opin Cell Biol. 1999;11:255–260. DR4, with or without addition of soluble TRAIL. To 7. Bernard D, Quatannens B, Vandenbunder B, Abbadie C. begin identifying the mechanisms of their resistance, we Rel/NF-kappaB transcription factors protect against tumor compared expression of the members of several families of necrosis factor (TNF)-related apoptosis-inducing ligand apoptotic regulators between MCF-7 cells and cell lines (TRAIL)-induced apoptosis by up-regulating the TRAIL sensitive to DR4 overexpression. decoy receptor DcR1. J Biol Chem. 2001;276:27322–27328. 8. LeBlanc HN, Ashkenazi A. Apo2L/TRAIL and its death Among the panel of apoptotic regulators tested, two and decoy receptors. Cell Death Differ. 2003;10:66–75. genes, bcl-2 and FLIP, were expressed at a higher level in 9. Roth W, Isenmann S, Nakamura M, et al. Soluble decoy MCF-7 cells as compared to the DR4-sensitive cell lines. receptor 3 is expressed by malignant gliomas and suppresses bcl-2 and FLIP have been shown to suppress TRAIL- CD95 ligand-induced apoptosis and chemotaxis. Cancer induced apoptosis18,20,21,36,37 and are good candidates for Res. 2001;61:2759–2765.

Cancer Gene Therapy DR4 efficiently kills breast cancer cells I Kazhdan and RA Marciniak 698 10. Seitz S, Wassmuth P, Fischer J, et al. Mutation analysis and myeloma cells: therapeutic implications. Oncogene. mRNA expression of trail-receptors in human breast cancer. 2002;21:5673–5683. Int J Cancer. 2002;102:117–128. 26. Secchiero P, Gonelli A, Carnevale E, et al. TRAIL promotes 11. Shin MS, Kim HS, Lee SH, et al. Mutations of tumor the survival and proliferation of primary human vascular necrosis factor-related apoptosis-inducing ligand receptor 1 endothelial cells by activating the Akt and ERK pathways. (TRAIL-R1) and receptor 2 (TRAIL-R2) genes in meta- Circulation. 2003;107:2250–2256. static breast cancers. Cancer Res. 2001;61:4942–4946. 27. Degli-Esposti MA, Dougall WC, Smolak PJ, et al. The 12. Hopkins-Donaldson S, Ziegler A, Kurtz S, et al. Silencing of novel receptor TRAIL-R4 induces NF-kappaB and protects death receptor and caspase-8 expression in small cell lung against TRAIL-mediated apoptosis, yet retains an incom- carcinoma cell lines and tumors by DNA methylation. Cell plete death domain. Immunity. 1997;7:813–820. Death Differ. 2003;10:356–364. 28. Blanc C, Deveraux QL, Krajewski S, et al. Caspase-3 is 13. Ozoren N, Fisher MJ, Kim K, et al. Homozygous deletion essential for procaspase-9 processing and cisplatin-induced of the death receptor DR4 gene in a nasopharyngeal cancer apoptosis of MCF-7 breast cancer cells. Cancer Res. cell line is associated with TRAIL resistance. Int J Oncol. 2000;60:4386–4390. 2000;16:917–925. 29. Devarajan E, Sahin AA, Chen JS, et al. Down-regulation of 14. Park WS, Lee JH, Shin MS, et al. Inactivating mutations of caspase 3 in breast cancer: a possible mechanism for KILLER/DR5 gene in gastric cancers. Gastroenterology. chemoresistance. Oncogene. 2002;21:8843–8851. 2001;121:1219–1225. 30. Papoff G, Hausler P, Eramo A, et al. Identification and 15. Bin L, Li X, Xu LG, Shu HB. The short splice form of characterization of a ligand-independent oligomerization Casper/c-FLIP is a major cellular inhibitor of TRAIL- domain in the extracellular region of the CD95 death induced apoptosis. FEBS Lett. 2002;510:37–40. receptor. J Biol Chem. 1999;274:38241–38250. 16. Bullani RR, Huard B, Viard-Leveugle I, et al. Selective 31. Faubion WA, Guicciardi ME, Miyoshi H, et al. Toxic bile expression of FLIP in malignant melanocytic skin lesions. salts induce rodent hepatocyte apoptosis via direct activa- J Invest Dermatol. 2001;117:360–364. tion of Fas. J Clin Invest. 1999;103:137–145. 17. Kelly MM, Hoel BD, Voelkel-Johnson C. Doxorubicin 32. Gu J, Kagawa S, Takakura M, et al. Tumor-specific pretreatment sensitizes prostate cancer cell lines to TRAIL transgene expression from the human telomerase reverse induced apoptosis which correlates with the loss of c-FLIP transcriptase promoter enables targeting of the therapeutic expression. Cancer Biol Ther. 2002;1:520–527. effects of the Bax gene to cancers. Cancer Res. 18. Kim Y, Suh N, Sporn M, Reed JC. An inducible pathway 2000;60:5359–5364. for degradation of FLIP protein sensitizes tumor cells 33. Koga S, Hirohata S, Kondo Y, et al. A novel telomerase- to TRAIL-induced apoptosis. J Biol Chem. 2002;277: specific gene therapy: gene transfer of caspase-8 utilizing the 22320–22329. human telomerase catalytic subunit gene promoter. Hum 19. Kim IK, Jung YK, Noh DY, et al. Functional screening of Gene Ther. 2000;11:1397–1406. genes suppressing TRAIL-induced apoptosis: distinct in- 34. Komata T, Kondo Y, Kanzawa T, et al. Caspase-8 gene hibitory activities of Bcl-XL and Bcl-2. Br J Cancer. therapy using the human telomerase reverse transcriptase 2003;88:910–917. promoter for malignant glioma cells. Hum Gene Ther. 20. Fulda S, Meyer E, Debatin KM. Inhibition of TRAIL- 2002;13:1015–1025. induced apoptosis by Bcl-2 overexpression. Oncogene. 35. Lin T, Huang X, Gu J, et al. Long-term tumor-free survival 2002;21:2283–2294. from treatment with the GFP-TRAIL fusion gene expressed 21. Lamothe B, Aggarwal BB. Ectopic expression of Bcl-2 and from the hTERT promoter in breast cancer cells. Oncogene. Bcl-xL inhibits apoptosis induced by TNF-related apopto- 2002;21:8020–8028. sis-inducing ligand (TRAIL) through suppression of cas- 36. Hietakangas V, Poukkula M, Heiskanen KM, et al. pases-8, 7, and 3 and BID cleavage in human acute Erythroid differentiation sensitizes K562 leukemia cells to myelogenous leukemia cell line HL-60. J Interferon Cytokine TRAIL-induced apoptosis by downregulation of c-FLIP. Res. 2002;22:269–279. Mol Cell Biol. 2003;23:1278–1291. 22. Ng CP, Bonavida B. X-linked inhibitor of apoptosis (XIAP) 37. Xiao C, Yang BF, Asadi N, Beguinot F, Hao C. Tumor blocks Apo2 ligand/tumor necrosis factor-related apoptosis- necrosis factor-related apoptosis-inducing ligand-induced inducing ligand-mediated apoptosis of prostate cancer cells death-inducing signaling complex and its modulation by c- in the presence of mitochondrial activation: sensitization by FLIP and PED/PEA-15 in glioma cells. J Biol Chem. overexpression of second mitochondria-derived activator of 2002;277:25020–25025. caspase/direct IAP-binding protein with low pl (Smac/ 38. Xue LY, Chiu SM, Oleinick NL. Staurosporine-induced DIABLO). Mol Cancer Ther. 2002;1:1051–1058. death of MCF-7 human breast cancer cells: a distinction 23. Nesterov A, Lu X, Johnson M, et al. Elevated AKT activity between caspase-3-dependent steps of apoptosis and the protects the prostate cancer cell line LNCaP from TRAIL- critical lethal lesions. Exp Cell Res. 2003;283:135–145. induced apoptosis. J Biol Chem. 2001;276:10767–10774. 39. Xue LY, Chiu SM, Oleinick NL. Photodynamic therapy- 24. Panka DJ, Mano T, Suhara T, Walsh K, Mier JW. induced death of MCF-7 human breast cancer cells: a role Phosphatidylinositol 3-kinase/Akt activity regulates c-FLIP for caspase-3 in the late steps of apoptosis but not for the expression in tumor cells. J Biol Chem. 2001;276:6893–6896. critical lethal event. Exp Cell Res. 2001;263:145–155. 25. Mitsiades CS, Mitsiades N, Poulaki V, et al. Activation of 40. Suyama E, Kawasaki H, Taira K. Identification of a caspase NF-kappaB and upregulation of intracellular anti-apoptotic 3-independent role of pro-apoptotic factor Bak in TNF- proteins via the IGF-1/Akt signaling in human multiple alpha-induced apoptosis. FEBS Lett. 2002;528:63–69.

Cancer Gene Therapy