Transfer of E2F-1 to Human Glioma Cells Results in Transcriptional Up-Regulation of Bcl-21

[CANCER RESEARCH 61, 6693–6697, September 15, 2001] Advances in Brief

Candelaria Gomez-Manzano, Paraskevi Mitlianga, Juan Fueyo, Ho-Young Lee, Min Hu, Kevin B. Spurgers, Tricia L. Glass, Dimpy Koul, Ta-Jen Liu, Timothy J. McDonnell, and W. K. Alfred Yung2 Departments of Neuro-Oncology [C. G-M., P. M., J. F., M. H., D. K., T. L. G., T-J. L., W. K. A. Y.], Thoracic and Head and Neck Medical Oncology [H-Y. L.], and Molecular Pathology [K. S., T. J. M.], The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe Boulevard, Houston, Texas 77030

Abstract characterized by abrogation of apoptosis as well as unregulated proliferation. It seems plausible that E2F-1 would protect proliferation by acti- Strong evidence exists to support the tenet that activation of E2F vating antiapoptotic genes. However, very few genes have been discov- transcription factors, via alterations in the p16-cyclin D-Rb pathway, is a ered that are both purely antiapoptotic and controlled by E2F-1. key event in the malignant progression of most human malignant gliomas. The oncogenic ability of E2F has been related to the E2F-mediated The Bcl-2 protein is one of the best known antiapoptotic molecules. up-regulation of several proteins that positively regulate cell proliferation. The bcl-2 gene was isolated from a common human follicular B-cell However, E2F may indirectly enhance proliferation by activating anti- lymphoma (8). Rather than inducing aberrant proliferation, like the vast apoptotic molecules. In this work, we sought to ascertain whether E2F-1- majority of oncogenes, Bcl-2 extends the life span of B cells by sup- mediated events involve the up-regulation of the antiapoptotic molecule pressing apoptosis (9). The relationship between Bcl-2 and the acquisi- Bcl-2. Western blot analyses showed up-regulation of Bcl-2 but not of tion of an antiapoptotic phenotype has been demonstrated in null Bcl-2 Bcl-xL by 24 h after the transfer of E2F-1. Northern blot studies showed and transgenic Bcl-2 animal models; Bcl-2-deficient mice showed abun- that transfer of E2F-1 also up-regulated Bcl-2 RNA. In support of these dant lymphocyte cell death (10). In addition, gain of function of Bcl-2 in findings and the concept that E2F-1 has a direct effect in the induction of mice renders their thymocytes resistant to apoptosis (11). The connection Bcl-2, we found a putative E2F binding site within the Bcl-2 sequence. between E2F-1 and Bcl-2 was established after the discovery that Bcl-2 Subsequent gel-mobility shift and supershift experiments involving the CTCCGCGC site in the bcl-2 promoter showed that E2F-1 bound Bcl-2. can protect cells from E2F-1- mediated apoptosis (12, 13). Transactivation experiments consistently showed that ectopic E2F-1 acti- We undertook this work to ascertain whether E2F-1 can activate vated responsive elements located in the ؊1448/؊1441 region in the P1 the expression of Bcl-2. We found that E2F-1 can up-regulate the promoter region of the bcl-2 gene. As expected, other members of the E2F expression of Bcl-2 at the protein and mRNA levels. We also family of transcription factors such as E2F-2 and E2F-4 also transacti- identified an E2F binding site in the promoter region of Bcl-2 and vated the bcl-2 promoter. Our results demonstrate that E2F-1 modulates showed that E2F-1 protein binds this E2F site. We additionally the expression of the antiapoptotic molecule Bcl-2 and suggest that up- demonstrated that E2F-1, E2F-2, and E2F-4 can transactivate regulation of Bcl-2 may favor the oncogenic role of E2F-1 and other responsive Bcl-2 elements. This report links for first time the members of the E2F family of transcription factors. regulation of Bcl-2 to E2F-1 transactivation function. Our results suggest that activation of antiapoptotic genes, in addition to up- Introduction regulation of positive modulators of cell cycle progression and The oncogenic property of E2F-1 is thought to be related to its ability DNA replication, may favor the oncogenic function of E2F-1. to regulate the expression of genes critical for cell proliferation. In Materials and Methods support of the tenet that E2F-1 can behave as an oncogene, the p16-Rb pathway, which negatively regulates the ability of E2F-1 to transactivate Cell Lines and Culture Conditions. The U-251 MG, U-87 MG, and T98 cell-cycle and DNA-replication-related genes, is partially inactive in the G cell lines were obtained from the American Type Culture Collection (Ma- vast majority of cancer cells including most malignant gliomas (reviewed nassas, VA). All of the cell lines were maintained in DMEM/F12 medium (1/1, in Ref. 1). Moreover, transfer of E2F-1 to quiescent fibroblasts results in v/v) supplemented with 10% heat-inactivated FCS in a humidified atmosphere containing 5% CO at 37°C. Synchronization procedures are described else- entry into the cell cycle and eventually replication of DNA (2). In 2 where (14). Briefly, T98 G cells were serum-starved for 3 days by culturing in addition, transgenic overexpression of E2F-1 favors the formation of skin MCDB-105 serum-free medium (Sigma Chemical Co., St. Louis, MO.) and tumors in vivo (3). However, inconsistent with its role as an oncogene, then stimulated into synchronous cell cycling progression by replacing the E2F-1 also induces programmed cell death in vitro and in vivo (4, 5). The medium with DMEM containing 10% FCS. ability of E2F-1 to induce either cell proliferation or cell death can be Adenoviral Vectors and Infection Conditions. The generation and char- tissue-specific; transgenic ablation of the expression of the E2F-1 gene in acterization of the recombinant-deficient adenovirus vectors carrying E2F-1 mice resulted in the spontaneous generation of tumors in certain tissues and the virus control Ad5CMV-pA have been described in detail elsewhere (5). and atrophy and lack of proliferation in others (6, 7). E2F-1-mediated The adenoviral vectors carrying the cDNA of E2F-2 or E2F-4 (15) were the proliferation is connected to the direct activation of positive regulators of generous gift of Dr. Joseph R. Nevins (Duke University Medical Center, cell cycle- and DNA-replication enzymes. The cancer phenotype is Durham, NC). Cell lines were cultured and infected as reported previously (16). We used a multiplicity of infection (defined as the ratio of the number of infectious virions to the number of susceptible cells) of 100. Received 11/29/00; accepted 7/26/01. Immunoblotting Assay. Total cell lysates were prepared by incubating The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with cells for1hat4°Cinradioimmunoprecipitation assay buffer [150 mM NaCl, 18 U.S.C. Section 1734 solely to indicate this fact. 1% Triton X-100, 1% sodium deoxycholate, 0.1% SDS, 20 mM EDTA, and 50 1 Supported by NIH Grants RO1 CA83127 (to W. K. A. Y.), RO1 CA80748 (to J. F.), mM Tris (pH 7.4)] at different times after infection. Then, 20 ␮g of protein and PO1 CA78778 (to T. J. M.). from each sample was subjected to 7% or 15% SDS-Tris-glycine gel electro- 2 To whom requests for reprints should addressed, at Department of Neuro-Oncology, Box 100, The University of Texas M. D. Anderson Cancer Center, 1515 Holcombe phoresis and transferred to a nitrocellulose membrane (Schleicher & Schuell Boulevard, Houston, TX 77030. Phone: (713) 794-1285; Fax: (713) 794-4999. Inc., Keene, NH). The membrane was blocked with Blotto-Tween [3% nonfat 6693

Downloaded from cancerres.aacrjournals.org on September 27, 2021. © 2001 American Association for Cancer Research. E2F-1 UP-REGULATES Bcl-2 milk, 0.05% Tween 20, 0.9% NaCl, and 50 mM Tris (pH 7.5)] and incubated structs and with 1 ␮g of pRL-CMV (containing the cDNA encoding Renilla with the following primary antibodies: mouse anti-E2F-1 (KH95; Santa Cruz luciferase) by using the FuGENE 6 transfection reagent (Roche Diagnostics Biotechnology Inc., Santa Cruz, CA), rabbit anti-E2F-2 (C-20; Santa Cruz Corp., Indianapolis, IN). Cells were infected with Ad5CMV-E2F-1, AdE2F-2, Biotechnology Inc.), rabbit anti-E2F-4 (C-108; Santa Cruz Biotechnology AdE2F-4, or AdCMV-pA 1 h later. After 24 h, lysates were collected and Inc.), mouse anti-Bcl-2 (C-2; Santa Cruz Biotechnology Inc.), mouse anti- assayed for luciferase activity by using the dual-luciferase reporter assay Bcl-x (PharMingen, San Diego, CA), mouse antihuman p53 (D0–7; DAKO), system (Promega). Luciferase activity from untreated control cells was used and mouse antihuman actin IgG (Amersham Corp., Arlington Heights, IL). for the background signal. All of the firefly luciferase values were normalized The secondary antibodies were horseradish peroxidase-conjugated antirabbit to the Renilla luciferase readout values and expressed as x-fold induction IgG, antimouse IgG (both from Amersham), and antigoat IgG (Santa Cruz relative to that of Ad5CMV-pA-infected cells. Biotechnology Inc.). The membranes were developed according to Amer- Luciferase reporter assays were also performed with the E2F-1 expression sham’s enhanced chemiluminescence protocol. vector pXCJL-E2F-1 (18). This plasmid contains an E2F-1 expression cassette Northern Blotting. U251 MG cells (5 ϫ 106) were seeded onto a 10-cm comprising the human cytomegalovirus promoter E2F-1 cDNA and the SV40 plate and allowed to adhere overnight. The next day, the cells were infected early polyadenylation signal. A similar construct, pXCJL-CMV-pA, lacking with Ad5CMV-E2F-1, AdE2F-2, AdE2F-4, or Ad5CMV-pA at a dose of 100 the E2F-1 cDNA was used as a control. multiplicity of infection. The total cellular RNA was isolated 36 h after Flow Cytometric Analyses of Cell Cycle. To measure DNA content, 106 infection by the acid-guanidium thiocyanate method. For the Northern blotting, cells were trypsinized, fixed in 70% cold ethanol, and incubated with pro- 15 ␮g of total cellular RNA prepared from each sample were subjected to pidium iodide (5 mg/ml) and RNase A (1 mg/ml) for 20 min at 37°C. All of electrophoresis on a 1% agarose gel containing 2% formaldehyde, stained with the measurements were made with an EPICS profile flow cytometer (Coulter ethidium bromide, photographed, transferred to a nylon membrane (Zetaprobe; Corp., Hialeah, FL) equipped with an air-cooled argon ion laser emitting 488 Bio-Rad Laboratories, Hercules, CA), and hybridized to an [␣-32P]dCTP- nm at 15 mW. Multicycle (Phoenix Flow System, San Diego, CA) program labeled Bcl-2 cDNA probe. Random priming was performed with the Prime It was used for data analysis. kit (Stratagene, La Jolla, CA), after which the membrane was washed in high-stringency conditions and autoradiographed for 24–48 h. Results Nuclear Extracts and Electrophoretic Gel Mobility Shift Analysis. Cell Cycle Analyses and Expression of Endogenous E2F-1 and U-251 MG cells (106 cells in 100-mm dishes) were washed with PBS and Bcl-2. Because E2F activity plays a major role in inducing cell resuspended in 400 ␮l of hypotonic buffer [20 mM HEPES (pH 7.9), 10 mM KCl, 0.2 mM EDTA, 1 mM DTT, 1 mM phenylmethylsulfonyl fluoride, 0.5 ␮g/ml progression from G1 to S phase, we assessed whether the up-regula- leupeptin, and 0.5 ␮g/ml aprotinin] and incubated at 4°C for 15 min. Cells were tion of E2F correlated temporally with up-regulation of the endoge- then lysed by adding 0.1% NP40 and vortexing; nuclei were pelleted and resus- nous Bcl-2 protein in T98 G cells as follows. T98 G cells were made pended in 200 ␮l of hypertonic buffer [20 mM HEPES (pH 7.9), 400 mM NaCl, 10 quiescent by serum starvation and then stimulated with serum and mM EDTA, 1 mM DTT, 1 mM phenylmethylsulfonyl fluoride, 0.5 ␮g/ml leupeptin, harvested at different time points from 0 to 28 h (Ref. 14; Fig. 1). and 0.5 ␮g/ml aprotinin]. After thorough mixing on a rotating wheel at 4°C for 15 After serum stimulation, we observed a more or less synchronized min, the suspension was centrifuged at 14,000 ϫ g for 5 min. The supernatant was cell-cycle progression with a predominant G0/G1 phase at 0 h, a used as the nuclear extract for the gel mobility shift assay. progressive accumulation of cells in the S phase of the cell cycle Bandshift assay were performed as follows: nuclear extracts (5 ␮g) were between 16 and 20 h, and an increased presence of the number of cells preincubated with 2 mg of the polydeoxynucleotide poly-dI:dC for 15 min at in the G -M phase by 24 h. As expected, in the Western blot analyses, 4°C and then incubated with labeled oligomer DNA for 15 min at 4°Cinthe 2 presence of 10 mM Tris-HCl (pH 7.5), 10 mM KCl, 1 mM EDTA, 20% glycerol, we detected an increase in the expression level of the endogenous E2F-1 protein 12 h after serum stimulation, immediately before cells 1mM DTT, and 5 mM MgCl2. An oligonucleotide (25-bp) containing the E2F(a) binding site (in capital letters; Ref. 14; sense: 5ЈatttaagCTCCGCGC- cctttctcaa3Ј; antisense: 3ЈtaaattcGAGGCGCGggaaagagtt5Ј) were synthesized (Life Technologies, Inc., Rockville, MD), annealed to each other, and forward- labeled with a [␥-32P]ATP using T4 polynucleotide kinase. For competition experiments, cold wild-type E2F(a) or mutant E2F(a) [mE2F(a), similar to E2F(a) but with a mutation in the binding site: sense: CTCCGATC; antisense: GAGGCTAG] oligonucleotides (10 ϫ or 100 ϫ molar excess) were mixed with the reaction mixtures before addition of the labeled probe. Reaction products were separated on 5% nondenaturing polyacrylamide gel (38:2 acrylamide:bis-acrylamide) in Tris-glycine electrophoresis buffer for2hat4°C at 180 V. The gel was then dried and autoradiographed with an amplified screen at Ϫ80°C. For antibody supershift experiments, nuclear extracts were preincubated with 4 ␮l of high-concentration anti-E2F-1 antibody (Geneka Biotechnology Inc., Montreal, Quebec, Canada) for1honicebefore the addition of labeled oligonucleotide probe. Plasmid Constructs. Bcl-2 promoter luciferase constructs were generated using the three-step cloning strategy described previously (17). Briefly, a 308-bp product was PCR amplified with primers generating XhoI and HindIII sites from a 7.8-kb genomic HindIII fragment of the bcl-2 gene. This 308-bp product covers a region 5Ј to the AccI site that is just 5Ј of the initiation codon. This product was cloned into the XhoI and HindIII sites of the pGL3 basic luciferase vector (Promega Corp., Madison, WI) producing pGL3–308 bcl-2. Next, a 486-bp XhoI, AccI restriction fragment was cloned into pGL3–308 bcl-2 producing pGL3–748 bcl-2. Finally, a 2100-bp XhoI fragment was Fig. 1. E2F-1 and Bcl-2 expression during synchronized cell-cycle progression. For this experiment cells were harvested at the same times for both flow cytometric analyses cloned into pGL3–748 bcl-2 producing pGL3–2.8 bcl-2. (Promoter sequence and immunoblot assays (see “Material and Methods” and Ref. 14). A, the flow cytometric of bcl-2: GenBank accession nos. X51898 and M13994). plots display the cell-cycle distribution of quiescent T98 G glioma cells after serum Luciferase Reporter Assays. U-251 MG cells were seeded in 6-well stimulation. The axis represent number of cells on the ordinate and DNA content on the dishes at a density of 5 ϫ 105 cells/well and cultured for 24 h in DMEM/F12 abscissa; B, Western blots show the expression of the endogenous E2F-1 and Bcl-2 ␮ proteins at the indicated points after serum stimulation. The level of the protein actin is medium containing 10% FCS. Cells were then cotransfected with 5 gof showed as a loading control. Note that, as expected, increase in level of expression of the pGL3–2.8 Bcl-2-Luc, pGL3–748 Bcl-2-Luc, or pGL3–308 Bcl-2-Luc con- E2F-1 protein correlated temporally with cell entry into S phase. 6694

Downloaded from cancerres.aacrjournals.org on September 27, 2021. © 2001 American Association for Cancer Research. E2F-1 UP-REGULATES Bcl-2 accumulated in S phase, and continued increase up until 24 h (exit of cells from S phase). The expression level of Bcl-2 protein began to increase in parallel with the increase in the expression level of E2F-1. These results suggested that the expression of E2F-1 and Bcl-2 molecules might be connected. Transfer of E2F-1 to Glioma Cells Enhances Expression of Bcl-2. Next, we investigated whether overexpression of E2F-1 could up-regulate Bcl-2 in human glioma cells. Analysis of protein extracts from E2F-1-transduced U-251 MG cells, growing asynchronously, revealed large increases in Bcl-2 protein levels (Fig. 2). Consistent with the previous analyses of the endogenous E2F-1 and Bcl-2 expression in a synchronized population of human glioma cells, in these experiments, up-regulation of Bcl-2 was evident by 16 h after transfer of an exogenous E2F-1 and increased progressively, achieving their maximum by 24–48 h after infection (Fig. 2). Interestingly, extending the Western blot analyses to include Bcl-xL showed that the transfer Fig. 3. Up-regulation of bcl-2 by E2F. A, Northern blot of Bcl-2 RNA expression in of E2F-1 did not up-regulate Bcl-x , a member of the Bcl-2 family of U-251 MG cells 36 h after infection with Ad5CMV-E2F-1, AdE2F-2, AdE2F-4, or control L (CMV) constructs, or medium only (Mock). Consistent with the Western blot analyses, proteins, with antiapoptotic activity. The E2F-1-mediated up-regula- transfer of any of the three members of the E2F family of transcription factors resulted in tion of Bcl-2 was also observed in U-87 MG cells. Moreover, trans- up-regulation of Bcl-2 RNA. B, verification that equal amounts of total RNA had been duction of U-251 MG and U-87 MG cells with adenoviral vectors examined. Ethidium bromide staining of the agarose gel is presented as loading control. carrying E2F-2 or E2F-4, two other members of the E2F family of transcription factors of which their DNA-binding domains share a high homology with that of E2F-1, also produced up-regulation of sequence (TTTCGCGC) to seek sequence homology within the bcl-2 Bcl-2 (Fig. 2). These results suggest that the E2F family of transcrip- gene sequence, we found no homologous E2F sites in the bcl-2 tion factors was involved in the modulation of Bcl-2 expression in promoter. However, when we examined the bcl-2 promoter sequence these gliomas cells. using another E2F binding sequence (CTCCGCGC) [E2F(a)] present E2F-1 Up-Regulates Bcl-2 RNA. Because E2F-1 is a transcription in the human p21/cip1 promoter (14), we identified a 100% homol- factor, we hypothesized that E2F-1-mediated up-regulation of Bcl-2 ogous sequence in the Ϫ1448/Ϫ1441 P1 region of the bcl-2 promoter. should be associated with a concomitant increase in bcl-2 gene ex- E2F Proteins Bind to the Putative E2F Site. After identifying the pression. Northern blot analyses of U-251 MG cells infected with ade- putative E2F binding site in the bcl-2 promoter, we confirmed that novirus constructs containing E2F-1, E2F-2, or E2F-4 revealed increases nuclear extract-derived proteins could form DNA-protein complexes in Bcl-2 RNA levels 36 h after infection (Fig. 3), indicating that E2F-1 with the E2F probe (Fig. 4). In competition experiments with nuclear either transcriptionally activated Bcl-2 or stabilized Bcl-2 RNA. extracts from U-251 MG human glioma cells, binding of the cellular Identification of a Putative E2F Binding Site in the Human proteins to the E2F probe was competed out by the unlabeled wild- Bcl-2 Promoter. The presence of E2F-responsive elements in the type probe but not by the unlabeled mutated probe. Addition of an bcl-2 promoter would be a strong indication that E2F regulates Bcl-2 anti-E2F-1 antibody supershifted the complex, indicating that the at a transcriptional level. When we used the consensus E2F binding E2F-1 protein was forming part of the complexes (Fig. 4). The Bcl-2 Promoter Is Regulated by E2F. Although these findings provide strong evidence that E2F-1 transcriptionally regulates bcl-2, the ultimate test is to show that the exogenous E2F-1 protein can transactivate a chimeric construct encompassing the putative responsive elements of the E2F protein in the bcl-2 promoter and a reporter gene. For these experiments, we transfected U-251 MG glioma cells with one of three plasmid constructs, each containing different lengths of the bcl-2 promoter linked to a luciferase reporter gene: pGL3–2.8 bcl, pGL3–748 bcl-2, or pGL3–308 bcl-2. After transfection, the cells were infected with Ad5CMV-E2F-1 and examined for luciferase activity 24 h later. Cells that had been transfected with pGL3–2.8 bcl-2, which contained both the P1- and P2-responsive elements, showed 70.8 Ϯ 22 times the activity of the bcl-2 promoter as that of the control (Fig. 5). In contrast, no significant induction of luciferase activity was noted in cells that had been transfected with the shorter constructs that contained only the P2-responsive elements (Fig. 5). To ascertain whether the up-regulation of bcl-2 was attributable to the adenoviral-mediated high level of expression of the Fig. 2. Western blot analyses of Bcl-2 protein expression in human glioma cells exogenous E2F-1, we performed a similar luciferase assay transiently expressing exogenous E2F-1, E2F-2, or E2F-4. A, Bcl-2, Bcl-xL, and p21 protein levels expressed by U-251 MG glioma cells after infection with adenovirus carrying the E2F-1 transfecting glioma cells with the plasmid construct pXCJL-E2F-1. coding sequence, assayed at the indicated time points. The up-regulation of Bcl-2 in these These experiments showed that cells that were cotransfected with cells was evident within 16 h after infection with Ad5CMV-E2F-1. Transfer of E2F-1 did pGL3–2.8 bcl-2 displayed an increase in the luciferase activity that not result in an increased level of expression of Bcl-xL, another antiapoptotic molecule. As described previously (14), p21 protein levels increased after E2F-1 transfer. B and C, was 17.2, 4, and 3.4 times higher than that of the pXCJL-CMV- Bcl-2 protein levels expressed by U-251 MG (B) and U-87 MG (C) glioma cells 48 h after pA-transfected cells in three independent experiments. These results infection with adenoviruses carrying the E2F-1, E2F-2, and E2F-4 coding sequences or with the adenovirus control (CMV). Level of expression of actin is shown as a loading suggest that E2F-1 transactivates the bcl-2 promoter through the control. CTCCGCGC site located in the P1 promoter region. Transfer of 6695

transcriptionally responsive (reviewed in Ref. 22). In this regard, the transcriptional regulators of Bcl-2 that have been identified include the p53 tumor suppressor gene product, the products of the cellular and viral myb genes, and the product of the Wilms’ tumor gene wt1 (23). Overexpression of Bcl-2 is known to block E2F-1-induced apoptosis (12, 13). In the system described by Strom et al. (12), overexpression of E2F-1 overrode the survival functions provided by granulocyte colony-stimulating factors and trigger apoptosis. However, coexpression of both E2F-1 and Bcl-2 resulted in the acquisition of an antiapoptotic phenotype. Lind et al. (13) reported that cells lacking bcl-2 expression respond to growth factors withdrawal by liberating E2F-1 from inactive complexes resulting in cell death. Another aspect of the interplay between E2F-1 and Bcl-2 is the ability of Bcl-2 to retard the entry of cells into S phase (24, 25). Vairo et al. (26) reported

that Bcl-2 expression delays E2F-1 accumulation during G1 progression. These authors suggest that gain of E2F-1 function would dimin- ish the Bcl-2 cycle-inhibitory activity and potentially enhance its oncogenic impact. Taken collectively, these observations and our data imply the existence of a feedback regulatory loop that controls the expression and cell effect of E2F-1 and Bcl-2. Fig. 4. E2F-1 binding to the bcl-2 promoter. Electrophoretic mobility-shift assays were We have shown previously that transfer of E2F-1 to glioma cells performed with a radiolabeled 25-bp oligonucleotide containing the E2F-1 binding region [E2F(a)] (Lane 1, labeled oligonucleotide probe only). U-251 MG nuclear extracts were resulted in apoptosis (5). In this report, the experiments were per- mixed with the labeled E2F(a) probe (Lanes 2 and 7) or, for competition assays, with labeled E2F(a) and a 10- or 100-fold excess of unlabeled probe [E2F(a)] (Lanes 3 and 4) supershift ء .(or mutant probe lacking E2F binding activity [mE2F(a)] (Lanes 5 and 6 caused by addition of an antibody to E2F-1 (Lane 8).

E2F-2 or E2F-4 also increased luciferase activity but, again, only in cells that had been transfected with the bcl-2 construct that contained both the P1 and P2 regions. These results strongly suggest that the sequence CTCCGCGC located in the Ϫ1448/Ϫ1441 position of the bcl-2 promoter is an E2F binding site.

Discussion Our results showed that the adenoviral transfer of E2F-1 to glioma cells led to the up-regulation of Bcl-2, which suggests that E2F-1 favors proliferation through the induction of antiapoptotic genes in addition to being a positive regulator of cell cycle- and DNA replication-related genes. The region within the bcl-2 promoter that was responsive to E2F-1 was found to be in the P1 promoter region (17, 19). Electrophoretic mobility assays showed that the E2F-1 protein interacted directly with the E2F binding site. Moreover, an exogenous wild-type E2F-1 was able to up-regulate a reporter gene driven by the bcl-2 promoter. Evidence that other members of the E2F family of transcription factors, like E2F-2 and E2F-4, also modulated the expression of Bcl-2 suggests that the up-regulation of Bcl-2 is not part of a nonspecific reaction of the cells to the overexpression of exogenous E2F-1. The hypothesis that E2F-1 specifically up-regulates Bcl-2 is also supported by our finding that E2F-1 did not up-regulate other members of the Bcl-2 family of proteins such as Bcl-xL. Consistent with these results, Dong et al. (20) reported finding increased levels of the Bcl-2 protein, assessed by Western blotting, after adenovirus-mediated E2F-1 transfer in a melanoma cell line. In addition, Mu¨ller et al. (21), using microarrays technology, showed that retrovirally mediated transfer Fig. 5. E2F-1 transactivates Bcl-2-responsive elements. A, schematic representation of of E2F-1 to U2OS cells resulted in up-regulation of Bcl-2. Bcl-2-luciferase reporter constructs. Note that three constructs comprise the P2 region but We showed here that E2F-1 transcriptionally activates bcl-2. The only the pGL3–2.8 bcl2 construct encompasses the P1 region. B, U-251 MG glioma cells were cotransfected with the bcl-2 reporter constructs and the pRL-CMV vector, and 1 h expression of pro- and antiapoptotic molecules is regulated by mul- later the cells were treated with adenoviral vectors carrying E2F-1, E2F-2, E2F-4, or an tiple mechanisms. For example, the Bcl-2 family is regulated by empty expression cassette (CMV). Luciferase activity was determined 24 h after the cytokines and other death-survival signals at different levels. Some of infection. Nontransfected cells were used as the background. All values were normalized for expression of Renilla luciferase, which served as internal control for transfection these signals involve post-translational modification or conforma- efficiency and expressed as x-fold induction relative to that of the adenovirus control- tional changes. In addition, several members of the Bcl-2 family of infected cells (equal to 1). Each experiment was performed at least three times in duplicate. Shown are means of normalized luciferase measurements; bars, Ϯ SE. The proteins are transcriptionally regulated. For example, Bax seems to be E2F-mediated induction of the bcl-2 responsive elements within the pGL3–2.8 bcl-2 transcriptionally activated by p53, because Bcl-xL and Bcl-2 are also construct was at least 30-fold higher than the control-mediated induction. 6696

Downloaded from cancerres.aacrjournals.org on September 27, 2021. © 2001 American Association for Cancer Research. E2F-1 UP-REGULATES Bcl-2 formed under similar conditions to those reported previously (5). The 4. Wu, X., and Levine, A. J. p53 and E2F-1 cooperate to mediate apoptosis. Proc. Natl. up-regulation of Bcl-2 (antiapoptotic molecule and negative regulator Acad. Sci. USA, 91: 3602–3606, 1994. 5. Fueyo, J., Gomez-Manzano, C., Yung, W. K. A., Liu, T. J., Alemany, R., McDonnell, of cell cycle) by E2F-1 (positive regulator of cell and proapoptotic) T. J., Shi, X., Rao, J. S., Levin, V. A., and Kyritsis, A. P. Overexpression of E2F-1 seems to be counterintuitive. However, it is often the case for tran- in glioma triggers apoptosis and suppresses tumor growth in vitro and in vivo. Nat. scription factors involved in cancer that produce both positive and Med., 4: 685–690, 1998. 6. Field, S. J., Tsai, F. Y., Kuo, F., Zubiaga, A. M., Kaelin, W. G., Jr., Livingston, D. M., negative survival signals. In this regard, c-myc and E1A are able to Orkin, S. H., and Greenberg, M. E. E2F-1 functions in mice to promote apoptosis and induce proliferation and apoptosis signals. Importantly, the regions of suppress proliferation. Cell, 85: 549–561, 1996. E1A and c-myc that are responsible for transformation are also nec- 7. Yamasaki, L., Jacks, T., Bronson, R., Goillot, E., Harlow, E., and Dyson, N. J. Tumor induction and tissue atrophy in mice lacking E2F-1. Cell, 85: 537–548, 1996. essary for induction of apoptosis, indicating that the growth-promot- 8. Tsujimoto, Y., Gorham, J., Cossman, J., Jaffe, E., and Croce, C. M. The t(14;18) ing activities of c-myc and E1A are linked to its death-inducing chromosome translations involved in B-cell neoplasms result from mistakes in VDJ properties (recently reviewed in Ref. 27). Transfer of p53 to cancer joining. Science (Wash. DC), 229: 1390–1393, 1985. 9. Hockenbery, D., Nunez, G., Milliman, C., Schreiber, R. D., and Korsmeyer, S. J. cells results in the up-regulation of pro-apoptotic molecules such as Bcl-2 is an inner mitochondrial membrane protein that blocks programmed cell death. Bax, but also in the up-regulation of molecules that negatively regu- Nature (Lond.), 348: 334–336, 1990. lates its apoptotic function, such as hdm-2 (28). The up-regulation of 10. Veis, D. J., Sorenson, C. M., Shutter, J. R., and Korsmeyer, S. J. Bcl-2-deficient mice demonstrate fulminant lymphoid apoptosis, polycystic kidneys, and hypopigmented Bcl-2 is not the only paradoxical effect of the transfer of E2F-1 to hair. Cell, 75: 229–240, 1993. cancer cells. E2F-1 induces the expression of several genes related to 11. Sentman, C. L., Shutter, J. R., Hockenbery, D., Kanagawa, O., and Korsmeyer, S. J. cell-cycle progression but, at the same time, up-regulates proteins that bcl-2 inhibits multiple forms of apoptosis but not negative selection in thymocytes. Cell, 67: 879–888, 1991. negatively influence cell cycle progression, including p18INK4c (21). 12. Strom, D. K., Cleveland, J. L., Chellappan, S., Nip, J., and Hiebert, S. W. E2F-1 and Because transfer of E2F-1 results in the production of several lines of E2F-3 are functionally distinct in their ability to promote myeloid cell cycle progres- decision, the fate of an E2F-1-treated cell may rely on the status and sion and block granulocyte differentiation. Cell Growth Differ., 9: 59–69, 1998. 13. Lind, E. F., Wayne, J., Wang, Q. Z., Staeva, T., Stolzer, A., and Petrie, H. T. expression of other genes. For instance, the level of E2F-3 may be Bcl-2-induced changes in E2F regulatory complexes reveal the potential for inte- important in the proliferation/apoptosis decision-making process. grated cell cycle and cell death functions. J. Immunol., 162: 5374–5349, 1999. Thus, it has been postulated that E2F-1 and E2F-3 contribute to a pool 14. Hiyama, H., Iavarone, A., and Reeves, S. A. Regulation of the cdk inhibitor p21 gene during cell cycle progression is under the control of the transcription factor E2F. of free E2F activity that activates inappropriate proliferation once it Oncogene, 16: 1513–1523, 1998. reaches one critical threshold level (proliferation threshold) but apo- 15. DeGregori, J., Leone, G., Miron, A., Jakoi, L., and Nevins, J. R. Distinct roles for E2F ptosis once it exceeds a second, higher threshold level (29). The “free” proteins in cell growth control and apoptosis. Proc. Natl. Acad. Sci. USA, 94: 7245–7250, 1997. E2F-3 (like the free E2F-1) activity is arguably high in cancer cells 16. Gomez-Manzano, C., Fueyo, J., Kyritsis, A. P., McDonnell, T. J., Steck, P. A., Levin, with abnormally regulated retinoblastoma pathway, playing a role in V. A., and Yung, W. K. Characterization of p53 and p21 functional interactions in the neoplastic phenotype. Increasing the level of E2F-1 in this back- glioma cells en route to apoptosis. J. Natl. Cancer Inst., 89: 1036–1044, 1997. 17. Young, R. L., and Korsmeyer, S. J. A negative regulatory element in the bcl-2 ground will surpass the proliferative threshold and trigger apoptosis. 5Ј-untranslated region inhibits expression from an upstream promoter. Mol. Cell. Finally, despite the fact that overexpression of bcl-2 also follows Biol., 13: 3686–3697, 1993. retroviral- (21) and plasmid-mediated transfer of E2F-1, we cannot 18. Hunt, K. K., Deng, J., Liu, T. J., Wilson-Heiner, M., Swisher, S. G., Clayman, G., and Hung, M. C. Adenovirus-mediated overexpression of the transcription factor E2F-1 completely rule out the possibility that the induction of bcl-2 is only induces apoptosis in human breast and ovarian carcinoma cell lines and does not seen because of the use of an adenoviral system where E2F can be require p53. Cancer Res., 57: 4722–4726, 1997. highly overexpressed. 19. Seto, M., Jaeger, U., Hockett, R. D., Graninger, W., Bennett, S., Goldman, P., and Korsmeyer, S. J. Alternative promoters and exons, somatic mutation and deregulation The full spectrum of E2F target genes remains to be determined. of the Bcl-2-Ig fusion gene in lymphoma. EMBO J., 7: 123–131, 1988. The E2F-1 up-regulation of Bcl-2 is consistent with the up-regulation 20. Dong, Y. B., Yang, H. L., Elliott, M. J., Liu, T. J., Stilwell, A., Atienza, C., Jr., and of other molecules with antiapoptotic activities such as p21 (16, 30). McMasters, K. M. Adenovirus-mediated E2F-1 gene transfer efficiently induces apoptosis in melanoma cells. Cancer (Phila.), 86: 2021–2033, 1999. Interestingly, the p21 and bcl-2 promoters share the same E2F-1 21. Mu¨ller, H., Bracken, A. P., Vernell, R., Moroni, C., Christians, F., Grassilli, E., binding site (14). Because expression of these two molecules may Prosperini, E., Vigo, E., Oliner, J. D., and Helin, K. E2Fs regulate the expression of enhance the resistance of cells to apoptosis and, therefore, favor the genes involved in differentiation, development, proliferation, and apoptosis. Genes Dev., 15: 267–285, 2001. ability of E2F-1 to behave as an oncogene, it is relevant to mention 22. Gross, A, McDonnell, J. M., and Korsmeyer, S. J. Bcl-2 family members and the that Bcl-2 and p21 are overexpressed in the majority of malignant mitochondria in apoptosis. Genes Dev., 13: 1899–1911, 1999. gliomas (31, 32). In this regard, it will also be interesting to see 23. Smith, M. D., Ensor, E. A., Coffin, R. S., Boxer, L. M., and Latchman, D. S. Bcl-2 transcription from the proximal P2 promoter is activated in neuronal cells by the whether tumors generated by the transgenic expression of E2F-1 Brn-3a POU family transcription factor. J. Biol. Chem., 273: 16715–16722, 1998. overexpress Bcl-2 and p21. Because E2F-1 is able to induce apopto- 24. O’Reilly, L. A., Huang, D. C., and Strasser, A. The cell death inhibitor Bcl-2 and its sis, this hypothesis also predicts that the apoptosis observed in the homologues influence control of cell cycle entry. EMBO J., 15: 6979–6990, 1996. 25. Linette, G. P., Li, Y., Roth, K., and Korsmeyer, S. J. Cross talk between cell death and null-Bcl-2 mice might be attributable, at least in part, to E2F-1. cell cycle progression: BCL-2 regulates NFAT-mediated activation. Proc. Natl. Acad. Sci. USA, 93: 9545–9552, 1996. Acknowledgments 26. Vairo, G., Soos, T. J., Upton, T. M., Zalvide, J., DeCaprio, J. A., Ewen, M. E., Koff, A., and Adams, J. M. Bcl-2 retards cell cycle entry through p27Kip1, pRB relative We thank Polly Y. Lee (Department of Neuro-Oncology, The University of p130, and altered E2F regulation. Mol. Cell. Biol., 20: 4745–4753, 2000. Texas M. D. Anderson Cancer Center) for technical assistance and Christine 27. Breckenridge, D. G., and Shore, G. C. Regulation of apoptosis by E1A and Myc Wogan (Department of Scientific Publications, The University of Texas M. D. oncoproteins. Crit. Rev. Eukaryot. Gene Expr., 10: 273–280, 2000. 28. Oren, M. Regulation of the p53 tumor suppressor protein. J. Biol. Chem., 274: Anderson Cancer Center) for editorial assistance. 36031–36034, 1999. 29. Ziebold, U., Reza, T., Caron, A., and Lees J. A. E2F3 contributes both to the References inappropriate proliferation and to the apoptosis arising in Rb mutant embryos. Genes 1. Alemany, R., Gomez-Manzano, C., Balague, C., Yung, W. K. A., Curiel, D. T., Dev., 15: 386–391, 2001. Kyritsis, A. P., and Fueyo, J. Gene therapy for gliomas: molecular targets, adenoviral 30. Polyak, K., Waldman, T., He, T. C., Kinzler, K. W., and Vogelstein, B. Genetic vectors, and oncolytic adenovirus. Exp. Cell Res., 252: 1–12, 1999. determinants of p53-induced apoptosis and growth arrest. Genes Dev., 10: 1945– 2. Johnson, D. G., Schwarz, J. K., Cress, W. D., and Nevins, J. R. Expression of 1952, 1996. transcription factor E2F1 induces quiescent cells to enter S phase. Nature (Lond.), 31. Alderson, L. M., Castleberg, R. L., Harsh, G. R., IV, Louis, D. N., and Henson, J. W. 365: 349–352, 1993. Human gliomas with wild-type p53 express bcl-2. Cancer Res., 55: 999–1001, 1995. 3. Pierce, A. M., Fisher, S. M., Conti, C. J., and Johnson, D. G. Deregulated expression 32. Jung, J. M., Bruner, J. M., Ruan, S., Langford, L. A., Kyritsis, A. P., Kobayashi, T., of E2F1 induces hyperplasia and cooperates with ras in skin tumor development. Levin, V. A., and Zhang, W. Increased levels of p21WAF1/Cip1 in human brain Oncogene, 16: 1267–1276, 1998. tumors. Oncogene, 11: 2021–2028, 1995.

6697

Downloaded from cancerres.aacrjournals.org on September 27, 2021. © 2001 American Association for Cancer Research. Transfer of E2F-1 to Human Glioma Cells Results in Transcriptional Up-Regulation of Bcl-2

Candelaria Gomez-Manzano, Paraskevi Mitlianga, Juan Fueyo, et al.

Cancer Res 2001;61:6693-6697.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/61/18/6693

Cited articles This article cites 32 articles, 16 of which you can access for free at: http://cancerres.aacrjournals.org/content/61/18/6693.full#ref-list-1

Citing articles This article has been cited by 6 HighWire-hosted articles. Access the articles at: http://cancerres.aacrjournals.org/content/61/18/6693.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/61/18/6693. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.