| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Tumor Biology |
Molecular Biology Institute [B. D., C. T. D.], Department of Pediatrics, Gwynne Hazen Cherry Memorial Laboratories [H. H., C. T. D.], and Jonsson Comprehensive Cancer Center [C. T. D.], University of California Los Angeles, Los Angeles, California 90024
 |
ABSTRACT |
|---|
 Top ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Tumor-associated chromosomal translocations found in EFTs3 fuse the NH2 terminus of the EWS gene to the COOH terminus of one of five members of the ETS family of transcription factors (FLI1, ERG, ETV, FEV, and E1AF; reviewed in Ref. 2 ). These EWS/ETS fusion proteins have been shown to play a dominant oncogenic role in cellular transformation and are required for growth of EFT cell lines (3, 4, 5) . Given that the NH2 terminus of EWS can behave as a potent transcriptional activation domain, it is believed that EWS/ETS fusions can function as aberrant ETS proteins (6) . This notion is supported by the observation that normal FLI1 can neither transform NIH-3T3 cells nor can modulate the same target genes as EWS/FLI1 (7 , 8) . Thus, it is likely that EWS/ETS proteins promote oncogenesis by transcriptionally modulating a set of target genes that are quantitatively and/or qualitatively distinct from those of their normal ETS counterparts. Some of these genes have been identified and shown to participate in EWS/ETS-mediated cellular transformation (9 , 10) .
Using microarray analysis, we have recently identified a more comprehensive cohort of target genes that were commonly regulated by three structurally distinct EWS/ETS fusions (11) . In evaluating these genes, many could be functionally categorized into groups such as regulators of cell cycle, signal transduction, or various metabolic processes. Trying to distinguish those target genes that promote EWS/ETS oncogenesis from those that are biologically less involved remains a challenge and is not easily discerned from a gene’s primary biochemical function.
In this work, we demonstrate that the metabolic regulator uridine phosphorylase is a biologically relevant EWS/ETS target gene. This gene encodes a pyrimidine nucleoside phosphorylase that is involved in the recycling of nucleotides. This enzyme catalyzes the reversible conversion of uridine nucleosides to uracil and ribose-1-phosphate (Ref. 12 ; UPase E.C. 2.3, 2.4). Uridine phosphorylase had previously been shown to be increased in numerous murine and human tumors and tumor cell lines and is up-regulated by activation of RAS (13, 14, 15) . We now show that ectopic expression of uridine phosphorylase alone is sufficient to promote anchorage-independent growth of NIH-3T3 cells. The mode of EWS/FLI1 regulation was also examined, and it was shown that modulation of uridine phosphorylase is DNA binding dependent and that EWS/FLI1 and RAS transcriptionally up-regulate a uridine phosphorylase promoter construct through a RRE. Chromatin immunoprecipitation experiments revealed that EWS/FLI1 physically interacts with the uridine phosphorylase promoter in vivo. Finally, expression of uridine phosphorylase renders EWS/FLI1-expressing cells susceptible to the pyrimidine analogue 5'-deoxy-5'-fluorouridine both in tissue culture and in vivo.
|
MATERIALS AND METHODS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-MSV tkneo and have been described previously (5
, 16)
. The EWS/FLI1 ZnSO4-inducible NIH-3T3 cell line used was described previously (7)
. Such cells were treated with 50 µM ZnSO4, and RNA was harvested at the given time points. The uridine phosphorylase promoter region was cloned using PCR primers specific for the promoter from genomic DNA isolated from NIH-3T3 cells. The isolated promoter was cloned into the basic pGL-Luciferase vector (Promega) Kpn/SacI. The various truncated promoter regions were cloned using PCR, with the full-length promoter as the DNA template, and also cloned Kpn/SacI into the basic pGL-Luciferase vector. Uridine phosphorylase cDNA was cloned using RT-PCR with mRNA harvested from NIH-3T3 cells. The UPase-300 
RRE construct was created using primers flanking RRE orientated in opposing directions. The subsequent PCR products contained blunted ends adjacent to the RRE and Kpn or SacI sites at the other end. These fragments were cloned together into the PGL-basic vector using the blunted ends to join the fragments together. The resulting fragment was sequence verified. The resultant cDNA was cloned into the replication-deficient retroviral construct pSR-MSV tkneo using EcoRI/HindIII and sequence verified. NIH-3T3 cells were passed under standard conditions in 5% calf serum DMEM that contains high glucose and glutamine.
Transfections and Infections.
To make the retroviral stocks, 293T cells were transfected with the retroviral vector and a 
-packaging plasmid as described previously (16)
. Briefly, 15 µg of the pSR constructs were CaPO4 transfected with 15 µg of ecotropic -packaging plasmid into 293T cells. After 2 days, viral collections were made for 24 h, filtered, and applied to cells. After infections, cells were selected with 450 µg/ml G418 for a period of 7–10 days.
Western and Northern Analysis.
Western and Northern blots were performed using standard procedures. For protein analysis, cell lysates were made using NP40 lysis buffer (50 mM Tris, 300 mM NaCl, 10% glycerol, and 1 mM EDTA). Eighty µg of total protein lysates were run on denaturing SDS gels, transferred to nitrocellulose membranes on a Bio-Rad semidry transfer apparatus, and decorated the -Flag monoclonal antibody (Sigma) in 5% milk, 1x TBS-T. Horseradish peroxidase-conjugated goat-antimouse secondary antibody (Transduction Labs) and Western Blot Chemiluminescent Reagent (NEN) were used for detection. Coomassie staining of the gel and ponceau staining of the membrane were used to ensure equal loading. For Northern blots, RNA lysates were made using Stat-60 (Tel-Test). Approximately 8 µg were run under denaturing conditions on a 1% agarose-formaldehyde gel, transferred to nitrocellulose membrane by capillary methods, and hybridized and washed using standard methods. Ethidium staining of the gel was used to ensure equal loading. The murine uridine phosphorylase cDNA probe was generated through RT-PCR of nucleotides 391-1017. The human cDNA uridine phosphorylase probe was generated through RT-PCR of nucleotides 446–908.
Transformation Assays and Tumor Assays.
A total of 1 x 106 NIH-3T3 fibroblasts expressing EWS/FLI1, uridine phosphorylase, or empty vector were s.c. injected into the nape region of the neck of CB17 SCID mice. All mice were female and 
6–8 weeks of age at the time of injection. Mice were observed until a grossly visible tumor of 1.5 cm in diameter was present, at which time the animals were sacrificed, and the tumors were dissected. Tumor cell lines were created by culturing minced tumors until cell lines were established and expanded. Agar assays were performed as described previously (16)
. Briefly, 5 x 104 NIH-3T3 cells expressing uridine phosphorylase, EWS/FLI1, or empty vector were seeded in semi-solid agar in high serum conditions (20%) and incubated for a period of 4 weeks. Photographs were taken at days 21 and 28; 21-day pictures are shown.
Reporter Assays.
A total of 1.25 µg of the pGL-Luciferase vectors containing the uridine phosphorylase promoter fragments was cotransfected with equal amounts of the pRL-SV40-Renilla Luciferase control into NIH-3T3 cells stably expressing EWS/FLI1, EWS/FLI1–340, RAS, or empty vector using superfect reagent (Qiagen). Forty-eight h later, cell lysates were collected using a passive lysis buffer (Promega). Firefly luciferase and the control renilla luciferase were detected using a dual luciferase detection kit (Promega) and a luminometer. Firefly luciferase values were normalized to one another by comparison with the renilla luciferase values.
Chromatin Immunoprecipitation Assays.
Ten 15-cm2 plates were seeded with the appropriate cell line and growth to confluency. Cells were cross-linked in 1% formaldehyde for 15 min at room temperature, collected, and B dounced on ice for 10 min. Nuclei were pelleted by microcentrifugation at 4°C and lysed by incubation in nuclear lysis buffer [1% SDS, 10 mM EDTA, 50 mM Tris-chloride (pH 8.1), 1 mM phenylmethylsulfonyl fluoride, 10 ng/ml aprotinin, and 10 ng/ml leupeptin]. The resulting chromatin solution was sonicated for four 15-s pulses (model 300; Fisher Sonice Dismembrator) to an average length of 0.5–1 kb and incubated overnight on ice at 4°C. After microcentrifugation, the supernatant was precleared with protein A-Sepharose 4B (Amersham Bioscience) for 2 h at 4°C. Supernatants were incubated with 80 µl antiflag M2-agarose at 4°C overnight (Sigma), extensively washed, and eluted (1% SDS, 50 mM NaHCO3). Half of the eluent (100 µl) was diluted in 900 µl of 1x PBS with protease inhibitors. The antiflag immunoprecipitation step and elution were repeated on the 1-ml sample. After the second immunoprecipitation, elutes were pooled, proteinase K treated at 37°C overnight, and heated at 65°C to reverse the formaldehyde cross-linking. DNA fragments were purified by a chloroform/phenol extraction and an ethanol precipitation, resuspended in water, and analyzed by PCR. PCR was performed using P32-dATP nucleotides, and products were separated on 6% polyacrylamide gels and visualized by autoradiography. Primer sequences: region A (-1117:-767) GGTGACCTTAGCCGAGAGTG(F) and AAGGCAAGAGCACCAGAGAA(R), region B (-867:-467) CAGGAACAGGGAAAGAGCAG(F) and AGCATGAGCCGTAGCAGAAC(R), region C (-567:-167) GCCACAGACTTTGATGATGG(F) and GCCCAGTAGGGGAAATGACT(R), region D (-317:-17) ATTCCCAAGCCTTGTCCTTT(F) and GTACCAGAGATGAATGGCCG(R), and mitochondria primers within the 16S RNase GTACCGCAAGGGAAAGATGA(F) and AGGTAGCTCGTTTGGTTTCG(R).
Growth Inhibition Assays.
A total of 2.5 x 104 NIH-3T3 cells expressing EWS/FLI1, EWS/FLI1–340, uridine phosphorylase, and empty vector were plated into 24-well plates. The next day, the normal media was exchanged for media containing the appropriate concentration of 5'deoxy-5'fluorouridine or 5'deoxy-5'fluorouracil (Sigma). The media was changed daily for 3 days, at which point the cells were treated with 400 µl of 1 mg/ml solution MTT/RPMI (Sigma/Invitrogen). The cells were incubated for 3 h at 37°C, at which time a solubility solution (1% Triton X-100, 0.04 M HCL, and 95% isopropanol) was added to the MTT solution. Absorbance of the resulting solution was measured at 540 nm on a microplate reader. A standard curve was established by plating known quantities of cells, treating these cells with the MTT/solubility solution, and measuring the absorbance at 540 nm. The number of cells in the experimental samples was determined by comparing their respective 540-nm absorbance with the standard curve.
|
RESULTS |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. Relative levels of induction for each EWS/ETS fusion paralleled those seen on microarray analyses (data not shown). Under these same conditions, normal FLI1 had no apparent inductive effect of uridine phosphorylase (Fig. 1B)
. These data indicate again the differences between normal ETS proteins and abnormal EWS/ETS fusions at the target gene level.
|
. Subsequent Northern analysis of these RNAs revealed near full induction of uridine phosphorylase by 8 h of Zn2+ stimulation in the EWS/FLI1 containing cells and no induction in the empty vector controls.
To confirm uridine phosphorylase was actually expressed in EFT cells, Northern analysis was performed on RNAs collected from three different EFT tumor-derived cell lines (Fig. 1D)
. HeLa cells, which have been previously shown to express high levels of uridine phosphorylase, served as a positive control (14)
. Uridine phosphorylase was found to be highly expressed in all three EFT cell lines. The multiple RNA species seem likely to represent alternatively spliced variants that have been described previously (14)
.
Ectopic Expression of Uridine Phosphorylase Supports Anchorage-independent Growth.
The induction of uridine phosphorylase by three different EWS/ETS proteins and its expression in EFT cell lines suggested that uridine phosphorylase could be contributing to EFT oncogenesis. Moreover, uridine phosphorylase is highly expressed in numerous murine and human tumors and tumor-derived cell lines, again suggesting a potential role in cellular transformation (14)
. Given these observations, we sought to examine the transformation potential of uridine phosphorylase in NIH-3T3 model systems.
A full-length uridine phosphorylase cDNA was obtained by RT-PCR from RNA extracted from NIH-3T3 cells expressing EWS/FLI1 using specific primers bracketing the coding region. The isolated cDNA was sequence verified, NH2-terminal epitope tagged, and cloned into a retroviral expression vector. NIH-3T3 cells were transduced with replication-deficient retroviruses containing empty vector, EWS/FLI1, or uridine phosphorylase. Polyclonal populations were selected using G418, and expression of uridine phosphorylase was confirmed through immunoblot analysis (Fig. 2A)
. A Mr 35,000 protein was detected in uridine phosphorylase transduced cells, which is consistent with the previously documented size of the murine uridine phosphorylase protein (15)
. Northern analysis revealed a 50-fold increase in the levels of ectopic uridine phosphorylase expression when compared with endogenous levels induced by EWS/FLI1 (data not shown).
|
. Colony number and size displayed by the uridine phosphorylase-expressing cell lines were comparable with the EWS/FLI1 expressing cell lines. This was a consistent finding over two independent uridine phosphorylase-transduced cell populations. Although uridine phosphorylase was capable of supporting anchorage-independent cell growth, it was unable to accelerate tumorigenesis in SCID mice. The same polyclonal cell lines used in the agar assays were injected s.c. into SCID mice to assess their tumorigenic potential. Although the EWS/FLI1 controls rapidly formed tumors, there was no difference in the rate of tumor formation between the empty vector and uridine phosphorylase expressing cell lines (data not shown).
EWS/FLI1 Modulates Uridine Phosphorylase in a DNA Binding-dependent Manner.
Because uridine phosphorylase displays transforming characteristics suggests that it could be a biologically significant component of the EWS/ETS target gene network. Furthermore, the observation that uridine phosphorylase transcript levels rise within 8 h of EWS/FLI1 induction suggested that EWS/FLI1 might be modulating uridine phosphorylase by directly binding transcriptional regulatory elements within the uridine phosphorylase promoter. Mutagenesis and promoter analyses were performed to better understand how EWS/ETS fusions modulate uridine phosphorylase expression.
First, the dependence on EWS/ETS DNA binding capability on uridine phosphorylase modulation was tested using the EWS/FLI1-340 point mutant. This mutant contains an arginine to asparagine substitution at amino acid position 340 and is unable to bind ETS consensus sequences in vitro (16)
. As before, NIH-3T3 cells were transduced with retrovirus containing epitope-tagged EWS/FLI1, EWS/FLI1-340, or empty vector. After antibiotic selection, immunoblot analyses were performed to confirm expression of the appropriate proteins (data not shown). Northern analysis from these cells revealed that in NIH-3T3 cells containing EWS/FLI1-340, uridine phosphorylase transcript levels were comparable with the empty vector controls (Fig. 3A)
. Expression levels of EWS/FLI1-340 have been previously shown to be significantly higher than those of EWS/FLI, indicating that such defects in uridine phosphorylase modulation are likely because of defects in DNA binding and not levels of protein expression (16)
. These data indicate that EWS/FLI1 modulation of uridine phosphorylase requires an intact ETS DNA binding domain.
|
. NIH-3T3 cells stably expressing EWS/FLI1, EWS/FLI1-340, and empty vector were transiently transfected with the appropriate promoter and control constructs. Cell lysates were harvested 48 h later and assayed for luciferase activity. EWS/FLI1 activates the full-length (-1117) uridine phosphorylase promoter 2.3-fold over the empty vector, whereas the EWS/FLI-340 mutant showed no increase in luciferase activity (Fig. 3C)
. This activation was seen in the -567 and -300 deletion constructs, where EWS/FLI1 activated the promoter constructs 2.3- and 2.2-fold, respectively. Activation of the uridine phosphorylase promoter by EWS/FLI1 is lost in the -120 deletion construct. These data indicate that an EWS/FLI1 response region existed in the 180-bp region between -300 and -120.
EWS/FLI1 and RAS Response Functionally Maps to a Common Domain within the Uridine Phosphorylase Promoter.
Previous studies have shown that overexpression of RAS leads to strong transcriptional up-regulation of uridine phosphorylase (15)
. To determine whether RAS and EWS/FLI1 modulated uridine phosphorylase by similar or distinct mechanisms, reporter gene assays were performed. First, the RAS response region in the uridine phosphorylase promoter was mapped in the same manner using the same nested deletion constructs used for EWS/FLI1 studies (see above). NIH-3T3 cells stably expressing RAS or empty vector control were transiently transfected with one of three uridine phosphorylase promoter reporter constructs. As before, cell lysates were collected 48 h after transfection, and luciferase activity was assessed. As with EWS/FLI1, RAS activated the -1117, -567 and -300 promoter fragments to approximately the same extent: 3.3-, 3.3-, and 3.1-fold, respectively (Fig. 4A)
. Also consistent with EWS/FLI1, the -120 promoter fragment failed to respond to activated RAS.
|
RRE). As before, this construct was transiently transfected into NIH-3T3 cells containing EWS/FLI1, activated RAS, or empty vector, and luciferase assays were performed on resultant cell lysates (Fig. 4B)
. Deletion of the RRE converted a promoter fragment with near full activity to one that was unresponsive to either EWS/FLI1 or RAS. These data indicate that EWS/FLI1 and RAS map to the same response region for transcriptional activation of uridine phosphorylase promoter constructs.
EWS/FLI1 Directly Interacts with the Uridine Phosphorylase Promoter in Vivo.
That EWS/FLI1 and RAS share a common response region within the uridine phosphorylase promoter suggests that EWS/FLI1 might use intermediate RAS-signaling mechanisms to modulate uridine phosphorylase. This would be consistent with a previous observation that EWS/FLI1 up-regulates RAS signaling through the MAP kinase kinase/ERK pathway (19)
. Alternatively, EWS/FLI1 may modulate uridine phosphorylase through direct physical association with regulatory elements within the uridine phosphorylase promoter. Given that the EWS/FLI1-RAS response region contains a RRE, an in vivo DNA substrate of ETS family members, we sought to examine whether EWS/FLI1 directly associates with the uridine phosphorylase promoter (18)
.
Chromatin immunoprecipitation assays were performed to determine whether EWS/FLI1 binds the uridine phosphorylase promoter in vivo. NIH-3T3 cells stably expressing NH2-terminal double flag-tagged EWS/FLI1 or empty vector were formaldehyde cross-linked and subjected to a two cycle anti-Flag immunoprecipitation (see "Materials and Methods"). To demonstrate an equal quantity of resultant PCR-amplifiable template across all samples, we used primers to mitochondria DNA as a measure of nonspecific immunoprecipitated product (Fig. 5)
. Site-specific PCR was used to assay for any relative increase in the presence of the uridine phosphorylase promoter in the EWS/FLI1 versus empty vector immunoprecipitated chromatin preparations. In two independent assays, the EWS/FLI1-expressing cell lines displayed significant enrichment of the uridine phosphorylase promoter within the proximal regulatory regions when compared with the empty vector control. These data indicate that EWS/FLI1 is able to directly interact with the uridine phosphorylase promoter in vivo.
|
NIH-3T3 cells expressing empty vector, EWS/FLI1, and EWS/FLI1-340 were plated at uniform density and treated with concentrations of 5dFUrd, ranging from 5 to 1000 µM. After 3 days of treatment, cells expressing EWS/FLI1 displayed increased sensitivity to 5dFUrd-induced growth inhibition when compared with empty vector and the EWS/FLI1-340 mutant (Fig. 6A)
. The IC50 of EWS/FLI1-expressing cells was 7 µM, whereas cells containing empty vector and EWS/FLI-340 showed an IC50 of 43 µM, representing a 6-fold difference in 5dFUrd sensitivity between these cell lines.
|
.
Finally, we wished to determine whether the sensitivity of EWS/FLI1-expressing cells to 5dFUrd was directly because of up-regulation of uridine phosphorylase and not simply a pleotropic effect of 5dFUrd. To address this issue, NIH-3T3 cells ectopically expressing uridine phosphorylase (see Figure 2
) were treated with 5dFUrd, and growth inhibition was measured. Cells stably transduced with a uridine phosphorylase expression vector proved to be the most sensitive to 5dFUrd growth inhibition and demonstrated an IC50 of <1 µM (Fig. 6C)
. The difference in the sensitivity to 5dFUrd between NIH-3T3 cells expressing uridine phosphorylase and EWS/FLI1 directly correlates with the relative differences in uridine phosphorylase expression level between the two cell lines. These data indicate that the growth inhibition displayed in EWS/FLI1-expressing cells directly correlate with an increase in uridine phosphorylase activity. In sum, these results demonstrate that EWS/FLI1 up-regulation of uridine phosphorylase renders cells more sensitive to 5dFUrd.
5'Deoxy 5'fluoro Uridine Attenuates EWS/FLI1-mediated Tumor Formation.
The ability to specifically target EWS/FLI1-expressing cells in tissue culture with 5dFUrd suggests that this drug might be effective at inhibiting EWS/FLI1-mediated tumorigenesis. Thus, we sought to extend our finding to an in vivo tumor model system. Because of the difficulty in developing a reliable xenograft tumor model system based on EFT cell lines, a previously described NIH-3T3 strategy was used (3
, 5)
. In this model system, NIH-3T3 cells expressing EWS/FLI1 or empty vector control were injected s.c. into SCID mice, and the time to tumor formation under various conditions was determined. Mice were treated with either a mock injection of 1x PBS/10% DMSO or 50 mg/kg 5dFUrd in 1x PBS/10% DMSO, administered i.p. Treatments began 3 days after cell injection, with the mice receiving treatments days 3–7 and days 10–12 for a total of eight treatments.
The cohort of mice that was treated with 5dFUrd showed a significant delay in the rate of tumor formation (Fig. 7)
. Mice treated with the mock injection formed 1.5-cm tumors on an average of 14.8 days, whereas mice treated with 5dFUrd formed 1.5-cm tumors on an average of 23.3 days. This data indicates that 5dFUrd can effectively target EWS/FLI1-expressing cells in an in vivo tumor system.
|
|
DISCUSSION |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
50-fold higher than the levels observed in EWS/FLI1-transformed cells. Nevertheless, it is surprising that any transformation effect was evident at all. Anchorage-independent growth and tumor induction are different measures of aberrant cell growth and are likely to require different biological characteristics that are not necessarily overlapping. Although anchorage-independent growth measures the capacity for growth without supportive extracellular matrix, tumor formation in SCID mice imposes additional cellular demands, including invasion of surrounding tissues and neovascularization. Differences between these assays have been observed for EWS/ETV1 and various EWS/FLI1 DNA binding mutants (5 , 16) . It may be that uridine phosphorylase contributes to pathways that initiate cellular proliferation and not to pathways required for tumor induction or maintenance.
Given the differences in cellular physiology between normal and malignant cells, shifts in metabolic gene regulation are often found in cellular transformation. However, the capacity of an individual metabolic gene such as uridine phosphorylase to promote cellular transformation is unexpected. Among the few examples that link metabolic gene function to cellular transformation include ornithine decarboxylase and thymidine phosphorylase. Ornithine decarboxylase functions in polyamine biosynthesis, which is critical for cellular proliferation and protects the cell from various extracellular insults (20 , 21) . Interestingly, ornithine decarboxylase is a target gene of C-MYC and promotes cell survival through the inhibition of apoptosis (22) . Thymidine phosphorylase (TP) is identical to the angiogenic platelet-derived endothelial cell growth factor and thymidine phosphorylase enzymatic activity is required for the angiogenic activity of platelet-derived endothelial cell growth factor/TP (23 , 24) . That the enzymatic activities of thymidine phosphorylase and uridine phosphorylase are similar could result in similar angiogenic functions between these proteins (21) . It may be that the association of pyrimidine nucleoside phosphorylase activity with angiogenesis is analogous to the association of polyamine biosynthesis with cell survival. Another possibility is that uridine phosphorylase is promoting cellular transformation by impacting other physiological mechanisms directly linked to cellular proliferation or survival. Additional examination of the impact of uridine phosphorylase overexpression on global cellular metabolism may prove insightful in delineating these relationships.
The kinetics of uridine phosphorylase transcriptional modulation by EWS/FLI1 suggested that uridine phosphorylase could be a direct target of EWS/ETS fusions. Chromatin immunoprecipitation assays revealed a direct in vivo interaction between EWS/FLI1 and the uridine phosphorylase promoter. The identification of such direct target genes confirms the longstanding hypothesis that EWS/FLI1 can behave as an aberrant ETS transcription factor. Among the diverse range of EWS/FLI1 target genes identified, relatively few have been distinguished as direct transcriptional targets of EWS/FLI1 through chromatin immunoprecipitation. Other recently defined direct target candidate genes include ID-2 and tenascin C (25 , 26) . With the development of an efficient and reliable EWS/ETS chromatin immunoprecipitation assay, genome-wide screening strategies can be applied to identify additional EWS/FLI1 direct target genes (27) .
Both EWS/FLI1 and RAS map to the same response region, which consists primarily of a composite ETS/AP-1 site within the uridine phosphorylase promoter. There is a long precedent of both physical and functional interaction between ETS family members and AP-1 transcription factors in gene regulation (reviewed in Ref. 28 ). RAS-dependent activation of a RRE can require an interaction between the ETS protein, ternary complex factor, and AP proteins (18) . Normal FLI1 has been shown to functionally interact with AP-1 in the transcriptional regulation of MMP-1 and hemeoxygenase genes (29 , 30) . Finally, there is in vitro evidence that Fos/Jun heterduplexes can form a ternary complex with ERG, an ETS protein that has high amino acid similarity to FLI1 (31) . These data suggest that EWS/FLI1 may be interacting with AP-1 proteins to transcriptionally up-regulate uridine phosphorylase. With a direct target gene now in hand, these theories can be tested explicitly.
Using NIH-3T3 cells expressing EWS/FLI1 as a tumor model, we demonstrated that 5dFUrd at a dose of 50 mg/kg can effectively attenuate tumor formation in this system. Although 5dFUrd slowed tumor formation, it did not abrogate it altogether. Increasing the dose of 5dFUrd (up to 200 mg/kg) did not increase the degree of tumor inhibition, indicating saturation of this system (data not shown). It would suggest that at least in our NIH-3T3 system, growth inhibitory effects of 5dFUrd are directly proportional to uridine phosphorylase levels. These data indicate that EWS/FLI1 promotes oncogenesis through a number of pathways of which up-regulation of uridine phosphorylase is only one. Given that present chemotherapeutic approaches to cancer treatment involve the simultaneous use of multiple compounds, coupling 5dFUrd with other anticancer agents is a strategy worth pursuing in the context of EFTs.
Using the biochemical activities of a gene induced in a malignant cell to target it with a toxic compound is a common approach to drug design. Given that EWS/FLI1 induces dramatic changes in global gene expression (>2000 genes modulated), it is likely that the biochemical properties of other target genes could be used to therapeutically target EFT cells (11) . Moreover, given the likely overlap in target genes between EWS/FLI1 and RAS, it is possible that drugs targeted toward RAS target genes may also antagonize EWS/FLI1 mediated tumor formation.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 This work was supported by NCI Grant CA 87771 from the National Institutes of Health. 
2 To whom requests for reprints should be addressed, at 10833 Le Conte Avenue, A2-140 MDCC, Los Angeles, CA 90024. Phone: (310) 825-0704; Fax: (310) 267-2848; E-mail: cdenny{at}ucla.edu
3 The abbreviations used are: EFT, Ewings family tumor; RRE, RAS response element; RT-PCR, reverse transcription-PCR; MTT, 3-(4,5dimethylthiazol-2-yl)-2,5-diphenyltetrazolium; AP-1, activator protein-1; 5dFUrd, 5'-deoxy 5'-fluorouridine; 5-FUra, 5-fluoruracil.
Received 12/10/02. Accepted 5/ 8/03.
|
REFERENCES |
|---|
|
TopABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. Wakahara, T. Ohno, M. Kimura, T. Masuda, S. Nozawa, T. Dohjima, T. Yamamoto, A. Nagano, G. Kawai, A. Matsuhashi, et al. EWS-Fli1 Up-Regulates Expression of the Aurora A and Aurora B Kinases Mol. Cancer Res., December 1, 2008; 6(12): 1937 - 1945. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Riggi, M.-L. Suva, D. Suva, L. Cironi, P. Provero, S. Tercier, J.-M. Joseph, J.-C. Stehle, K. Baumer, V. Kindler, et al. EWS-FLI-1 Expression Triggers a Ewing's Sarcoma Initiation Program in Primary Human Mesenchymal Stem Cells Cancer Res., April 1, 2008; 68(7): 2176 - 2185. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 E. C. Torchia, K. Boyd, J. E. Rehg, C. Qu, and S. J. Baker EWS/FLI-1 Induces Rapid Onset of Myeloid/Erythroid Leukemia in Mice Mol. Cell. Biol., November 15, 2007; 27(22): 7918 - 7934. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Potikyan, R. O.V. Savene, J. M. Gaulden, K. A. France, Z. Zhou, E. S. Kleinerman, S. L. Lessnick, and C. T. Denny EWS/FLI1 Regulates Tumor Angiogenesis in Ewing's Sarcoma via Suppression of Thrombospondins Cancer Res., July 15, 2007; 67(14): 6675 - 6684. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Kim, C. T. Denny, and R. Wisdom Cooperative DNA Binding with AP-1 Proteins Is Required for Transformation by EWS-Ets Fusion Proteins. Mol. Cell. Biol., April 1, 2006; 26(7): 2467 - 2478. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Zou, H. Ichikawa, M. L. Blackburn, H.-M. Hu, A. Zielinska-Kwiatkowska, Q. Mei, G. J. Roth, H. A. Chansky, and L. Yang The Oncogenic TLS-ERG Fusion Protein Exerts Different Effects in Hematopoietic Cells and Fibroblasts Mol. Cell. Biol., July 15, 2005; 25(14): 6235 - 6246. [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 |
represents ETS core sequences (GGAA), and 
represents an AP-1 binding site. C, analysis of uridine phosphorylase promoter activation in response to EWS/FLI1 (ESF), EWS/FLI1–340 (340), and empty vector (Tk). For each promoter construct, the raw luciferase values of each cellular condition was compared with the empty vector control. Thus, the degree of activation is represented as a relative measurement compared with the empty vector control. The data are representative of at least three independent experiments done in duplicate for each promoter construct.
