Int. J. Cancer: 115, 202–213 (2005) ' 2005 Wiley-Liss, Inc.
Interleukin-6 induces transcriptional activation of vascular endothelial growth factor (VEGF) in astrocytes in vivo and regulates VEGF promoter activity in glioblastoma cells via direct interaction between STAT3 and Sp1 Se´bastien Loeffler1, Be´renge`re Fayard1, Joachim Weis1,2 and Jakob Weissenberger1* 1 Division of Neuropathology, Institute of Pathology, University of Bern, Bern, Switzerland 2 Institute of Neuropathology, University Hospital, Aachen, Germany Interleukin-6 (IL-6) expression is strongly correlated with the degree of human glioma malignancy and necessary for tumor formation in a mouse model of spontaneous astrocytomas. Yet, exactly how IL-6 contributes to malignant progression of these brain tumors is still unclear. We have scrutinized the mechanism of transcriptional activation of vascular endothelial growth factor (VEGF) expression by IL-6 in the mouse brain and in glioblastoma cells. We demonstrate here that IL-6 drives transcriptional upregulation of VEGF in astrocytes in vivo using glial fibrillary acidic protein (GFAP)-IL-6/VEGF-green fluorescent protein (GFP) double transgenic mice. We further show that IL-6-induced VEGF transcription and VEGF secretion by human glioblastoma cells is dependent on signal transducer and activator of transcription 3 (STAT3). By progressive 50 -deletion analysis we defined the minimal VEGF promoter region for IL-6-responsiveness to nucleotides 288/250. Surprisingly, this promoter region is rich in GC-boxes and does not contain STAT3 binding elements. Electrophoretic mobility shift and supershift assays revealed binding of Sp1 and Sp3 to the 288/250 element upon IL-6 stimulation. Interestingly, preincubation with STAT3 antibody prevented the binding of Sp1 and Sp3 to the 288/250 element, indicating that STAT3 is involved in IL-6-driven Sp1/Sp3 protein-DNA complex formation. Physical interaction of STAT3 and Sp1 was demonstrated by coimmunoprecipitation. The functional relevance of the STAT3/ Sp1 association was corroborated by transient transfection experiments, which showed that overexpression of constitutively active STAT3 increased the minimal VEGF promoter activity. Taken together, our study suggests that IL-6 promotes tumor angiogenesis in gliomas and describes a novel transcriptional activation mechanism for STAT3 in the context of a STAT3 binding element (SBE)-free promoter. ' 2005 Wiley-Liss, Inc. Key words: IL-6; VEGF; STAT3; Sp1; glioma; angiogenesis
Gliomas are the most frequent primary brain tumors. Among these tumors, glioblastomas (GBMs) are the most aggressive. They are highly invasive and considered to be among the deadliest of human cancers.1 A hallmark of glioma progression is that transition from lower-grade tumors to GBMs is accompanied by a striking induction of angiogenesis.1 Therefore, inhibition of angiogenesis may represent a potentially promising strategy in the treatment of gliomas. Angiogenesis, the formation of new blood vessels, is a crucial process for solid tumors in order to acquire nutrients for continued growth and metastatic spread.2 Vascular endothelial growth factor (VEGF) is one of the most potent inducers of angiogenesis identified so far. The ability of a VEGF antisense construct to reduce angiogenicity and tumorigenicity of human U87MG GBM cells suggested that VEGF plays a key role in glioma development.3 In human brain cancers, a close correlation was found between VEGF expression, tumor vascularization and grade of malignancy,4 with 95% of GBMs staining positive for VEGF.5 The signals were localized in both perinecrotic and vital tumor areas suggesting that both hypoxia and other stimuli may control the expression of VEGF in gliomas.5 Interleukin-6 (IL-6) may represent one of these stimuli, since IL-6 is a potential stimulator of VEGF expression6 and is also implicated in the pathogenesis of gliomas. Indeed, the IL-6 gene is amplified in primary glioblastomas7 and human GBM cells release Publication of the International Union Against Cancer
biologically active IL-6 in vitro and in vivo.8 IL-6 has been shown to promote growth of U87MG cells by an autocrine mechanism.9 Moreover, a strong and significant association between glioma aggressiveness and IL-6 gene expression was found.10 Finally, by abrogation of IL-6 gene expression in glial fibrillary acidic protein (GFAP)-viral src kinase (v-src) transgenic mice that phenocopy the neuropathology of human astrocytomas,11 we have recently shown that IL-6 is crucial for glioma development.12 IL-6 is a member of the neuropoietic cytokine family, which also includes, among others, ciliary neurotrophic factor (CNTF) and leukemia inhibitory factor (LIF). IL-6 not only promotes the survival of certain neuronal cell populations13 but also the reactive transformation of astrocytes.14,15 The complex of IL-6 and its primary receptor, IL-6 receptor a (gp80), binds to the transmembrane signal transducer gp130 and activates gp130-associated janus kinases (JAKs) leading to the recruitment and phosphorylation of the signal transducer and activator of transcription 3 (STAT3). Active STAT3 dimerizes, translocates to the nucleus and binds to enhancer elements of target genes, thereby inducing transcriptional activation.16 IL-6 can also activate NF-IL6 (nuclear factor interleukin-6 or C/EBPb), Ras-GTPases/mitogen-activated protein kinase (Ras/MAPK) and PI(3) kinase/Akt pathways.17–20 The diversity of IL-6 signaling explains its pleiotropic function. In fact, IL-6 acts on a wide variety of cells regulating immune response, acute phase reaction and hematopoiesis21 and promotes the growth of several human cancers such as prostate carcinoma, multiple myeloma, colorectal cancer, renal cell carcinoma and Kaposi’s carcinoma.22–26 In cervical cancer, it has been shown recently that IL-6 promotes tumor growth by VEGF-dependent angiogenesis.27 Combined, these findings suggest that elevated IL-6 levels in the brain could promote glial tumor progression via stimulation of angiogenesis. Therefore, our study was conducted to investigate the mechanism of IL-6-induced VEGF transcription in the mouse brain, in glioblastoma cells and in NIH3T3 model cells.
S. Loeffler’s present address is: IFOM-FIRC Institute of Molecular Oncology, Via Adamello 16, 20139 Milano, Italy. Abbreviations: AP2, activator protein 2; BCA, bicinchoninic acid; CNTF, ciliary neurotrophic factor; ECL, enhanced chemiluminescence; DTT, dithiothreitol; ECL, enhanced chemiluminescence; EGF, epidermal growth factor; EMSA, electrophoretic mobility shift assay; GBM, glioblastoma; GFAP, glial fibrillary acidic protein; GFP, green fluorescent protein; IgG, immunoglobulin E; IL-6, interleukin-6; JAK, janus kinase; LIF, leukemia inhibitory factor; NeuN, neuronal nuclei; NF-IL6, nuclear factor interleukin-6 or C/EBPb; Egr-1, early growth response factor-1; nt, nucleotide; PTEN, phosphatase and tensin homolog deleted from chromosome 10; Ras/ MAPK, Ras-GTPases/mitogen-activated protein kinase; RIPA, radioimmunoprecipitation; STAT, signal transducer and activator of transcription; TBE, Tris-Borate-EDTA; SRL; VEGF, vascular endothelial growth factor; v-src, viral src kinase. Grant sponsor: Bernese Cancer League; Grant sponsor: Dr. med. h.c. Erwin Braun Foundation, Basel, Switzerland. *Correspondence to: Division of Neuropathology, Institute of Pathology, University of Bern, Murtenstrasse 31, CH-3010 Bern, Switzerland. Fax: þ41-31-632-9872. E-mail:
[email protected] Received 25 May 2004; Accepted after revision 29 September 2004 DOI 10.1002/ijc.20871 Published online 1 February 2005 in Wiley InterScience (www.interscience. wiley.com).
MECHANISM OF IL-6 INDUCED VEGF EXPRESSION
203
FIGURE 1 – IL-6 induces VEGF transcription in vivo. (a) Control brain section of a GFAP-IL-6 mouse devoid of fluorescent cells (left). A single GFP-positive astrocyte-like, stellate cell with branched processes is present in the white matter of the cerebellum of a 30-week-old VEGF-GFP mouse (middle), whereas numerous GFP-positive cells are found in the cerebellum of GFAP-IL-6/VEGF-GFP mice (right). Note the stellate shape of some GFP-positive cells and the radial glia-like shape of other stained cells, both compatible with astrocytic differentiation. (b) Detection of cerebral VEGF-GFP fluorescence in serial coronal sections of a 30-week-old GFAP-IL-6/VEGF-GFP mouse. The number of green fluorescent cells is highest in the cerebellum (panel 6). The prominent unspecific yellow labeling in several panels, especially in panel 7, is due to autofluorescence of erythrocytes and neurons. (c) VEGF-GFP reporter activity observed at different ages (10, 30 and 60 weeks) in the cerebellum of GFAP-IL-6/VEGF-GFP mice. Astrocyte-like cells with stellate and radial processes are especially frequent at 30 and 60 weeks. (d) GFP fluorescence from a cerebellar coronal section of a 40-week-old GFAP-IL-6/VEGF-GFP mouse (upper left), GFAP immunofluorescence (upper middle). GFP-positive cells with stellate and radial processes colocalized with GFAP-immunoreactive cells (upper right), confirming the astrocytic phenotype of these cells. GFP-labeling does not colocalize with the neuronal marker NeuN (lower panels).
Material and methods Cell culture NIH3T3, U87MG and Cos 7 cells were grown in Dulbecco’s modified Eagle’s medium (Sigma-Aldrich, Buchs, Switzerland) supplemented with 4 mM Glutamax (Invitrogen-Gibco, Basel, Switzerland), 100 units/ml penicillin/streptomycin (InvitrogenGibco) and 10% fetal calf serum (Invitrogen-Gibco). Animals To assess transcriptional regulation of VEGF in vivo we made use of transgenic mouse models. Glial fibrillary acidic protein (GFAP)-IL-6 transgenic mice overexpressing IL-6 selectively in astrocytes14 were a generous gift from Iain L. Campbell, The Scripps Research Institute, La Jolla, CA. VEGF-green fluorescent protein (GFP) mice carrying the reporter gene GFP under the control of the VEGF promoter28 were contributed generously by Brian Seed, Massachusetts General Hospital, Boston, MA. Heterozygous GFAP-IL-6 and VEGF-GFP mice as well as double transgenic offspring GFAP-IL-6/VEGF-GFP were genotyped by PCR. Details of the genotyping procedure including primer sequences are available upon request. Immunofluorescence To study the pattern of VEGF-GFP expression, transgenic mice were anesthetized with ketamine/xylazine (12.5/2 mg/ml) and
perfused through the left ventricle with 10 ml 0.1 M phosphatebuffered saline (PBS; pH 7.4) containing 0.1% wt/vol NaNO3, 1% wt/vol MgCl2 6 H2O, 10 U/ml liquemin, followed by 40 ml 4% paraformaldehyde in PBS. Brains were removed and postfixed in the same fixative for 6–18 hr, then immersed with 18% sucrose in PBS with 0.1% NaN3 for cryoprotection. Coronal sections (5 mM) were cut with a cryostat and viewed under a fluorescence microscope (Leica Microsystems, Glattbrugg, Switzerland). For GFAP and NeuN immunofluorescence, sections were blocked in Trisbuffered saline (TBS) with 1% BSA and 5% normal goat serum for 20 min and then incubated with rabbit anti-GFAP polyclonal antibody (Dako, Glostrup, Denmark) or mouse anti-NeuN monoclonal antibody (Chemicon International, Temecula, CA) diluted 1:100 in TBS with 1% BSA and 5% normal goat serum for 60 min at room temperature. Sections were then incubated with CY3-conjugated goat-anti-mouse or rabbit secondary antibodies (Jackson Immunoresearch Laboratories, West Grove, PA) diluted 1:400 in TBS with 1% BSA and 5% normal goat serum for 60 min and thereafter mounted with the Dako fluorescent mounting medium (Dako) and visualized under a fluorescence microscope. ELISA for VEGF in cell cultures For the generation of conditioned medium, 5 104 U87MG cells were seeded in 24-well plates in growth medium overnight. After washing 2 times with PBS, cells were starved in serum-free medium for 24 hr. Supernatant was collected at different time
204
LOEFFLER ET AL.
FIGURE 1 – CONTINUED.
MECHANISM OF IL-6 INDUCED VEGF EXPRESSION
205
FIGURE 1 – CONTINUED.
points after incubation with IL-6 (100 ng/ml; R&D Systems, Abingdon, UK), centrifuged cell-free and stored at 208C. Cell extracts were lysed and the protein concentration was determined with a BCA kit (Pierce, Rockford, IL). VEGF concentrations were assessed by CYTELISA (CytImmune Sciences, College Park, MD) following the instructions of the manufacturer and normalized to protein content. Plasmids The VEGF promoter-luciferase construct hVEGF (2018/þ50)Luc as well as the 50 -deletion constructs were generous gifts from Ulrike Fiedler and Dieter Marme´, Tumour Biology Center, University of Freiburg, Germany.29 Briefly, the nucleotide sequence 2018 to þ50 bp of the human VEGF promoter had been cloned
into the promoterless firefly luciferase reporter vector pAH 1409 to generate the plasmid hVEGF (2018/þ50); 50 -deletions were made using restriction sites within the VEGF promoter and the pAH 1409 multiple cloning site. The mammalian expression vectors harboring a dominant-negative STAT3 (STAT3F) or a constitutive active STAT3 (STAT3C) were obtained by courtesy from Joo-Yeon Yoo, The Johns Hopkins University School of Medicine, Baltimore, MD.30 The human Sp1 expression vector was kindly provided by Giuliana Napolitano and Luigi Lania, University of Naples ‘‘Federico II’’, Naples, Italy.31 Transfections and luciferase assays Transient transfection assays, using Transfast reagent (Promega, Madison, WI) for NIH3T3 cells and Neuroporter (Gene Therapy
206
LOEFFLER ET AL.
FIGURE 1 – CONTINUED.
Systems, San Diego, CA) for U87MG cells, were performed at 60% cell confluence with 1 mg of reporter plasmid plus 100 ng of Renilla luciferase plasmid (Promega) as an internal control for transfection efficiency. Cells were serum-starved for 24 hr before stimulation with IL-6 (10 ng/ml for NIH3T3 cells; 100 ng/ml for U87MG cells) for 24 hr. For inhibition experiments, cells were cotransfected with 200 ng of STAT3F or parent vector pYN3218. Cell extracts were prepared using the Dual Luciferase Assay System (Promega). Luciferase activity in cell extracts was measured using a Dynatech luminometer (Dynatech Laboratories, Chantilly, VA). Firefly luciferase activity was normalized to Renilla luciferase activity. Western blots Cells were harvested and washed with ice-cold PBS. The cells were lysed in RIPA buffer containing 50 mM Tris pH 8.0, 150 mM NaCl, 0.1% SDS, 1% NP-40, 0.1% sodium deoxycholate, protease inhibitor cocktail (Roche, Basel, Switzerland) and 1 mM Na3VO4, 20 mM NaF, 1 mM Na4P2O7. A total of 40 mg of total protein, quantified with a BCA kit (Pierce), were separated by SDS-PAGE and transferred to Protran (0.45 mm) nitrocellulose membranes (Schleicher & Schuell, Dassel, Germany). The membranes were blocked with 5% milk in PBS and subsequently incubated with the phospho-STAT3 (Tyr705) antibody (Cell Signalling, Beverly, MA) diluted at 1:1,000 for 3 hr at room temperature, or with STAT3 (C-20) antibody (Santa Cruz Biotechnology, Santa Cruz, CA) at a dilution of 1:1,000 for 1 hr at room temperature. After incubation with the secondary antibodies conjugated to horseradish peroxidase (1:2,000; Dako) for 1 hr, the bands were visualized by ECL (Amersham Biosciences Europe, Du¨bendorf, Switzerland). Nuclear extract preparation Cells were washed and scraped in ice-cold PBS and transferred to centrifuge tubes. Samples were spun at 4,000g for 3 min at 48C. The pellet was resuspended in buffer containing 10 mM HEPES, pH 7.9, 1.5 mM MgCl2, 10 mM KCl to swell on ice for 10 min. The samples were vortexed for 10 sec and spun at 4,000g for
3 min. The resulting pellet was resuspended in buffer containing 20 mM HEPES, pH 7.9, 420 mM NaCl, 1.5 mM MgCl2, 0.2 mM EDTA, 25% glycerol and incubated on ice for 20 min. These samples were spun at 16,800g for 2 min and the supernatant was transferred to Eppendorf tubes and stored at 808C until use. All buffers contained protease inhibitor cocktail (Roche) and 1 mM Na3VO4, 20 mM NaF, 1 mM Na4P2O7. Electrophoretic mobility shift assay Oligonucleotides were radiolabeled using the T4 polynucleotide kinase (New England Biolabs, Beverly, MA) and [g-32P] ATP (Amersham). A total of 5 mg protein was incubated with 35 fmol of [g-32P] ATP-labeled probe in binding buffer (10 mM HEPES pH 7.9, 1 mM EDTA, 5 mM MgCl2, 10% glycerol, 5 mM DTT, protease inhibitor cocktail and 1 mg of poly(dI-dC)) for 10 min on ice. Supershift assays were performed by addition of 2 mg of each antibody (TransCruz Gel Supershift reagents; Santa Cruz Biotechnology) to the binding reaction incubating for 15 min on ice before adding [g-32P] ATP-labeled probe. The protein-DNA complexes were separated on 4.5% polyacrylamide gel containing 7.5% glycerol in 0.25 TBE at 15 V/cm (160 V) for 4 hr at 48C. The gel was dried without fixation and exposed to the phosphor imager. Coimmunoprecipitation Cos7 cells were transfected with pYN3218-STAT3 and pEVR2Sp1 using Lipofectamine 2000 according to the manufacturer’s recommendations. At 48 hr after transfection, the cells were lysed in RIPA buffer and cleared by centrifugation at 15,700g for 30 min at 48C. Cell lysates containing equal amount of total protein were precleared for 1 hr with 40 ml of 50% (wt/vol) protein A-Sepharose beads and incubated overnight with Sp1 (PEP 2) antibodies (Santa Cruz Biotechnology, Santa Cruz, CA) or control (immunoglobulin G [IgG]) antibodies. The beads were added for 2 hr and then washed extensively with lysis buffer; bound proteins were fractionated on 8% SDS-PAGE, transferred to a nitrocellulose membrane, and incubated with STAT3 or Sp1 antibodies. The bands were visualized by ECL.
MECHANISM OF IL-6 INDUCED VEGF EXPRESSION
207
Statistical analysis Statistical differences were evaluated by 2-tailed Student’s t-tests for paired observations. Differences were considered statistically significant at p < 0.05. All data are expressed as mean 6 standard error (SE).
Results IL-6 induces VEGF transcription in the mouse brain To study the transcriptional regulation of VEGF by IL-6 in vivo, we used GFAP-IL-6 transgenic mice overexpressing IL-6 in astrocytes14 and crossbred them with VEGF-GFP reporter mice carrying the reporter gene GFP under the control of the human VEGF promoter.28 In VEGF promoter activation, GFP is expressed and can be visualized in coronal sections using a fluorescence microscope. In single transgenic GFAP-IL-6 mice and nontransgenic littermates of VEGF-GFP mice (VEGF-GFP/) that served as controls, we detected only yellow-brownish autofluorescence, mainly by erythrocytes and Purkinje cells, but no green fluorescence (Fig. 1a). In VEGF-GFP transgenic mice, only single fluorescent cells were observed, whereas in double-transgenic GFAPIL-6/VEGF-GFP mice, a strong GFP expression and concomitant green fluorescence by numerous cells was found (Fig. 1a). We then studied the spatial VEGF-GFP expression in the brains of 30-week-old double-transgenic GFAP-IL-6/VEGF-GFP mice by performing contiguous coronal sections from the olfactory bulbs to the medulla oblongata (Fig. 1b). We found the highest VEGFGFP expression in the cerebellum (Fig. 1b; panel 6) reflecting the pattern of astrogliosis found in the GFAP-IL-6 mouse.14 Analysis of the temporal VEGF-GFP distribution revealed that GFP expression increased from the age of 10-30 weeks (Fig. 1c). Thereafter, from 30 weeks until the end of the observation period of 60 weeks, GFP expression remained at the same level. The green fluorescent cells had long, branched, cytoplasmic processes, resembling astrocytes (Fig. 1c). To confirm their astrocytic phenotype and to exclude that these cells were neurons, we performed double-immunofluorescence of cryostat sections with GFAP or NeuN antibodies. GFP-labeling was found to colocalize with GFAP-immunofluorescence (Fig. 1d), demonstrating that these cells were indeed astrocytes. In contrast, the GFP-labeling was not found in neurons. The fluorescent astrocytes developed distinct signs of reactive transformation including cytoplasmic swelling and elongated, coarse cell processes (Fig. 1c and d) between the age of 10-30 weeks due to the chronic IL-6 overexpression from the GFAP-IL-6 transgene.14 IL-6 increases VEGF secretion and transactivates the VEGF promoter in human GBM cells To scrutinize the effects of IL-6 on VEGF expression in human gliomas, human U87MG GBM cells were exposed to IL-6 and analyzed for secreted VEGF165 protein. Unstimulated U87MG cells exhibit a high endogenous VEGF expression level32,33 and IL-6 promotes growth of this cell line by an autocrine mechanism.9 Therefore, higher IL-6 concentrations (100 ng/ml) were required for an induction of VEGF expression compared to other IL-6-sensitive model cell lines such as NIH3T3 or HepG2 (10 ng/ml) (data not shown). IL-6 significantly increased the VEGF secretion beyond 24 hr (p < 0.025; Fig. 2a) but this augmentation lessened at 48, 72 and 96 hr by the accumulation of basal VEGF secretion (p > 0.05). To analyze the transactivating effects of IL-6 on VEGF expression, functional transfection studies using the plasmid hVEGF (2018/þ50)-Luc carrying the human VEGF promoter sequence from 2018 to þ50 upstream of the luciferase reporter gene were performed in U87MG cells. In this assay, IL-6 treatment (100 ng/ml) resulted in a significant increase in VEGF promoter activity (p < 0.02; Fig. 2b), which correlated with the secreted VEGF protein levels found in the ELISA (Fig. 2a).
FIGURE 2 – IL-6 increases VEGF protein release and transactivates the VEGF promoter in the human glioma cell line U87MG. (a) U87MG cells exposed to IL-6 (100 ng/ml) for various periods of time were assayed by ELISA for VEGF protein release into the cell culture supernatant and normalized to milligrams of total protein. Results are expressed as mean 6 SE and significantly increased protein levels are indicated (*p < 0.025). (b) U87MG cells were transiently transfected with 0.5 mg of hVEGF (-2018/þ50)-Luc, 50 ng of Renilla luciferase construct, then treated with IL-6 (100 ng/ml) or left untreated for 36 hr and analyzed for luciferase activity. The bar graph shown is representative of three independent experiments, each performed in triplicate (*p < 0.02).
Dominant-negative STAT3 abolishes IL-6-induced VEGF transcription IL-6 is known to signal mainly through STAT3 and MAPK.34,35 In our time course studies with IL-6-stimulated cells, phosphorylation of STAT3 at tyrosine 705 occurred after 5 min in both cell lines, NIH3T3 and U87MG (Fig. 3a). To evaluate the relevance of the STAT3 and MAPK signaling pathways in the transcriptional upregulation of VEGF by IL-6, cells were either cotransfected with a dominant-negative STAT3 mutant (STAT3F)30 or pretreated with the MEK1 inhibitor PD98059. IL-6-induced VEGF transcription was completely abolished by cotransfection of STAT3F in NIH3T3 cells as well as in U87MG cells (Fig. 3b and c). Pretreatment of NIH3T3 cells with PD98059 caused only a moderate decrease in IL-6-mediated VEGF transcription (data not shown) indicating a STAT3-mediated signal transduction in IL-6induced VEGF expression. The minimal IL-6 responsive element is located between nucleotides 85 and 50 of the human VEGF promoter We next employed 50 -progressive deletion mutants of the VEGF promoter29 to locate the critical IL-6-responsive element in the
208
LOEFFLER ET AL.
FIGURE 4 – The IL-6 response element is located between nucleotides 85 and 50 of the human VEGF promoter. (a,b) Progressive 50 -deletions were introduced between positions 2018 and 50 bp. NIH3T3 and U87MG cells were transiently transfected with 1 mg of each construct and 100 ng of Renilla luciferase plasmid. After transfection, cells were serum-starved for 24 hr and then stimulated with IL-6 (10 ng/ml for NIH3T3; 100 ng/ml for U87MG) for 24 hr or left untreated. Luciferase activity was normalized for transfection efficiency. Data represent the means of 3 experiments, each performed in triplicate. Significantly increased activities are indicated (*p < 0.05).
FIGURE 3 – IL-6-induced VEGF promoter activity is STAT3dependent. (a) Time course experiment showing IL-6-induced phosphorylation of STAT3 in NIH3T3 cells and in U87MG cells. Cells were serum-starved for 24 hr and then treated for the indicated times with IL-6 (10 ng/ml). Phosphorylation of STAT3 was detected by Western blot analysis using anti-phospho-STAT3 antibodies. Membranes were stripped and reprobed with anti-STAT3 antibodies. Data shown represent a typical result obtained from 3 independent experiments. (b,c) Functional transactivation assays: NIH3T3 and U87MG cells were transiently transfected with 1 mg of hVEGF (2018/þ50)Luc, 100 ng of Renilla luciferase plasmid and with 200 ng of parental vector or expression vector carrying dominant-negative STAT3 (STAT3F). After transfection, cells were serum-starved for 24 hr and then stimulated with IL-6 (10 ng/ml for NIH3T3, 100 ng/ml for U87MG cells) for 24 hr or left untreated. Luciferase activity was normalized for transfection efficiency. Results are expressed as mean 6 SE and significantly increased activities are indicated (*p < 0.0005 for NIH3T3; p < 0.02 for U87MG).
human VEGF promoter. Use of these deletion mutants in functional promoter assays revealed that loss of the region spanning nucleotide (nt) 2018 to 85 had no substantial influence on VEGF promoter activity (Fig. 4a). In contrast, removal of nt 85 to 50 from the proximal VEGF promoter repressed basal promoter activity as well as IL-6 responsiveness by about 100-fold, indicating that the region between 85 to 50 contains the enhancer-like elements required for full basal and IL-6-inducible expression of the VEGF gene in NIH3T3 cells (Fig. 4a). This experiment was repeated in the human glioblastoma cell line
U87MG with similar results (Fig. 4b). By electrophoretic mobility shift assay (EMSA), 3 complexes, 2 major and 1 minor, bound to this element both under basal conditions and after IL-6-stimulation (Fig. 5a). IL-6 treatment of U87MG cells resulted in a slight increase of the 3 DNA-protein complexes over time, with complex 1 showing the strongest augmentation. Maximal binding of the 3 complexes occurred after 20 min (Fig. 5a; lane 4), whereas prolonged exposure to IL-6 decreased complex formation partially (Fig. 5a; lane 5). IL-6 induces binding of Sp1 and Sp3 to the 85/50 element of the human VEGF promoter Computer-assisted sequence analysis36 of the 85/50 fragment of the human VEGF promoter revealed the presence of 3 Sp1 binding sites (GC-boxes I-III), 2 AP-2 binding sites and 2 early growth response factor-1 (Egr-1) binding sites (Fig. 5b), but no STAT3 binding element (SBE). To further characterize the nuclear factors binding to the VEGF (88/50) element, gel supershift analysis was performed in the presence of Sp1, Sp2, Sp3, Sp4, AP-2 and Egr-1 antibodies in IL-6-stimulated U87MG cells (Fig. 5c). Sp2, Sp4, AP-2 and Egr-1 antibodies had no effect on the binding activities (Fig. 5c; lanes 3, 5, 7–8), whereas addition of Sp1 antibody resulted in a supershift of the complex 1 (Fig. 5c; lane 2). Complexes 2 and 3 were supershifted by Sp3 (Fig. 5c; lane 4). Complex 3 corresponds to the shorter isoform (60–70 kDa) of Sp3.37 EMSA competition experiments confirmed these results (data not shown). Single mutations of the 3 Sp1 binding sites within the 88/50 VEGF element were then introduced and used in luciferase assays. Mutation of the Sp1 binding sites
MECHANISM OF IL-6 INDUCED VEGF EXPRESSION
209
FIGURE 5 – Sp1, Sp2 and Sp3 transcription factors bind to the 88/50 element of the human VEGF promoter. (a) Nuclear extracts from untreated or IL-6stimulated (100 ng/ml) U87MG cells were prepared at the time points indicated and subjected to EMSA using the 88/50 VEGF fragment as 32P-labeled probe. The autoradiograph shown is representative for 3 independent experiments. (b) Illustration of potential binding sites for the transcription factors Sp1, Egr-1 and activator protein 2 (AP2) within the 88/50 VEGF promoter sequence. (c) EMSA supershift experiment: Nuclear extracts of untreated or stimulated U87MG cells (IL-6 100 ng/ml for 15 min) were incubated with the radiolabeled probe hVEGF 88/50 in the absence or presence of specific antibodies recognizing Sp1, Sp2, Sp3, Sp4, AP2 and Egr-1. The autoradiographs shown are representative of 3 independent experiments.
located at 73/66 and 58/52 of the VEGF promoter resulted in nearly complete loss of basal promoter activity and IL-6 responsiveness in both cell lines, U87MG and NIH3T3 (data not shown). In addition, only overexpressed Sp1 could stimulate both basal and IL-6-induced VEGF transcription in a dose-dependent manner, but not Sp3 (data not shown). In summary, these experiments validate the absolute requirement for Sp1 in the regulation of VEGF transcription by IL-6.
STAT3 drives the minimal VEGF promoter through direct interaction with Sp1 We then attempted to investigate the molecular mechanism responsible for IL-6-induced VEGF transcription mediated by Sp1 in the context of an SBE-less minimal VEGF promoter. To evaluate the functional importance of STAT3 in this setting, transactivation experiments with the hVEGF(88/50) reporter gene
210
LOEFFLER ET AL.
FIGURE 6 – STAT3 drives the minimal VEGF promoter through direct interaction with Sp1. Functional transactivation assays: (a) NIH3T3 cells were transiently transfected with 1 mg of hVEGF (85/þ50)-Luc, 100 ng of Renilla luciferase plasmid and with 200 ng of parental vector or dominant-negative STAT3 (STAT3F). After transfection, cells were serum-starved for 24 hr and then stimulated with IL-6 (10 ng/ml for NIH3T3) for 24 hr or left untreated. Luciferase activity was normalized for transfection efficiency. (b) NIH3T3 cells were transiently transfected with 500 ng of hVEGF (85/þ50)-Luc, 50 ng of Renilla luciferase plasmid and with 600 ng of parental vector or indicated vector dose of constitutive active STAT3 (STAT3C). After transfection, cells were serum-starved for 48 hr and then lysed unstimulated. Luciferase activity was normalized for transfection efficiency. Data represent the mean of 3 experiments, each performed in triplicate. Significantly increased activities are indicated (*p < 0.05). (c) EMSA supershift experiment: Nuclear extracts of IL-6-stimulated U87MG cells (100 ng/ml for 15 min) were incubated with the radiolabeled probe hVEGF 88/50 in the presence of either a control antibody or a-STAT3. Addition of a-STAT3 interferes with Sp1/Sp3 protein-DNA complex formation. The autoradiographs shown are representative of 3 independent experiments. (d) Coimmunoprecipitation experiment: Cos7 cells were transiently transfected with pYN3218-STAT3 and pEVR2-Sp1. After 48 hr, equal protein amounts from whole-cell lysates were subjected to immunoprecipitation using an antibody against Sp1 or an anti-rabbit IgG. The immunoprecipitated proteins were separated and subjected to immunoblot analysis. No complexes were detected in lysates immunoprecipitated with normal rabbit immunoglobulin G. The blots shown are representative of at least 3 independent experiments.
construct were performed in NIH3T3 cells. Cotransfection of dominant-negative STAT3F completely abolished IL-6-induced VEGF transcription from the minimal VEGF promoter (Fig. 6a),
whereas cotransfection of constitutive active STAT3C induced VEGF promoter activity in a dose-dependent fashion, even in the absence of IL-6 stimulation (Fig. 6b). More intriguingly, the addi-
MECHANISM OF IL-6 INDUCED VEGF EXPRESSION
211
tion of anti-STAT3 antibody to the EMSA reaction disrupted the binding of Sp1 and Sp3 to the 88/50 VEGF element (Fig. 6c). An equivalent amount of control antibody had no effect. These data show that STAT3 is directly involved in IL-6-driven proteinDNA complex formation of Sp1/Sp3 and the 88/50 VEGF promoter element. To further investigate a potential physical interaction between STAT3 and Sp1, we overexpressed STAT3 and Sp1 in Cos7 cells. Whole-cell lysates were immunoprecipitated with Sp1 antiserum, separated by SDS-PAGE and immunoblotted for STAT3. Figure 6d shows that STAT3 associates with Sp1, and thereby conveys an IL-6-induced STAT3 signal through an SBEless promoter element. Discussion Under normal conditions, the expression levels of IL-6 and VEGF are low in the adult mammalian central nervous system.38,39 In human gliomas, however, both factors are expressed at increased levels that correlate with the grade of malignancy.4,10,38 We investigated the underlying mechanism of the functional relationship between IL-6 and VEGF in glioma angiogenesis. A proangiogenic role for IL-6 in the central nervous system is supported by recent data which demonstrated that IL-6 induces proliferation and capillary tube formation of murine brain microvessel endothelial cells in vitro.40 Moreover, IL-6 promoted the progression of cervical cancer by activating angiogenesis, and modulated the activity of VEGF through STAT3.27 Similarly, vIL-6, a viral version of IL-6 produced by Kaposi’s sarcoma-associated herpes virus (KSHV), induced VEGF production in NIH3T3 cells, and vIL-6 producing NIH3T3 cells gave rise to highly vascularized tumors in nude mice.24 Using genetically engineered mice, we demonstrate here that forced IL-6 expression in the mouse brain induces astrogliosis and disposes astrocytes to synthesize GFP from a human VEGF promoter, indicating that IL-6 is directly involved in the regulation of VEGF expression in vivo. Furthermore, by doubleimmunofluorescence labeling for GFAP and NeuN, we proved that the GFP expressing cells are indeed astrocytes. Thus, it seems that the extensive angiogenesis described previously in brains of GFAP-IL-6 mice14 is at least in part attributable to astrocyte-borne VEGF. These data are consistent with an auto/ paracrine upregulation of VEGF expression by IL-6 in astrocytes, which eventually will ensue in neovascularization. The existence of such IL-6 activation loops in glioblastoma cells has been demonstrated in vitro. Incubation of GBM cells with a neutralizing anti-IL-6 antibody inhibited proliferation and induced apoptosis.9,41 So far, IL-6 was shown to induce VEGF expression in ratderived cell lines and in human myeloma cells in vitro.6,42 Here, we examined IL-6-induced VEGF transcription in human glioblastoma cells (U87MG) and in NIH3T3 model cells.9,24 As expected, IL-6 increased secretion of VEGF protein from U87MG cells. Functional promoter assays confirmed that increased VEGF secretion occurred through augmentation of VEGF transcription. However, this increase was more prominent in NIH3T3 cells than in U87MG cells, because U87MG cells exhibit a higher basal level of VEGF expression,33 which might originate from inactivation of the phosphatase and tensin homolog deleted from chromosome 10 (PTEN),43 overexpression of epidermal growth factor (EGF) receptor44 or autocrine secretion of IL-6.9 Apart from hypoxia, the stimuli for VEGF induction during glial tumor progression are not well defined. In our study we focused on IL-6 as a novel stimulator of glioma angiogenesis and scrutinized the underlying transcriptional mechanism. Functional transactivation assays showed that a region spanning from nt -88 to 50 is indispensable for basal and IL-6-triggered VEGF promoter activity in U87MG cells and in NIH3T3 cells, consistent with previous reports.29,45,46 This GC-rich region contains binding
FIGURE 7 – New mechanism of transcriptional regulation by STAT3. (a) Depiction of the classical transcriptional mechanism regulated by STAT3: IL-6 induces STAT3 activation and DNA-binding through a STAT3-binding element (SBE) in the promoter of a target gene. (b) New proposed mechanism of STAT3-regulated VEGF transcription: Upon IL-6 stimulation, activated STAT3 translocates to the nucleus and tethers to prebound Sp1 on the SBE-less 88/50 promoter element, thereby activating VEGF transcription.
sites for Sp1, AP-2 and Egr-1 transcription factors. Out of the candidate binding factors only Sp1 and Sp3 yielded a supershifted band in EMSAs after addition of the respective antibodies. IL-6 stimulation slightly enhanced the DNA binding of Sp1/Sp3 in GBM cells. In HepG2 cells, this enhanced DNA-protein complex formation by IL-6 was even stronger (data not shown). Sp1 belongs to a family of zinc finger proteins that are ubiquitously expressed.37 Whereas Sp1 has been found to act predominantly in a positive manner, Sp3 can also function as a repressor or in Sp1-mediated gene activation.37 Overexpression experiments revealed that Sp1 strongly activates both basal and IL-6-induced VEGF transcription, and that Sp3 does not inhibit, but rather enhances Sp1 activity in IL-6-induced VEGF transcription. This is in line with a report showing that decoy oligonucleotides resembling Sp1 binding sites suppressed VEGF expression, cell growth and invasiveness of the human GBM cell line U251 in vitro. Simultaneous mutation of the Sp1 binding sites II and III leaving an intact Sp1 binding site I revealed that GC-boxes II and III are indispensable for full basal and IL-6-stimulated VEGF promoter activity, whereas GC-box I is apparently not essential for IL-6regulated VEGF expression. This situation is similar to the one observed in gastric cancer cells stimulated by oxidative stress.46 IL-6 is known to signal primarily via the STAT and MAPK pathways.35 In contrast to HGF-induced VEGF transcription,
212
LOEFFLER ET AL.
where Sp1 phosphorylation is dependent on PI3-K and MEK1/2 signalling,47 pretreatment of NIH3T3 cells with either PD98059 or LY294002 led only to a 30% reduction of IL-6-triggered VEGF transcription (data not shown), suggesting that the MAPK and PI3-K/Akt signaling pathways are only partially involved. However, transfection of NIH3T3 cells and U87MG cells with dominant-negative STAT3 reduced IL-6-mediated VEGF expression to control levels, attesting that STAT3 is absolutely required for IL-6-regulated VEGF expression. These results are in line with those of Wei et al.27 and others showing that constitutive STAT3 activity upregulates VEGF expression and tumor angiogenesis.48,49 A putative upstream STAT binding element at nt 842 to 849 in the VEGF promoter was recently ascribed to be involved in STAT3-mediated VEGF expression.48 However, in our progressive deletion analysis, IL-6 is still increasing VEGF transcription even in the absence of this upstream promoter region, indicating that IL-6-induced VEGF transcription is not dependent on the supposed STAT binding element at nt 842 to 849 in the VEGF promoter. Because both dominant-negative STAT3 and constitutively active STAT3 are capable of regulating IL-6-triggered VEGF transcription through the SBE-less 88/50 VEGF promoter, we hypothesize that IL-6 drives VEGF expression via STAT3 and Sp1 in a novel noncanonical way. This hypothesis was comforted by the fact that STAT3 antibodies blocked the IL6-induced binding of Sp1/Sp3 to the 88/50 VEGF promoter. Our findings are reminiscent of a situation in which STAT3 activates transcription without binding itself to a palindromic enhancer element. In fact, our coimmunoprecipitation experiments imply that STAT3 directly interacts with a preexisting Sp1/DNA complex, thus allowing the transmission of an extracellular IL-6
stimulus via STAT3 signaling in the context of an SBE-less promoter (Fig. 7). There is evidence that transcription factors can activate transcription without DNA binding, e.g., that activin A can trigger transactivation of the VEGF promoter through physical association of Smad and Sp1.50 As for STAT3 and Sp1, it has already been shown that Stat3 and Sp1 cooperatively activate the C/EBPdelta promoter,51 albeit through adjacent Sp1 and STAT3 binding sites. However, the 88/50 VEGF promoter contains only Sp1 binding sites, but no STAT3 binding elements. We describe here a new mechanism of transcriptional regulation by which IL-6-activated STAT3 tethers to Sp1 bound to the SBE-less 88/50 VEGF promoter, thereby inducing VEGF transcription. The characterization of this mechanism may be relevant for the design of novel therapeutic approaches in VEGF-mediated angiogenesis in gliomas. Acknowledgements We thank Dr. I.L. Campbell (The Scripps Research Institute, La Jolla, CA) and Dr. B. Seed (Massachusetts General Hospital, Boston, MA) for their kind donation of GFAP-IL-6 and VEGFGFP transgenic mice. We thank Dr. U. Fiedler and Dr. D. Marme´ (Tumour Biology Center, University of Freiburg, Germany) for the generous gift of VEGF promoter-luciferase constructs. We thank Dr. J.-Y. Yoo (The Johns Hopkins University School of Medicine, Baltimore, MD), Dr. G. Napolitano and Dr. L. Lania (University of Naples ‘‘Federico II’’, Naples, Italy) for their respective gifts of the expression vectors for wild-type STAT3, STAT3F and constitutive active STAT3, as well as for human Sp1. We thank Dr. A. Kappeler for excellent technical support. We thank Dr. F. Bianchi for statistical analysis.
References 1. 2. 3.
4. 5.
6. 7. 8. 9. 10.
11. 12. 13. 14.
Maher EA, Furnari FB, Bachoo RM, Rowitch DH, Louis DN, Cavenee WK, DePinho RA. Malignant glioma: genetics and biology of a grave matter. Genes Dev 2001;15:1311–33. Folkman J. Seminars in Medicine of the Beth Israel Hospital, Boston. Clinical applications of research on angiogenesis. N Engl J Med 1995; 333:1757–63. Cheng SY, Huang HJ, Nagane M, Ji XD, Wang D, Shih CC, Arap W, Huang CM, Cavenee WK. Suppression of glioblastoma angiogenicity and tumorigenicity by inhibition of endogenous expression of vascular endothelial growth factor. Proc Natl Acad Sci USA 1996;93:8502–7. Chaudhry IH, O’Donovan DG, Brenchley PE, Reid H, Roberts IS. Vascular endothelial growth factor expression correlates with tumour grade and vascularity in gliomas. Histopathology 2001;39:409–15. Pietsch T, Valter MM, Wolf HK, von Deimling A, Huang HJ, Cavenee WK, Wiestler OD. Expression and distribution of vascular endothelial growth factor protein in human brain tumors. Acta Neuropathol (Berl) 1997;93:109–17. Cohen T, Nahari D, Cerem LW, Neufeld G, Levi BZ. Interleukin 6 induces the expression of vascular endothelial growth factor. J Biol Chem 1996;271:736–41. Tchirkov A, Rolhion C, Bertrand S, Dore JF, Dubost JJ, Verrelle P. IL-6 gene amplification and expression in human glioblastomas. Br J Cancer 2001;85:518–22. Van Meir E, Sawamura Y, Diserens AC, Hamou MF, de Tribolet N. Human glioblastoma cells release interleukin 6 in vivo and in vitro. Cancer Res 1990;50:6683–8. Goswami S, Gupta A, Sharma SK. Interleukin-6-mediated autocrine growth promotion in human glioblastoma multiforme cell line U87MG. J Neurochem 1998;71:1837–45. Rolhion C, Penault-Llorca F, Kemeny JL, Lemaire JJ, Jullien C, Labit-Bouvier C, Finat-Duclos F, Verrelle P. Interleukin-6 overexpression as a marker of malignancy in human gliomas. J Neurosurg 2001;94:97–101. Weissenberger J, Steinbach JP, Malin G, Spada S, Rulicke T, Aguzzi A. Development and malignant progression of astrocytomas in GFAPv-src transgenic mice. Oncogene 1997;14:2005–13. Weissenberger J, Loeffler S, Kappeler A, Kopf M, Lukes A, Afanasieva TA, Aguzzi A, Weis J. IL-6 is required for glioma development in a mouse model. Oncogene 2004;23:3308–16. Thier M, Marz P, Otten U, Weis J, Rose-John S. Interleukin-6 (IL-6) and its soluble receptor support survival of sensory neurons. J Neurosci Res 1999;55:411–22. Campbell IL, Abraham CR, Masliah E, Kemper P, Inglis JD, Oldstone MB, Mucke L. Neurologic disease induced in transgenic mice by cerebral
15.
16. 17. 18.
19.
20. 21. 22.
23. 24.
25. 26. 27.
overexpression of interleukin 6. Proc Natl Acad Sci USA 1993;90: 10061–5. Brunello AG, Weissenberger J, Kappeler A, Vallan C, Peters M, Rose-John S, Weis J. Astrocytic alterations in interleukin-6/Soluble interleukin-6 receptor alpha double-transgenic mice. Am J Pathol 2000;157:1485–93. Heinrich PC, Behrmann I, Muller-Newen G, Schaper F, Graeve L. Interleukin-6-type cytokine signalling through the gp130/Jak/STAT pathway. Biochem J 1998;334(Pt 2):297–314. Akira S. IL-6-regulated transcription factors. Int J Biochem Cell Biol 1997;29:1401–18. Wei LH, Kuo ML, Chen CA, Chou CH, Cheng WF, Chang MC, Su JL, Hsieh CY. The anti-apoptotic role of interleukin-6 in human cervical cancer is mediated by up-regulation of Mcl-1 through a PI 3-K/Akt pathway. Oncogene 2001;20:5799–809. Ogata A, Chauhan D, Teoh G, Treon SP, Urashima M, Schlossman RL and Anderson KC. IL-6 triggers cell growth via the Ras-dependent mitogen-activated protein kinase cascade. J Immunol 1997;159: 2212–21. Zauberman A, Zipori D, Krupsky M, Ben-Levy R. Stress activated protein kinase p38 is involved in IL-6 induced transcriptional activation of STAT3. Oncogene 1999;18:3886–93. Kishimoto T, Akira S, Narazaki M, Taga T. Interleukin-6 family of cytokines and gp130. Blood 1995;86:1243–54. Adler HL, McCurdy MA, Kattan MW, Timme TL, Scardino PT, Thompson TC. Elevated levels of circulating interleukin-6 and transforming growth factor-beta1 in patients with metastatic prostatic carcinoma. J Urol 1999;161:182–7. Akira S, Kishimoto T. The evidence for interleukin-6 as an autocrine growth factor in malignancy. Semin Cancer Biol 1992;3:17–26. Aoki Y, Jaffe ES, Chang Y, Jones K, Teruya-Feldstein J, Moore PS, Tosato G. Angiogenesis and hematopoiesis induced by Kaposi’s sarcoma-associated herpesvirus-encoded interleukin-6. Blood 1999; 93:4034–43. Aoyagi T, Takishima K, Hayakawa M, Nakamura H. Gene expression of TGF-alpha, EGF and IL-6 in cultured renal tubular cells and renal cell carcinoma. Int J Urol 1996;3:392–6. Kinoshita T, Ito H, Miki C. Serum interleukin-6 level reflects the tumor proliferative activity in patients with colorectal carcinoma. Cancer 1999;85:2526–31. Wei LH, Kuo ML, Chen CA, Chou CH, Lai KB, Lee CN, Hsieh CY. Interleukin-6 promotes cervical tumor growth by VEGFdependent angiogenesis via a STAT3 pathway. Oncogene 2003;22: 1517–27.
MECHANISM OF IL-6 INDUCED VEGF EXPRESSION
28. Fukumura D, Xavier R, Sugiura T, Chen Y, Park EC, Lu N, Selig M, Nielsen G, Taksir T, Jain RK, Seed B. Tumor induction of VEGF promoter activity in stromal cells. Cell 1998;94:715–25. 29. Finkenzeller G, Sparacio A, Technau A, Marme D, Siemeister G. Sp1 recognition sites in the proximal promoter of the human vascular endothelial growth factor gene are essential for platelet-derived growth factor-induced gene expression. Oncogene 1997;15:669–76. 30. Yoo JY, Wang W, Desiderio S, Nathans D. Synergistic activity of STAT3 and c-Jun at a specific array of DNA elements in the alpha 2-macroglobulin promoter. J Biol Chem 2001;276:26421–9. 31. Majello B, De Luca P, Suske G, Lania L. Differential transcriptional regulation of c-myc promoter through the same DNA binding sites targeted by Sp1-like proteins. Oncogene 1995;10:1841–8. 32. Valter MM, Wiestler OD, Pietsche T, Pietsch T. Differential control of VEGF synthesis and secretion in human glioma cells by IL-1 and EGF. Int J Dev Neurosci 1999;17:565–77. 33. Feldkamp MM, Lau N, Rak J, Kerbel RS, Guha A. Normoxic and hypoxic regulation of vascular endothelial growth factor (VEGF) by astrocytoma cells is mediated by Ras. Int J Cancer 1999;81:118–24. 34. Hirano T, Nakajima K, Hibi M. Signaling mechanisms through gp130: a model of the cytokine system. Cytokine Growth Factor Rev 1997;8:241–52. 35. Kamimura D, Ishihara K, Hirano T. IL-6 signal transduction and its physiological roles: the signal orchestration model. Rev Physiol Biochem Pharmacol 2003;149:1–38. 36. Quandt K, Frech K, Karas H, Wingender E, Werner T. MatInd and MatInspector: new fast and versatile tools for detection of consensus matches in nucleotide sequence data. Nucleic Acids Res 1995;23: 4878–84. 37. Suske G. The Sp-family of transcription factors. Gene 1999;238:291– 300. 38. Plate KH, Risau W. Angiogenesis in malignant gliomas. Glia 1995; 15:339–47. 39. Van Wagoner NJ, Benveniste EN. Interleukin-6 expression and regulation in astrocytes. J Neuroimmunol 1999;100:124–39. 40. Fee D, Grzybicki D, Dobbs M, Ihyer S, Clotfelter J, Macvilay S, et al. Interleukin 6 promotes vasculogenesis of murine brain microvessel endothelial cells. Cytokine 2000;12:655–65. 41. Rahaman SO, Harbor PC, Chernova O, Barnett GH, Vogelbaum MA, Haque SJ. Inhibition of constitutively active Stat3 suppresses proliferation and induces apoptosis in glioblastoma multiforme cells. Oncogene 2002;21:8404–13.
213
42. Dankbar B, Padro T, Leo R, Feldmann B, Kropff M, Mesters RM, Serve H, Berdel WE, Kienast J. Vascular endothelial growth factor and interleukin-6 in paracrine tumor-stromal cell interactions in multiple myeloma. Blood 2000;95:2630–6. 43. Pore N, Liu S, Haas-Kogan DA, O’Rourke DM, Maity A. PTEN mutation and epidermal growth factor receptor activation regulate vascular endothelial growth factor (VEGF) mRNA expression in human glioblastoma cells by transactivating the proximal VEGF promoter. Cancer Res 2003;63:236–41. 44. Maity A, Pore N, Lee J, Solomon D, O’Rourke DM. Epidermal growth factor receptor transcriptionally up-regulates vascular endothelial growth factor expression in human glioblastoma cells via a pathway involving phosphatidylinositol 30 -kinase and distinct from that induced by hypoxia. Cancer Res 2000;60:5879–86. 45. Benckert C, Jonas S, Cramer T, Von Marschall Z, Schafer G, Peters M, Wagner K, Radke C, Wiedenmann B, Neuhaus P, Hocker M, Rosewicz S. Transforming growth factor beta 1 stimulates vascular endothelial growth factor gene transcription in human cholangiocellular carcinoma cells. Cancer Res 2003;63:1083–92. 46. Schafer G, Cramer T, Suske G, Kemmner W, Wiedenmann B, Hocker M. Oxidative stress regulates vascular endothelial growth factor-A gene transcription through Sp1- and Sp3-dependent activation of two proximal GC-rich promoter elements. J Biol Chem 2003;278:8190–8. 47. Reisinger K, Kaufmann R, Gille J. Increased Sp1 phosphorylation as a mechanism of hepatocyte growth factor (HGF/SF)-induced vascular endothelial growth factor (VEGF/VPF) transcription. J Cell Sci 2003;116(Pt 2):225–38. 48. Niu G, Wright KL, Huang M, Song L, Haura E, Turkson J, Zhang S, Wang T, Sinibaldi D, Coppola D, Heller R, Ellis LM, et al. Constitutive Stat3 activity up-regulates VEGF expression and tumor angiogenesis. Oncogene 2002;21:2000–8. 49. Wei D, Le X, Zheng L, Wang L, Frey JA, Gao AC, Peng Z, Huang S, Xiong HQ, Abbruzzese JL, Xie K. Stat3 activation regulates the expression of vascular endothelial growth factor and human pancreatic cancer angiogenesis and metastasis. Oncogene 2003;22:319–29. 50. Wagner K, Peters M, Scholz A, Benckert C, Ruderisch HS, Wiedenmann B, Rosewicz S. Activin A stimulates vascular endothelial growth factor gene transcription in human hepatocellular carcinoma cells. Gastroenterology 2004;126:1828–43. 51. Cantwell CA, Sterneck E, Johnson PF. Interleukin-6-specific activation of the C/EBPdelta gene in hepatocytes is mediated by Stat3 and Sp1. Mol Cell Biol 1998;18:2108–17.
