| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Tumor Biology |
Departments of General Visceral and Transplantation Surgery [C. B., S. J., P. N.], Hepatology, Gastroenterology, Endocrinology and Metabolism [T. C., Z. v. M., G. S., M. P., K. W., B. W., M. H., S. R.], and Pathology [C. R.], Charité, Humboldt-University, 13353 Berlin, Germany
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In recent years, considerable experimental evidence has accumulated, emphasizing the role of angiogenesis for tumor growth and metastasis (4, 5, 6) . Cancer cells, often during the premalignant stages of tumor development, can activate quiescent endothelial cells through a process termed "angiogenic switch," which is characterized by overexpression of proangiogenic factors paralleled by decreased expression of antiangiogenic factors (7 , 8) .
Among these proangiogenic factors, VEGF plays a predominant role for the induction of tumor-associated angiogenesis (9, 10, 11) . VEGF was originally discovered as a mediator of both vascular permeability and endothelial cell proliferation (12 , 13) and is widely expressed during development and in various pathological states, most notably tumorigenesis. Generally, VEGF is detected in most malignant epithelial tumor cells: VEGF165 is the predominant isoform; and VEGF206 is rarely expressed (10 , 11 , 14) . The biological effects of VEGF are mediated mainly by the tyrosine kinase receptors VEGFR-1 (fms-like tyrosine kinase) and VEGFR-2 [KDR/fetal liver kinase 1 (15 , 16) ]. Both receptors are predominantly expressed in vascular endothelial cells of tumor vessels (10 , 11 , 17) .
In addition to VEGF, TGF-ß (particularly TGF-ß1) has featured prominently among cytokines studied for their capability to regulate angiogenesis, both in vitro and in vivo (18) . TGF-ß1 is a secreted multifunctional protein that modulates a diverse set of cellular responses such as cell proliferation, differentiation, angiogenesis, and metastasis (18, 19, 20) . TGF-ß1 signaling is initiated by assembling two serine/threonine receptor complexes, known as type I and type II receptors, that activate Smad complexes (19) . After TGF-ß binding to its receptors, the receptor-regulated Smads (R-Smads), Smad2 and Smad3, are phosphorylated by the type I receptor and associate with the common partner Smad4. The resulting heteromultimer translocates to the nucleus, where it regulates expression of TGF-ß target genes (21) .
Little is currently known about the expression of the VEGF/VEGFR and TGF-ß/TGF-ß receptor systems in human CCC. The current study was therefore initiated to evaluate the expression of both cytokines and their receptors in human CCC and to explore the potential interaction between these factors at a molecular level.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Tissue Samples.
CCC tissue samples were obtained from 19 individuals who underwent surgical resection at the Department of Surgery at Charité University Hospital. Fifteen samples contained normal liver tissue at resection margins and served as control tissues.
Immunohistochemistry.
Immunohistochemical analysis of surgically resected formalin-fixed, paraffin-embedded tissue samples was carried out using the alkaline phosphatase/anti-alkaline phosphatase method exactly as described previously (22)
. To confirm the specificity of the observed immunohistochemical signal, two independent approaches were used: (a) serial dilution of the primary Ab until the signal disappeared; and (b) preimmune rabbit IgG as the first Ab that failed to reveal relevant staining. Sections were analyzed by two independent investigators. MVD was determined by light microscopy according to the procedure of Weidner et al. (23)
.
In Situ Hybridization.
Sense and antisense 35S-labeled cRNA probes were prepared from human cDNAs subcloned into pBluescript KS vectors: a 517-bp fragment of VEGF121 cDNA was subcloned into a SmaI site; a 1080-bp fragment of fms-like tyrosine kinase cDNA was subcloned into a Sall and a NotI site; and a 1400-bp fragment of KDR was subcloned into a BamIII site. Prehybridization, hybridization, washing procedures, and RNase digestion of mismatched sequences as well as autoradiography were performed as described previously (22)
.
Cell Cultures.
TFK-1 and EGI-1 cells were cultured in DMEM or RPMI 1640, respectively, supplemented with 10% (v/v) FCS, 100 units/ml penicillin, and 100 µg/ml streptomycin. All cells were grown as a subconfluent monolayer in a humidified atmosphere containing 5% CO2 at 37°C.
ELISA for VEGF and TGF-ß1 in Cell Cultures.
For generation of conditioned media, 2 x 104 cells were plated in 24-well plates in growth medium overnight. After washing with PBS, cells were switched to serum-free UltraCulture medium. Conditioned medium was collected after incubation with or without TGF-ß1 (10 ng/ml), centrifuged to remove floating cells, and stored frozen at -20°C. Cell extracts were centrifuged at 15,000 rpm for 15 min, and the protein concentration in each supernatant was determined by colorimetric Bradford protein assay. VEGF and TGF-ß1 concentrations were assessed by ELISA (Quantikine; R&D Systems), following the instructions supplied by the manufacturer and normalized to protein content.
PCR for VEGF Splice Forms.
Total RNA from TFK-1 and EGI-1 cells was isolated using the RNAzol B reagent following the instructions supplied by the manufacturer. Amplification of reverse transcription products was performed as described previously (22
, 24)
. Used primer sequences recognize all known VEGF splice variants and were generated according to published sequences.
Quantification of VEGF by QC RT-PCR.
Total RNA extraction from tumor cells and reverse transcription of 2 µg of RNA were performed as described above. Competitors for VEGF165 and ß-actin were constructed using a "semi-nested" PCR technique as described previously (22)
. Wild-type PCR products were reamplified using combined (so-called "mimic") sense primers, which bind 80 to 120 bp downstream of the original sense primer but contain the original sense primer sequence at their 5'-ends. Only competitor constructs showing equal wild-type:competitor ratios before and after reamplification were used further for competitive PCR. The following amplification and quantification procedures were performed as described previously (22)
.
DNA Constructs and Reporter Plasmids.
The series of progressive hVEGF 5' deletion constructs, which are all based on the promoterless luciferase reporter gene vector pAH1409, have been reported previously (25)
. To study the characteristics of potential hVEGF-regulatory elements in a heterologous promoter system, an oligonucleotide comprising the region hVEGF -88 to -50 bp was synthesized and subcloned into the restriction sites HindIII (5') and XhoI (3') of the vector pT81-Luc (26)
, which contains the enhancerless herpes simplex TK viral promoter. All pT81 constructs were checked by restriction digest for the correct length of promoter segments and then confirmed by sequencing. Sequential deletion of the GC boxes from 5'- and 3'-end was performed and has been described in detail (26)
.
Transient Transfections.
TFK-1 and EGI-1 cells were transfected using the Effectene Transfection Reagent Kit (Qiagen). To correct for transfection efficiency, cotransfection with 100 ng/ml Renilla luciferase expression construct pRL-TK (Promega, Mannheim, Germany) was performed. After transfection, cells were deprived of serum and maintained in UltraCulture (Biowhittaker Inc., Walkersville, MD). Firefly and Renilla luciferase activities were detected in a monolight luminometer (EG Berthold, Bad Wildbach, Germany) using a dual luciferase reporter assay (Promega) according to the manufacturer’s instructions. Incubations were performed in triplicates, and the results were normalized for transfection efficiency. Values were expressed as the fold increase in luciferase activity compared with untreated controls.
EMSA.
Nuclear extracts from TFK-1 cells were prepared by using a nonionic detergent method. For the detection of DNA binding activity, nuclear protein extracts were incubated with radiolabeled double-stranded oligonucleotides comprising the -88 to -50 region of the hVEGF promoter. The following oligonucleotides were used in 100- or 200-fold excess as unlabeled competitors (5'
3'): (a) Sp1 consensus, attcgatcggggcggggcgagc; (b) Sp1 mutant, attcgatcggttcggggcgagc; (c) AP2 consensus, gatcgaactgaccgcccgcggcccgt; (d) AP2 mutant, gatcgaactgaccgcttgcggcccgt; (e) Egr-1 consensus, ggatccagcgggggcgagcgggggcga; and (f) Egr-1 mutant, ggatccagctagggcgagcgggggcga.
DNA binding reactions were performed as described previously (26) . For competition experiments, nuclear extracts were incubated with a 100x molar excess of double-stranded competitor oligonucleotide. For supershift experiments, nuclear extracts were incubated with 1 µg of anti-Sp1, anti-Sp3, anti-AP2, or anti-Egr-1 Abs for 10 min at room temperature before the addition of radiolabeled probes. DNA-protein complexes were electrophoresed on a 6% nondenaturing polyacrylamide gel, dried, and exposed to Kodak BioMax MR films (Amersham, Braunschweig, Germany) at -80°C.
Statistical Analysis.
Statistical differences were evaluated by one-way ANOVA using GraphPad statistical software (GraphPad Software Inc., San Diego, CA). Differences were considered statistically significant at P < 0.05. Unless otherwise stated, all data are expressed as mean ± SE.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. The MVD for the 19 tumor specimens was calculated at 126.5 ± 45.1 according to the procedure of Weidner et al. (23)
. Immunostaining with a polyclonal Ab specific for hVEGF revealed a strong expression of VEGF in epithelial tumor cells in 19 of 19 tumor specimens investigated (Fig. 1B)
. To further corroborate the cellular source of VEGF production, we performed in situ hybridization using a 35S-labeled cRNA probe specific for hVEGF. In analogy to the results obtained by immunohistochemistry, VEGF was strongly expressed in epithelial tumor cells of all 19 tumor specimens, whereas the surrounding mesenchymal and endothelial cells were consistently negative (Fig. 1, E and F)
. Although VEGF mRNA was consistently expressed over tumor cells throughout the tumor, a selective enhancement of the hybridization signal intensity was observed in the areas directly adjacent to necrotic foci (Fig. 1E)
. To confirm the specificity of the VEGF mRNA expression pattern, sense cRNA was analyzed in parallel, which revealed no specific hybridization pattern different from background (Fig. 1F)
.
|
and VEGFR-2 (Fig. 1D)
showed that both receptors were expressed exclusively in endothelial cells adjacent to tumor cells. In situ hybridization demonstrated tumor endothelial cells as the only cellular source of both VEGFR mRNA transcripts (Fig. 1, G and H)
. Again, no significant hybridization signal was observed using sense cRNA (data not shown).
Expression of TGF-ß1 and Its Receptors in Human CCC.
To investigate the potential interaction of VEGF with TGF-ß1, we next had to characterize the expression pattern for TGF-ß1 and its receptors in CCC. Immunostaining with a polyclonal Ab directed against TGF-ß1 revealed that all CCCs examined consistently expressed TGF-ß1 (19 of 19 specimens). Two distinct expression patterns were observed. In 15 of 19 tumor specimens, a strong TGF-ß1 signal was detected over mesenchymal cells surrounding tumor cells, but only weak or negligible immunostaining was seen over tumor cells (Fig. 2A)
. In contrast, 4 of 19 tissue samples showed strong TGF-ß1 expression over tumor cells, whereas surrounding mesenchymal cells demonstrated only weak immunoreactivity (Fig. 2B)
. Immunostaining with polyclonal Abs directed against TßR-I and TßR-II consistently revealed a strong signal for both receptors over transformed epithelial tumor cells in 19 of 19 tumor samples (Fig. 2, C and D)
. These data demonstrate that the majority of human CCCs coexpress VEGF and TGF-ß1 as well as their respective receptors, suggesting a possible functional interdependence between these two cytokines.
|
. In contrast, no VEGF145 and VEGF206 mRNA transcripts could be detected. Using ELISAs specific for hVEGF and TGF-ß1, we detected both cytokines in supernatants and cell lysates of the two cell lines (data not shown).
|
. A significant stimulation of VEGF by TGF-ß1 was observed as early as 48 h after incubation (2.3-fold increase) and persisted throughout the observation period of 96 h. Similar results were obtained in EGI-1 cells, although the TGF-ß1 concentrations required for maximal induction of VEGF were higher compared with TFK-1 cells (100 ng/ml). Based on TGF-ß1 expression in tumor cells in vivo and in vitro, we considered the possibility that TGF-ß1 might also stimulate VEGF expression in an autocrine manner. We therefore investigated the effects of a neutralizing TGF-ß1 Ab on VEGF protein concentration in cell culture supernatants. Compared with an isotype-matched control Ab, addition of a neutralizing TGF-ß Ab resulted in a significant inhibition of VEGF concentrations in TFK-1 supernatants to 61 ± 9% of control (n = 4; P < 0.05; Fig. 3C
). This observation was further confirmed in two independent cell lines (HepG2 and HuH7) derived from human liver (data not shown). This implies that exogenous as well as autocrine secreted TGF-ß1 is able to stimulate VEGF expression.
Molecular Mechanism of TGF-ß1-induced VEGF Expression in TFK-1 Cells.
To elucidate whether increased VEGF production in response to TGF-ß1 is based on an increase in VEGF mRNA concentrations, we performed QC RT-PCR. VEGF mRNA concentrations increased upon TGF-ß1 treatment in a time-dependent manner. Induction of VEGF mRNA concentrations was observed as early as 12 h, and a maximal stimulation was observed after 48 h of TGF-ß1 incubation (5.6 ± 1.9-fold; n = 3; P < 0.05; Fig. 4A
). We next investigated the effect of TGF-ß1 on VEGF gene transcription. A hVEGF promoter-reporter construct ranging from its 5'-end -2018 to +50 was expressed in TFK-1 and EGI-1 cells using transient transfection assays. Treatment with TGF-ß1 resulted in increased VEGF promoter activity in both cell lines, although the effects in EGI-1 cells did not reach statistical significance (Fig. 4B)
. Because TGF-ß1 was more potent in TFK-1 cells, we further determined the dose dependency of TGF-ß1 on VEGF promoter activity using this cell line. We observed a dose-dependent increase of VEGF promoter activity in reponse to TGF-ß1, with significant stimulation observed at 1 ng/ml, and maximal stimulation occurring at 100 ng/ml (Fig. 4C)
. For additional experiments, we used a concentration of 10 ng/ml TGF-ß1.
|
demonstrate that 5' deletion analysis down to -85/+50 sustained basal VEGF promoter activity and conferred TGF-ß1 responsiveness, although to a slightly variable degree. In contrast, removal of the VEGF promoter region -85/-53 decreased basal VEGF promoter activity by about 95% and completely abolished TGF-ß1 responsiveness. To further confirm that the hVEGF -85/-53 promoter region confers TGF-ß1 responsiveness, a heterologous promoter system was used. This construct contains an enhancerless TK promoter in front of the luciferase reporter gene (pTK-Luc). Cells transfected with empty pTK-Luc showed very low promoter activity and did not respond to TGF-ß1 treatment. Subcloning of the hVEGF -88/-50 sequence into this construct (pTK-Luc/hVEGF-88/-50) resulted in strong enhancement of basal promoter activity, suggesting that the hVEGF -88/-50 sequence functions as a basal enhancer of the VEGF gene in CCC cells. Furthermore, presence of the hVEGF-88/-50 element conferred TGF-ß1 sensitivity to the pTK-Luc construct, resulting in a 2-fold increase in reporter gene activity (Fig. 5B)
.
|
. To elucidate which of these transcription factors participates in TGF-ß1 responsiveness, a mutational analysis of the potential binding sites was performed. hVEGF-88/+50/Sp1-mut and hVEGF-85/+50/Egr-mut comprised two nucleotide mutations known to abolish Sp1 binding or Egr-1 binding, respectively (25)
. Transactivation assays in TFK-1 cells revealed that mutation of the potential Egr-1 binding sites did not abolish TGF-ß1 responsiveness (Fig. 6B)
. In contrast, inactivating mutations of all potential Sp1 consensus binding sites resulted in a substantial decrease of basal VEGF promoter activity and completely abolished TGF-ß1 responsiveness (Fig. 6B)
.
|
shows that all three constructs displayed constitutive basal promoter activity, although to a lesser extent than the -88/-50 construct. TGF-ß1 responsiveness decreased somewhat upon deletion of Sp1 binding site I but showed a significant loss when two Sp1 binding sites (I and II) or only the Sp1 binding site next to the -50 end (III) was deleted. These data indicate that Sp1 binding site I is dispensable, whereas Sp1 binding sites II and III are indispensable for TGF-ß1 responsiveness.
Sp1 and Sp3 Transcription Factors Bind to the -85 to -50 region of the VEGF Promoter.
We next investigated which members of the Sp transcription factor family are able to bind to the potential consensus binding sites. EMSAs using a 32P-labeled double-stranded oligonucleotide that covers the -88 to -50 region of the hVEGF promoter and nuclear extracts from TFK-1 cells demonstrated the formation of two prominent DNA-protein complexes (Fig. 7A
, Lane 1). The two complexes could be completely competed away by addition of an excess of unlabeled VEGF -88/-50 oligonucleotide or consensus Sp1 binding oligonucleotide, whereas using an oligonucleotide containing a mutated Sp1 binding site did not affect complex formation (Fig. 7A
, Lanes 2–4). In contrast, DNA-protein complex formation could be reduced but not completely eliminated using a 100-fold molar excess of unlabeled AP2 and Egr-1 oligonucleotides and was not affected by unlabeled, mutated AP2 or Egr-1 oligonucleotides (Fig. 7A
, Lanes 5–8). To confirm that any of the two nucleo-protein complexes represent the binding of Sp1, AP2, or Egr-1 transcription factors, gel shift analysis was performed using specific anti-Sp1, anti-AP2, and anti-Egr1 Abs. Furthermore, anti-Sp3 Abs were used, based on the fact that Sp1 binding sites also recognize Sp3 (27)
. Addition of Sp1 Ab resulted in a supershift of a major component of complex I, and a Sp3 Ab completely supershifted complex II (Fig. 7B
, Lanes 2 and 3). Simultaneous addition of Sp1 and Sp3 Abs resulted in a complete supershift of complex II and a major supershift of complex I. Addition of AP2 and Egr-1 Abs, however, did not result in altered nucleo-protein complex (Fig. 7B
, Lanes 5 and 6), although their ability to supershift has independently been confirmed (data not shown). These results suggest that complex I represents predominant binding of Sp1 (and to a lesser extent, Sp3), whereas complex II represents binding of Sp3 to the -88 to -50 segment of the hVEGF promoter.
|
.
TGF-ß1 Stimulates Sp1-dependent Transactivation of hVEGF Promoter in TFK-1 Cells.
Because TGF-ß1-stimulated VEGF transcription was not accompanied by a change in the composition of Sp1 and Sp3 DNA-protein complexes, we next investigated the effect of TGF-ß1 on Sp1- or Sp3-dependent transactivation. We used a system in which either Sp1 or Sp3 was fused to the DNA-binding domain of the yeast transcription factor Gal4. After cotransfection of Sp1/Gal4 or Sp3/Gal4 with a Gal4/luciferase construct, the ability of the Sp-Gal4 fusion proteins to activate Gal4 DNA binding can be assessed. Cotransfection of Sp1/Gal4 with Gal4/luciferase resulted in a significant 2.6-fold increase of Gal4/luciferase promoter activity in response to TGF-ß1 treatment (Fig. 7D)
. Cotransfection of Sp3/Gal4 resulted in basal luciferase activity comparable with cotransfection with Sp1/Gal4. However, TGF-ß1 treatment had no significant effect on Sp3-mediated transactivation (Fig. 7D)
. These results suggest that the transactivation domain of Sp1 mediates TGF-ß1-induced transcriptional activation of the hVEGF promoter.
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Thus far, determination of MVD has not been evaluated in CCC. Using an endothelial cell-specific Ab, we determined the mean MVD as 126 ± 45 vessels/0.74 mm2 (n = 19). This is well within the range of MVD observed in tumors commonly considered to be well vascularized, such as node-positive breast carcinoma (101 ± 49) or prostate carcinoma with metastatic disease (77 ± 32; Ref. 23 ), indicating a considerable degree of tumor-associated angiogenesis in human CCC.
In view of the significant vascularization of CCCs, the analysis of VEGF and its receptors appears of particular relevance (28) . Two previous immunohistochemical studies investigated VEGF expression in human CCC and yielded conflicting results (29 , 30) . We were able to demonstrate that all 19 human CCC samples investigated expressed VEGF protein. Immunohistochemistry detects intracellular as well as receptor-bound VEGF, thereby precluding unambiguous determination of the cellular source of VEGF. We therefore performed in situ hybridization and demonstrated the malignant cholangiocyte as the exclusive cellular source of VEGF expression in vivo. Furthermore, we observed pronounced VEGF mRNA expression in areas surrounding necrotic foci. This observation has also been described in other malignancies (22 , 31 , 32) . Enhanced VEGF expression in the areas adjacent to low oxygen tension suggests a possible role for local hypoxia as a stimulus for VEGF up-regulation in CCC (33) .
To the best of our knowledge, the expression of VEGFRs in CCC has not been investigated. By means of immunohistochemistry and in situ hybridization, we observed expression of VEGFR-1 in 15 of 19 tumor samples and expression of VEGFR-2 in 10 of 19 tumor samples. Both VEGFR mRNA and protein were exclusively expressed by endothelial cells. VEGFRs are rarely expressed in quiescent endothelial cells, whereas enhanced expression of both receptors is frequently observed in the context of tumor neovascularization (10 , 11 , 31 , 32 , 34) . Taken together, expression of both VEGFRs indicates that not only VEGF but also its receptors are up-regulated upon malignant transformation in CCC.
A variety of stimuli have been described as activators of VEGF gene expression in other cell systems and also appear conceivable to be responsible for the induction of VEGF (and angiogenesis) in human CCC: (a) tumor hypoxia, as evidenced by enhanced perinecrotic VEGF mRNA expression in our study; (b) genetic alterations such as K-ras-activating mutations and expression of mutated p53, which have been reported as common genetic alterations in CCC (35 , 36) and also have been demonstrated to up-regulate VEGF in transformed cell lines (37 , 38) ; and (c) finally, various cytokines such as epidermal growth factor, platelet-derived growth factor, interleukin 6, and TGF-ß1 as well as oxidative stress are also known to induce VEGF expression (26 , 39 , 40) . Among those, TGF-ß1 plays a prominent role in the pathogenesis of liver disease (41, 42, 43) . Recently, Yokomuro et al. (44) demonstrated that TGF-ß1 lost its antiproliferative and proapoptotic effect on human CCC cells, suggesting a possible role for TGF-ß1 in propagating the malignant phenotype of CCC. For other gastrointestinal malignancies, such as pancreatic carcinoma (45) and colon cancers (46) , it has been well established that mutations disabling a component of the TGF-ß1 signaling pathway promote the malignant phenotype of a given tissue. Induction of VEGF represents such a potential mechanism by which TGF-ß1 might act as a promoter of tumorigenesis (47, 48, 49, 50) . We therefore focused our studies on the functional relation between TGF-ß1 and VEGF in human CCC.
Previous studies identified mesenchymal cells as the main cellular source of TGF-ß1 in the liver (51) . We detected TGF-ß1 predominantly in mesenchymal cells (15 of 19 samples), but we also observed expression in tumor cells in a smaller percentage of samples (4 of 19 samples), suggesting that TGF-ß1 expression is controlled in a cell type-specific manner in a given CCC. This observation reconciles previous work with one report demonstrating TGF-ß1 immunoreactivity in mesenchymal cells in 25 of 30 CCCs (52) , whereas another study demonstrated TGF-ß1 immunoreactivity in CCC mainly over tumor cells (53) . The expression of the TßRs in CCC has not been investigated. We demonstrated strong immunoreactivity for both receptors over tumor cells in all tumor specimens, whereas only scarce immunostaining was seen in surrounding mesenchymal cells, indicating that tumor cells can act as a target for TGF-ß1 in human CCC. In extension to previously published reports, we now provide evidence for coexpression of VEGF, TGF-ß1, and their respective receptors in human CCC.
To further explore the postulated TGF-ß1-VEGF interaction, we established an in vitro system of two human CCC cell lines, TFK-1 and EGI-1. Using RT-PCR and VEGF- and TGF-ß1-specific ELISAs, we confirmed coexpression of VEGF and TGF-ß1 in both tumor cell lines. In addition, TGF-ß1 treatment resulted in a significant time-dependent increase of VEGF protein levels. This observation is consistent with previous data showing TGF-ß1-induced VEGF protein concentration in human osteoblasts (49) and in human keratinocytes (50) . Furthermore, we demonstrated that antagonizing endogenous TGF-ß1 using a neutralizing TGF-ß1 Ab results in inhibition of VEGF expression, which represents the first report of an autocrine and paracrine regulatory pathway for TGF-ß1-regulated VEGF expression.
Although the stimulatory effect of TGF-ß1 on VEGF expression has been described in a variety of in vitro systems, the molecular mechanisms responsible for this induction remain unsolved (40 , 47, 48, 49, 50 , 54) . We performed QC RT-PCR analysis and demonstrated that TGF-ß1 stimulated VEGF expression mainly at a pretranslational level.
Using a hVEGF promoter-reporter construct, we found that TGF-ß1 treatment results in a dose-dependent activation of VEGF promoter constructs in human CCC cell lines. Further 5' deletion analysis identified the VEGF promoter region -85 to -53 as responsible for basal VEGF promoter activity as well as TGF-ß1 responsiveness. Interestingly, this GC-rich region of the VEGF promoter has been demonstrated to represent the minimal region required for basal as well as regulated VEGF promoter activity in various cell systems, such as platelet-derived growth factor-mediated up-regulation in NIH3T3 cells (25) , p42/p44 mitogen-activated protein kinase-mediated up-regulation in fibroblasts (55) , interleukin 1-mediated up-regulation in cardiac myocytes (56) , p73-mediated down-regulation in leukemic cells (57) , and oxidative stress-mediated induction in gastric cancer cells (26) .
Mutational analysis of the putative transcription factor binding sites demonstrated that the Sp1/Sp3 binding sites but not the Egr-1 binding sites are required for TGF-ß1-mediated transactivation of the hVEGF promoter. Sequential deletion analysis of the three Sp1/Sp3 binding sites revealed that Sp1 binding site I was dispensable, whereas Sp1 binding sites II and III were indispensable for TGF-ß1-mediated transactivation. However, all three binding sites were required for maximal TGF-ß1 responsiveness, a situation similar to that observed for oxidative stress in gastric cancer cells (26) .
It has recently become clear that, aside from the Smad binding elements, canonical Sp1 sites can function as TGF-ß1-responsive elements for gene activation (58) . Although not completely understood, this is due to functional cooperation of Smad proteins with Sp1, which involves the physical interaction of these two transcription factor families (59, 60, 61, 62) and occurs independent from direct association of Smad proteins with DNA (62) . The molecular mechanisms underlying this functional cooperativity between Smads and Sp1 have not yet been fully elucidated, although several potential modes of action have been proposed (59, 60, 61, 62) .
We currently do not know which of these proposed mechanisms is operative in TGF-ß-mediated induction of VEGF gene transcription. However, our EMSA analysis revealed no change in DNA binding of Sp1 or Sp3 upon TGF-ß1 treatment, suggesting that a mechanism other than altering the DNA binding affinity of Sp transcription factors is responsible for TGF-ß-mediated transactivation. We therefore analyzed the transactivating capability using fusion proteins of Sp1 and Sp3 with the DNA binding domain of the yeast transcription factor Gal4. We observed that expression of fusion proteins between Gal4 DNA binding domain and Sp1 but not Sp3 renders the Gal4 reporter construct TGF-ß1 responsive. Thus, Sp1 but not Sp3 can be activated by TGF-ß1 treatment to increase transcription. Therefore, in addition to "classical" TGF-ß target genes (60, 61, 62) , VEGF represents the first proangiogenic factor that is transcriptionally regulated by TGF-ß1 through the Sp1 family of transcription factors.
In summary, we provided evidence for coexpression of VEGF and TGF-ß1 as well as their respective receptors upon malignant transformation in human CCC. Furthermore, TGF-ß1 can stimulate VEGF gene transcription in malignant cholangiocytes in a paracrine and/or autocrine manner through an Sp1-dependent mechanism. This regulatory pathway might contribute to the "angiogenic switch" and the malignant phenotype of human CCC.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 S. R. was supported by grants from Deutsche Forschungsgemeinschaft, Deutsche Krebshilfe, Berliner Krebshilfe, Sander Stiftung, Else Kröner Fresenius Stiftung, and Sonnenfeld Stiftung. M. H. was supported by Deutsche Forschungsgemeinschaft and the Bundesministerium für Bildung und Forschung (NBLIII). 
2 To whom requests for reprints should be addressed, at Department of Hepatology, Gastroenterology, Endocrinology and Metabolism, Charité, Campus Virchow-Klinikum, Humboldt-University, Augustenburger Platz 1, 13353 Berlin, Germany. Phone: 49-30-450-553733; Fax: 49-30-450-553940; E-mail: stefan.rosewicz{at}charite.de
3 The abbreviations used are: CCC, cholangiocellular carcinoma; VEGF, vascular endothelial growth factor; VEGFR, VEGF receptor; TGF, transforming growth factor; EMSA, electrophoretic mobility shift assay, MVD, microvessel density; QC, quantitative competitive; RT-PCR, reverse transcription-PCR; Ab, antibody; TßR, TGF-ß1 receptor; hVEGF, human VEGF; TK, thymidine kinase; EGR, early growth response protein; AP-2, activator protein 2.
Received 5/20/02. Accepted 12/27/02.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
2(I) collagen expression in human glomerular mesangial cells. J. Biol. Chem., 276: 6983-6992, 2001.This article has been cited by other articles:
|
|
 |
|
  V. J. HASSID, F. A. ORLANDO, Z. T. AWAD, D. TAN, T. KHOURY, B. H. AHMED, and S. J. ALRAWI Genetic and Molecular Abnormalities in Cholangiocarcinogenesis Anticancer Res, April 1, 2009; 29(4): 1151 - 1156. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. L. Clifford, K. Deacon, and A. J. Knox Novel Regulation of Vascular Endothelial Growth Factor-A (VEGF-A) by Transforming Growth Factor {beta}1: REQUIREMENT FOR Smads, {beta}-CATENIN, AND GSK3{beta} J. Biol. Chem., December 19, 2008; 283(51): 35337 - 35353. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Basu, A. G. Contreras, D. Datta, E. Flynn, L. Zeng, H. T. Cohen, D. M. Briscoe, and S. Pal Overexpression of Vascular Endothelial Growth Factor and the Development of Post-Transplantation Cancer Cancer Res., July 15, 2008; 68(14): 5689 - 5698. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Thelen, A. Scholz, C. Benckert, W. Weichert, E. Dietz, B. Wiedenmann, P. Neuhaus, and S. Jonas Tumor-Associated Lymphangiogenesis Correlates with Lymph Node Metastases and Prognosis in Hilar Cholangiocarcinoma Ann. Surg. Oncol., March 1, 2008; 15(3): 791 - 799. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Obama, S. Satoh, R. Hamamoto, Y. Sakai, Y. Nakamura, and Y. Furukawa Enhanced Expression of RAD51 Associating Protein-1 Is Involved in the Growth of Intrahepatic Cholangiocarcinoma Cells Clin. Cancer Res., March 1, 2008; 14(5): 1333 - 1339. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Ishida, J.-L. Gao, and P. M. Murphy Chemokine Receptor CX3CR1 Mediates Skin Wound Healing by Promoting Macrophage and Fibroblast Accumulation and Function J. Immunol., January 1, 2008; 180(1): 569 - 579. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Jungert, A. Buck, G. v. Wichert, G. Adler, A. Konig, M. Buchholz, T. M. Gress, and V. Ellenrieder Sp1 Is Required for Transforming Growth Factor-{beta}-Induced Mesenchymal Transition and Migration in Pancreatic Cancer Cells Cancer Res., February 15, 2007; 67(4): 1563 - 1570. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S.-H. Jeon, B.-C. Chae, H.-A Kim, G.-Y. Seo, D.-W. Seo, G.-T. Chun, N.-S. Kim, S.-W. Yie, W.-H. Byeon, S.-H. Eom, et al. Mechanisms underlying TGF-{beta}1-induced expression of VEGF and Flk-1 in mouse macrophages and their implications for angiogenesis J. Leukoc. Biol., February 1, 2007; 81(2): 557 - 566. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Jungert, A. Buck, M. Buchholz, M. Wagner, G. Adler, T. M. Gress, and V. Ellenrieder Smad-Sp1 complexes mediate TGF{beta}-induced early transcription of oncogenic Smad7 in pancreatic cancer cells Carcinogenesis, December 1, 2006; 27(12): 2392 - 2401. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. McMahon, M. Charbonneau, S. Grandmont, D. E. Richard, and C. M. Dubois Transforming Growth Factor beta1 Induces Hypoxia-inducible Factor-1 Stabilization through Selective Inhibition of PHD2 Expression J. Biol. Chem., August 25, 2006; 281(34): 24171 - 24181. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Gu, S. Hagiwara, Q. Fan, M. Tanimoto, M. Kobata, M. Yamashita, T. Nishitani, T. Gohda, Z. Ni, J. Qian, et al. Role of receptor for advanced glycation end-products and signalling events in advanced glycation end-product-induced monocyte chemoattractant protein-1 expression in differentiated mouse podocytes Nephrol. Dial. Transplant., February 1, 2006; 21(2): 299 - 313. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Charalambous, L. B. Pen, Y. S. Su, J. Milan, T. C. Chen, and F. M. Hofman Interleukin-8 Differentially Regulates Migration of Tumor-Associated and Normal Human Brain Endothelial Cells Cancer Res., November 15, 2005; 65(22): 10347 - 10354. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. Bradbury, D. Clarke, C. Seedhouse, L. Corbett, J. Stocks, and A. Knox Vascular Endothelial Growth Factor Induction by Prostaglandin E2 in Human Airway Smooth Muscle Cells Is Mediated by E Prostanoid EP2/EP4 Receptors and SP-1 Transcription Factor Binding Sites J. Biol. Chem., August 26, 2005; 280(34): 29993 - 30000. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Pages and J. Pouyssegur Transcriptional regulation of the Vascular Endothelial Growth Factor gene-a concert of activating factors Cardiovasc Res, February 15, 2005; 65(3): 564 - 573. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Colombel, S. Filleur, P. Fournier, C. Merle, J. Guglielmi, A. Courtin, A. Degeorges, C. M. Serre, R. Bouvier, P. Clezardin, et al. Androgens Repress the Expression of the Angiogenesis Inhibitor Thrombospondin-1 in Normal and Neoplastic Prostate Cancer Res., January 1, 2005; 65(1): 300 - 308. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Motomura, H. Kanbayashi, W. I. Khan, Y. Deng, P. A. Blennerhassett, P. J. Margetts, J. Gauldie, K. Egashira, and S. M. Collins The gene transfer of soluble VEGF type I receptor (Flt-1) attenuates peritoneal fibrosis formation in mice but not soluble TGF-{beta} type II receptor gene transfer Am J Physiol Gastrointest Liver Physiol, January 1, 2005; 288(1): G143 - G150. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Bostrom, A. F. Zebboudj, Y. Yao, T. S. Lin, and A. Torres Matrix GLA Protein Stimulates VEGF Expression through Increased Transforming Growth Factor-{beta}1 Activity in Endothelial Cells J. Biol. Chem., December 17, 2004; 279(51): 52904 - 52913. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Wang, J. H. Kwak, S. I. Kim, Y. He, and M. E. Choi Transforming Growth Factor-{beta}1 Stimulates Vascular Endothelial Growth Factor 164 via Mitogen-activated Protein Kinase Kinase 3-p38{alpha} and p38{delta} Mitogen-activated Protein Kinase-dependent Pathway in Murine Mesangial Cells J. Biol. Chem., August 6, 2004; 279(32): 33213 - 33219. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-L. Shie, G. Wu, J. Wu, F.-F. Liu, R. J. Laham, P. Oettgen, and J. Li RTEF-1, a Novel Transcriptional Stimulator of Vascular Endothelial Growth Factor in Hypoxic Endothelial Cells J. Biol. Chem., June 11, 2004; 279(24): 25010 - 25016. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y. Ishida, T. Kondo, T. Takayasu, Y. Iwakura, and N. Mukaida The Essential Involvement of Cross-Talk between IFN-{gamma} and TGF-{beta} in the Skin Wound-Healing Process J. Immunol., February 1, 2004; 172(3): 1848 - 1855. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |
