Original Article

Oncogene (2008) 27, 6034–6043; doi:10.1038/onc.2008.203; published online 30 June 2008

DAX1, a direct target of EWS/FLI1 oncoprotein, is a principal regulator of cell-cycle progression in Ewing's tumor cells

E García-Aragoncillo1,5, J Carrillo1,5, E Lalli2, N Agra1, G Gómez-López3,4, Á Pestaña1 and J Alonso1

  1. 1Laboratorio de Genética y Patología Molecular de Tumores Sólidos Infantiles, Departamento de Biología Molecular y Celular del Cáncer, Instituto de Investigaciones Biomédicas ‘A Sols’, CSIC-UAM, Madrid, Spain
  2. 2Institut de Pharmacologie Moléculaire et Cellulaire, CNRS UMR6097, Universitéde Nice-Sophia Antipolis, Valbonne, France
  3. 3Fundación Biomédica del CHUVI, Hospital Dorrebullón, Vigo, Pontevedra, Spain
  4. 4Unidad de Bioinformática, Centro Nacional de Investigaciones Oncológicas, Madrid, Spain

Correspondence: Dr J Alonso, Laboratorio de Genética y Patología Molecular de Tumores Sólidos Infantiles, Departamento de Biología Molecular y Celular del Cáncer, Instituto de Investigaciones Biomédicas ‘A Sols’, CSIC-UAM, C/Arturo Duperier 4, Madrid 28029, Spain. E-mails: fjaviera@iib.uam.es

5These authors contributed equally to this work.

Received 21 January 2008; Revised 12 May 2008; Accepted 19 May 2008; Published online 30 June 2008.

Top

Abstract

The molecular hallmark of the Ewing's family of tumors is the presence of balanced chromosomal translocations, leading to the formation of chimerical transcription factors (that is, EWS/FLI1) that play a pivotal role in the pathogenesis of Ewing's tumors by deregulating gene expression. We have recently demonstrated that DAX1 (NR0B1), an orphan nuclear receptor that was not previously implicated in cancer, is induced by the EWS/FLI1 oncoprotein and is highly expressed in Ewing's tumors, suggesting that DAX1 is a biologically relevant target of EWS/FLI1-mediated oncogenesis. In this study we demonstrate that DAX1 is a direct transcriptional target of the EWS/FLI1 oncoprotein through its binding to a GGAA-rich region in the DAX1 promoter and show that DAX1 is a key player of EWS/FLI1-mediated oncogenesis. DAX1 silencing using an inducible model of RNA interference induces growth arrest in the A673 Ewing's cell line and severely impairs its capability to grow in semisolid medium and form tumors in immunodeficient mice. Gene expression profile analysis demonstrated that about 10% of the genes regulated by EWS/FLI1 in Ewing's cells are DAX1 targets, confirming the importance of DAX1 in Ewing's oncogenesis. Functional genomic analysis, validated by quantitative RT–PCR, showed that genes implicated in cell-cycle progression, such as CDK2, CDC6, MCM10 or SKP2 were similarly regulated by EWS/FLI1 and DAX1. These findings indicate that DAX1 is important in the pathogenesis of the Ewing's family of tumors, identify new functions for DAX1 as a cell-cycle progression regulator and open the possibility to new therapeutic approaches based on DAX1 function interference.

Keywords:

Ewing's tumors, EWS/FLI1 oncoprotein, DAX1, NR0B1, RNA interference

Top

Introduction

The Ewing's family of tumors (EFTs) are a group of highly malignant and primitive tumors arising in children and adolescents (Kovar, 1998). The hallmark of EFT is the presence of specific balanced chromosomal translocations leading to the formation of chimeric proteins, joining the N-terminal region from the EWS gene product in frame with the C-terminal portion of one out of five different members of the ETS family of transcription factors: FLI1, ERG or FEV (ERG subfamily) and E1AF or ETV1 (PEA3 subfamily). The EWS/FLI1 combination is the most frequent and is present in nearly 85% of all cases (de Alava and Gerald, 2000; Arvand and Denny, 2001).

A large body of evidence has shown that EWS/FLI1 and the other chimeric proteins are fundamental in Ewing's tumor oncogenesis. Because EWS/FLI1 works as an aberrant transcription factor, very much attention has been put in the last years to identify EWS/FLI1 gene targets and to analyse their contribution to Ewing's tumorigenesis (reviewed in Janknecht, 2005; Kovar, 2005). In this sense, we have recently described that DAX1 (NR0B1), an unusual orphan nuclear receptor, is induced by EWS/FLI1 and is expressed at high levels in Ewing's tumors, suggesting that DAX1 could be important in EWS/FLI1-mediated oncogenesis (Mendiola et al., 2006).

DAX1 is an important regulator of steroidogenesis, gonadal development and sex determination (Lalli and Sassone-Corsi, 2003). It is responsible for X-linked adrenal hypoplasia congenita, a disease where patients display adrenal insufficiency and electrolyte imbalance as a consequence of inactivating mutations in the DAX1 gene (Zanaria et al., 1994). In addition, chromosomal duplications in Xp21 encompasssing the DAX1 gene cause phenotypic sex reversal in XY individuals (dosage sensitive sex reversal; Bardoni et al., 1994). The function of DAX1 in Ewing's tumors is probably different from control of hormone production, as these tumors are not known to produce steroids. In this study, we aimed at characterizing the biological function of DAX1 in Ewing's tumors and their contribution to Ewing's pathogenesis. We first demonstrate that DAX1 is a direct transcriptional target of EWS/FLI1 in Ewing's tumor cells. Next, we show that DAX1 silencing, using an inducible RNA interference (RNAi) system, induces growth arrest of Ewing's tumor cells and severely impairs their growth in a xenograft mouse model. These data indicate that DAX1 is a major player of EWS/FLI1-mediated proliferation. We then performed gene expression profile studies to identify genes and pathways implicated in DAX1-dependent Ewing's cell proliferation. These experiments show that an elevated percentage of the genes regulated by EWS/FLI1 are DAX1 targets, of which a significant proportion correspond to genes implicated in G1/S transition of the cell cycle. In summary, our data demonstrate that DAX1 is a key transcriptional downstream target of EWS/FLI1 and that it is an important contributor to Ewing's cell growth by promoting cell proliferation. These findings may provide new clues to develop therapeutic strategies for the treatment of the EFTs through the modulation of DAX1 expression and/or activity.

Top

Results and discussion

DAX1 is a direct transcriptional target of the EWS/FLI1 oncoprotein

We have recently shown that DAX1 (NR0B1), an unusual orphan nuclear receptor, is induced by EWS/FLI1 and is expressed at high levels in Ewing's tumors. (Mendiola et al., 2006). However, in our previous study, we did not investigate if DAX1 is a direct target of the EWS/FLI1 transcription factor in Ewing's cells. To this purpose, we first analysed the effect of wild-type EWS/FLI1 and two EWS/FLI1 mutants, defective in DNA binding, on DAX1 induction in 293 cells. As shown in Figure 1a, DAX1 was only induced by wild-type EWS/FLI1 whereas the two EWS/FLI1 mutants did not induce DAX1 expression, although they were expressed at comparable levels. Next we analysed if EWS/FLI1 was able to induce the transcriptional activation of a 1.6kb DAX1 promoter fragment. As shown in Figure 1b, wild-type EWS/FLI1 induced a significant increment in the transcriptional activity of the DAX1 promoter, whereas both EWS/FLI1 mutants failed to activate it, consistently with the results shown above. The effect of EWS/FLI1 on DAX1 promoter transactivation was completely lost on the 1.2kb DAX1 promoter fragment (Figure 1b). Analysis of the DAX1 promoter sequence showed that the region responsible for EWS/FLI1-dependent transactivation harbors several concatenated GGAA motifs (Figure 1b). This motif is the core of the consensus DNA-binding sequence of the ETS family of transcription factors (Wasylyk et al., 1993).

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

DAX1 is a direct transcriptional target of EWS/FLI1. (a) 293 cells were transfected with the expression vectors tkneo (empty vector), wild-type EWS/FLI1, a EWS/FLI1 mutant in which the entire ets domain has been deleted (Δ65) and a EWS/FLI1 mutant in which three key amino acids of the DNA-binding domain have been substituted to impair DNA binding (triple mutant). After 72h, total protein extracts were analysed by western blot and probed sequentially with anti-FLI1 and anti-DAX1 antibodies. Blots were stripped and incubated with an anti-β-tubulin antibody as a control for loading and transferring. DAX1 expression is induced by wild-type EWS/FLI1, but not by EWS/FLI1 mutants defective in DNA-binding activity. (b) 293 cells were co-transfected with the expression vectors described above, a reporter plasmid containing either a 1.6 or a 1.2kb fragment of the DAX1 promoter and a pSV-β-galactosidase expression plasmid used as a control for transfection efficiency. The figure shows the relative CAT (1.6kb promoter) or luciferase (1.2kb promoter) activity (mean±standard error of four experiments done in triplicate) compared to cells transfected with the empty expression vector tkneo and the reporter plasmid. Wild-type EWS/FLI1 induced activation only of the 1.6kb DAX1 promoter, whereas no activation was observed with both EWS/FLI1 mutants. The sequence of the DAX1 promoter present only in the 1.6kb promoter fragment including the GGAA-rich region is shown. The position of the primers (underlined) used for the chromatin immunoprecipitation (ChIP) experiments is also shown. (c) A673/TR cells expressing high levels of the tetracycline repressor were infected with an inducible lentiviral expression vector encoding short-hairpin RNAs (shRNAs) designed against EWS/FLI1 and stable clones were selected. Four independent clones of A673/TR/shEF cells were stimulated with doxycycline to induce the expression of the corresponding shRNAs. After 72h, total protein extracts were analysed by western blot and probed sequentially with anti-FLI1, anti-DAX1 and anti-β-tubulin antibodies. Induction of EWS/FLI1-specific shRNAs efficiently reduced the levels of EWS/FLI1 and DAX1 proteins. (d) ChIP demonstrates that EWS/FLI1 interacts in vivo with the GGAA-rich region of the DAX1 promoter. A673/TR/shEF (clone 18) cells were cultured in the absence or in the presence of doxycycline during 72h, formaldehyde cross-linked, sonicated, inmunoprecipitated with an anti-FLI1 specific or an irrelevant antibody and assayed by PCR to detect the presence of a 340bp fragment corresponding to the GGAA-rich DAX1 promoter region (the core of the DNA-binding consensus sequence of the ETS family of transcription factors). A specific band was only observed when complex DNA-protein complexes were immunoprecipitated with the anti-FLI1-specific antibody from A673/TR/shEF cells in the absence of doxycycline but not from cells cultured in the presence of doxycycline, in which EWS/FLI1 expression has been silenced. (e) The number of GGAA repeats present in the polymorphic GGAA-rich region correlates with the levels of DAX1 mRNA expression. The length of the polymorphism and the DAX1 mRNA levels relative to the housekeeping gene TATA-binding protein (TBP) were measured by PCR/capillary fluorescent electrophoresis and quantitative RT–PCR, respectively, in six Ewing's cell lines. For heterozygote cell lines derived from female patients with two X chromosomes, the mean of both alleles was calculated. Points corresponding to RD-ES and SK-PN-DW Ewing's tumor cells (both homozygotes), with indication of GGAA repeat number, are shown.

Full figure and legend (125K)

We then analysed if the EWS/FLI1 transcription factor binds this GGAA-rich promoter region in vivo in A673 Ewing's cells using the chromatin immunoprecipitation (ChIP) assay. For this experiment, we took advantage of a system of inducible RNAi in Ewing's tumor cells previously established by our group (Carrillo et al., 2007). Briefly, the A673 Ewing's cell line was engineered to express high levels of the tetracycline repressor (A673/TR) and then stably infected with an EWS/FLI1-specific short-hairpin RNA (shRNA) expression vector in which shRNA expression is under the control of a doxycycline-responsive promoter (A673/TR/shEF). As shown in Figure 1c, doxycycline treatment of A673/TR/shEF cells significantly reduced expression of EWS/FLI1, concomitantly to a dramatic reduction in the levels of DAX1 protein, consistently with our previous results using transient transfection of synthetic small-interfering RNA (siRNA) against EWS/FLI1 (Mendiola et al., 2006). As EWS/FLI1 silencing was efficiently achieved with this shRNA-inducible system, we performed ChIP experiments on cells cultured in the absence (high EWS/FLI1 expression) or presence (EWS/FLI1 knockdown) of doxycycline to study the binding of EWS/FLI1 on the GGAA-rich region of the DAX1 promoter in vivo. As shown in Figure 1d, a specific PCR product using primers delimiting this region was observed from A673/TR/shEF Ewing's cells cultured in the absence of doxycycline (and thus expressing EWS/FLI1). By contrast, no PCR products were observed in the anti-FLI1 immunoprecipitate of A673/TR/shEF Ewing's cells incubated in presence of doxycycline, which as shown in Figure 1c, does not express EWS/FLI1.

The structure of this promoter region, constituted by an array of ets core motifs, suggests that synergistic activation by multiple EWS/FLI1 monomers could be important for DAX1 expression. Interestingly, the number of GGAA repeats present in the genomic DNA of different individuals is variable, which means that this region is polymorphic (Guo et al., 1996). Thus, we analysed if the number of GGAA repeats in the DAX1 promoter correlates with the levels of DAX1 mRNA expression in Ewing's cells. The GGAA-rich region was amplified by PCR in six Ewing's cell lines (A673, SK-N-MC, RD-ES, A4573, TTC-466 and SK-PN-DW) and amplicons were analysed by fluorescent capillary electrophoresis to determine their length. Fragment length ranked from 299bp in RD-ES cells to 336bp in SK-PN-DW. Sequencing of these fragments demonstrated that these sizes correspond to 17 and 26 GGAA repeats, respectively. We next determined the levels of DAX1 mRNA in these Ewing's cell lines by quantitative RT–PCR. As shown in Figure 1e, the length of the polymorphic region (and thus the number of GGAA repeats) was significantly positively correlated with the levels of DAX1 mRNA expression.

Altogether, these data indicate that EWS/FLI1 induces the expression of DAX1 through a polymorphic region present in the DAX1 promoter. This region, constituted by a concatemerization of ets core motifs, represents a new type of DNA sequence implicated in EWS/FLI1-mediated oncogenesis. In fact, the DNA motifs that are able to interact in vivo with EWS/FLI1 remain largely unknown, as DNA-binding characteristic of EWS/FLI1 and coupled gene transactivation could be different dependent of the target genes. For example, EWS/FLI1 has been shown to bind bona fide ets-binding sites in some genes (Nakatani et al., 2003), whereas cooperation with other transcription factors resulted to be necessary in others (Kim et al., 2006). In addition, several studies have demonstrated that EWS/FLI1 and FLI1 does not induce the same set of genes, despite that both share the same DNA-binding motif (Hahm et al., 1999; Nishimori et al., 2002). This suggests that promoter specificity is not only dependent of the ets-binding domain. In this sense, DAX1 itself is activated by EWS/FLI1 in 293 and HeLa cells, but not by FLI1 (Mendiola et al., 2006). Additional investigations, combining high throughput ChIP strategies with global gene expression microarrays will be neccesary to determine in detail what DNA motifs or combination of them are involved in EWS/FLI1-mediated oncogenesis. In this sense, identification of GGAA-rich regions in promoters, linked to gene expression profiling, may represent an attractive method to discover direct target genes of the EWS/FLI1 oncoprotein.

DAX1 silencing blocks Ewing's cell proliferation

To analyse the contribution of DAX1 to Ewing's tumorigenesis, we developed an Ewing's cell line expressing an inducible DAX1-specific shRNA (A673/TR/shDAX). As shown in Figure 2a, induction of the DAX1-specific shRNA by doxycycline efficiently reduced the levels of DAX1 mRNA (not shown) and protein, whereas it had no effect on EWS/FLI1 expression. Thus, our system results ideal to study the specific effects of DAX1 on Ewing's oncogenesis without affecting EWS/FLI1 expression. As expected, control cell lines expressing an inducible shRNA designed against the green fluorescent protein (GFP) had no effect on either EWS/FLI1 or DAX1 expression.

Figure 2.
Figure 2 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

DAX1 silencing induces growth arrest in Ewing's tumor cells. A673/TR cells expressing high levels of the tetracycline repressor were infected with an inducible lentiviral expression vector encoding short-hairpin RNAs (shRNAs) designed against DAX1 and stable clones were selected. (a) Two independent clones of A673/TR/shDAX cells and control A673/TR/shGFP cells were stimulated with doxycycline to induce the expression of the corresponding shRNAs. After 72h, total protein extracts were analysed as described above. Induction of DAX1-specific shRNAs efficiently reduced the levels of the DAX1 protein, but had no effect on the levels of the EWS/FLI1 protein. (b) DAX1 knockdown significantly reduced DNA synthesis in A673 Ewing's tumor cells. Control cells (A673/TR/shGFP), one clone of A673/TR/shEF cells and two independent clones of A673/TR/shDAX1 were stimulated with doxycycline for 72h and DNA synthesis quantified by 5-bromo-2-deoxyuridine (BrdU) incorporation into DNA. The figure shows the mean±standard error of three independent experiments done in triplicate. Data are shown as the percentage of BrdU incorporation versus unstimulated cells, which was arbitrarily set to 100. (*P<0.01 versus control cells stimulated with doxycycline). (c) A673/TR/shEF and A673/TR/shDAX1 cells were incubated for 72h in the absence or in the presence of doxycycline to induce the expression of the corresponding shRNAs. Cell-cycle distribution was analysed by flow cytometry. Both EWS/FLI1 and DAX1 silencing induce cell growth arrest with concomitant accumulation of cells in the G1 phase of cell cycle. The figure shows the mean of two independent experiments.

Full figure and legend (69K)

Several studies have shown that inhibition of EWS/FLI1 expression in Ewing's cells using for example constitutive expression of antisense cDNA (Tanaka et al., 1997) or transient transfection of specific siRNAs (Prieur et al., 2004), produces growth arrest of Ewing's tumor cells. We thus studied the effects of EWS/FLI1 and DAX1 knockdown on Ewing's tumor cell proliferation and/or cell death using our systems of conditional siRNA expression. As expected, induction of the EWS/FLI1-specific shRNA with doxycycline in A673/TR/shEF cells reduced 5-bromo-2-deoxyuridine (BrdU) incorporation of about 50% (Figure 2b). DAX1 knockdown in A673/TR/shDAX cells upon doxycycline stimulation also produced a reduction in the rate of BrdU incorporation that was, interestingly, in the same range as the one observed upon EWS/FLI1 knockdown. Note that DAX1 silencing on these cells does not affect the levels of the EWS/FLI1 protein (Figure 2a), indicating that EWS/FLI1-dependent cell proliferation is mediated to a great extent by DAX1. Flow cytometry demonstrated that EWS/FLI1 knockdown produced a significant accumulation of cells in the G1 phase of the cell cycle and a concomitant reduction in the percentage of cells in the S and G2/M phases (Figure 2c). Again the effect of DAX1 knockdown on cell cycle was comparable to that observed for EWS/FLI1, indicating that DAX1 silencing also induced growth arrest in Ewing's tumor cells (Figure 2c). Apoptosis was not observed upon DAX1 silencing (data not shown). In agreement with the role of DAX1 in Ewing's cell proliferation, DAX1 knockdown reduced significantly the size of colonies growing in soft agar, although the number of colonies per plate was similar to that of control cells (Figure 3a). This result suggests that, as shown above, DAX1 expression is required for Ewing's tumor cell proliferation but is probably not related to cell transformation itself. Lastly, DAX1 silencing also inhibited very efficiently the formation of tumors in nude mice (Figure 3b). Indeed, the effect of DAX1 knockdown in vivo was more pronounced than that observed on colony size. Thus, the effect of DAX1 on cell proliferation could be even more potent in vivo, where interaction with the microenvironment and complex processes such as angiogenesis are necessary in tumor development.

Figure 3.
Figure 3 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

DAX1 silencing blocks Ewing's tumor growth in vivo. (a) A673/TR/shGFP (control cells), A673/TR/shEF (clone 18) and A673/TR/shDAX (clone 13) cells were grown in soft agar in the absence or in the presence of doxycycline for 25 days. Culture dishes were photographed and colony number and mean colony area were calculated. EWS/FLI1 silencing reduced dramatically the number and the size of the colonies growing in semisolid medium. DAX1 knockdown reduced significantly the size of the colonies, but had no effect on the total number of colonies (mean±standar error of a representative experiment done in triplicate; *P<0.0001, **P<0.005). (b) Nude mice (n=14–16 animals) were injected with A673/TR/shEF (clone 18) or A673/TR/shDAX (clone 11 and 13) cells, and split in two groups, one of which was given doxycycline in the drinking water to induce the expression of the corresponding short-hairpin RNA (shRNA). DAX1 knockdown inhibited tumor growth dramatically. As expected, EWS/FLI1 silencing blocked completely the formation of tumors.

Full figure and legend (46K)

The findings described above demonstrate that DAX1 is an important mediator of EWS/FLI1-dependent cell proliferation in Ewing's tumor cells. Recently, it has been shown that DAX1 silencing impaired Ewing's cell growth (Kinsey et al., 2006). However, it was not analysed in detail if these effects were because of reduced cell proliferation or increased cell death (apoptosis). Our results demonstrate, using an inducible model of RNAi in Ewing's cells, that DAX1 knockdown is associated to a significant reduction of DNA synthesis and accumulation of cells in the G1 phase of cell cycle whereas no effect on apoptosis was observed. Interestingly, the effect of DAX1 knockdown on Ewing's cell proliferation was quantitatively comparable to those observed upon EWS/FLI1 silencing. As DAX1 silencing blocked cell proliferation in presence of continued EWS/FLI1 expression, the stimulation of cell proliferation by EWS/FLI1 appears to be mainly mediated by DAX1.

EWS/FLI1 and DAX1 control expression of a significant overlapping set of genes in Ewing's tumor cells

How does DAX1 mediate cell proliferation in Ewing's tumor cells? In their physiological setting, DAX1 works as a negative transcriptional regulator of the biosynthesis of steroids (Lalli and Sassone-Corsi, 2003). However, the function of DAX1 in Ewing's tumors is probably different from control of hormone production, as these tumors are not known to produce steroids. Nevertheless, based on its known properties, DAX1 is supposed to also work as a transcriptional regulator in Ewing's tumor cells.

We thus analysed the gene expression profile induced by EWS/FLI1 and DAX1 silencing to identify genes and functional pathways regulated by both EWS/FLI1 and DAX1 that could account for the effect of DAX1 on Ewing's cell proliferation. To overcome the intrinsic technical and sampling variability, we performed microarray analysis of three independent clones and a polyclonal population for A673/TR/shEF, A673/TR/shDAX and A673/TR/shGFP cells stimulated with doxycycline during 72h. After filtering and normalizing, statistical analysis (P<0.001, Student's t-test) revealed that 3494 and 1366 gene probes were dysregulated by EWS/FLI1 and DAX1 knockdowns, respectively, when compared to A673/TR/shGFP control cells. These probes correspond to 2783 and 1081 annotated genes, respectively (Figure 4a). Notably, the number of genes regulated by DAX1 was significantly smaller than the number of genes regulated by EWS/FLI1, which is expected for an EWS/FLI1 downstream target as DAX1. As DAX1 is a direct downstream target of EWS/FLI1 and given that it mediates some of the effects that EWS/FLI1 exerts on Ewing's tumor cells, one should expect that a significant proportion of the genes regulated by EWS/FLI1 should also be regulated by DAX1. In fact, we found that 266 gene probes, corresponding to 229 annotated genes were consistently upregulated or downregulated by both EWS/FLI1 and DAX1 (Figure 4a, Supplementary Table S1). These results indicate that approximately 10% of genes regulated by EWS/FLI1 are DAX1 targets. As shown by χ2-statistical analysis, the number of genes that are regulated by both EWS/FLI1 and DAX1 is much higher than the number expected by chance alone (P<0.0001, Figure 4a).

Figure 4.
Figure 4 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

EWS/FLI1 and DAX1 regulate a subgroup of genes implicated in cell-cycle progression. (a) χ2-Analysis shows that the number of genes regulated by both EWS/FLI1 and DAX1 is much higher than the one expected by chance alone (P<0.0001). (b) Expression Analysis Systematic Explorer (EASE) analysis shows that genes functionally related to cell cycle are overrepresented in the set of genes regulated by EWS/FLI1 and DAX1 (P-value according to Fisher's exact test is shown). (c) Gene set enrichment analysis (GSEA) analysis identified that RB and CELLCYCLE pathways are similarly regulated by EWS/FLI1 and DAX1. FDR q-values, nominal P-values and enrichment score (ES) reflecting the degree to which the pathway is overrepresented at the top of the ranked list of genes, are shown (see ‘Materials and methods’ for details).

Full figure and legend (176K)

In an attempt to identify functions that could explain the relation between EWS/FLI1 and DAX1, we used the Expression Analysis Systematic Explorer (EASE) algorithm (Hosack et al., 2003) to look for gene ontology (GO) terms (Gene Ontology Consortium database; Ashburner et al., 2000) overrepresented in the set of genes regulated by EWS/FLI1 or DAX1, as well as in the list of genes consistently regulated by both proteins. Significantly overrepresented GO terms in each subset of genes (P<0.0001, Fisher's exact test) are detailed in Supplementary Table S2a–c. Noteworthy, the only GO term overrepresented in all subsets of genes was ‘cell cycle’, indicating that genes classified according to this term were significantly enriched in the set of genes commonly regulated by EWS/FLI1 and DAX1 (Figure 4b). However, it should be taken into account that overrepresentation of GO terms simply displays information about annotated functions of individual genes, rather than defined pathways. Thus, to identify pathways similarly affected by EWS/FLI1 and DAX1, we performed a gene set enrichment analysis (GSEA; Subramanian et al., 2005). Briefly, GSEA determines whether one list of genes (in our case genes belonging to a defined pathway) is enriched near the top of a second, rank ordered gene list (in our case the list of genes regulated by EWS/FLI1 or DAX1). We tested 215 depurated pathways of Biocarta (www.biocarta.com) against our ranked data set according to Student's t-test P-values. If the majority of the genes belonging to a pathway are localized at the top of the ranked list of genes, it implies that this pathway is differentially regulated by EWS/FLI1 or DAX1. GSEA analysis identified several pathways regulated by EWS/FLI1 and DAX1 (false discovery rate (FDR) q-value <0.25, Figure 4c and Supplementary Table S3a–b). Two pathways, RBPATHWAY and CELLCYCLEPATHWAY according to Biocarta denomination, implicated in the G1/S transition of the cell cycle, were regulated by both EWS/FLI1 and DAX1. Thus, functional genomic analysis of microarray data showed that the only biological functions and pathways overrepresented in both EWS/FLI1 and DAX1 sets of regulated genes correspond to genes implicated in cell-cycle progression, in particular, in the transition from G1 to S phase. These findings provide a functional explanation for the effect of DAX1 silencing on Ewing's cell proliferation.

We chose four genes based in their key contribution to cell-cycle regulation, but with different functions, to independently confirm these results by quantitative RT–PCR: the cyclin-dependent kinase (CDK) 2, essential for cell-cycle G1/S-phase transition; the F-box protein SKP2, involved in the ubiquitin-dependent degradation of members of the KIP family of CDK inhibitors; the mini-chromosome maintenance protein 10 (MCM10), involved in the initiation of eukaryotic genome replication and CDC6, a protein involved in the initiation of DNA replication that controls that DNA replication is completed before mitosis is initiated. As shown in Figure 5, CDK2, SKP2, MCM10 and CDC6 were downregulated upon EWS/FLI1 or DAX1 knockdown and thus were upregulated by both EWS/FLI1 and DAX1 in the A673 Ewing's tumor cell. We confirmed these results on a different Ewing's cell line (SK-PN-DW), also engineered to express EWS/FLI1 and DAX1-specific shRNAs upon doxycycline stimulation. Expression profiles of these cell-cycle genes were highly similar in both Ewing's cell lines upon both EWS/FLI1 and DAX1 knockdown (Figure 5). As EWS/FLI1 levels were not altered upon DAX1 silencing (Figures 2a and 5), we conclude that the effect of EWS/FLI1 on the expression of these cell-cycle regulators in Ewing's cells is mainly mediated through DAX1.

Figure 5.
Figure 5 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Cell-cycle genes are similarly regulated by EWS/FLI1 and DAX1 in Ewing's cells. A673/TR/shEF, A673/TR/shDAX, SK-PN-DW/TR/shEF and SK-PN-DW/TR/shDAX were cultured in the absence or in the presence of doxycycline to induce the expression of the corresponding short-hairpin RNA (shRNA). mRNA expression levels of the cell-cycle genes cyclin-dependent kinase (CDK) 2, SKP2, mini-chromosome maintenance protein 10 (MCM10) and CDC6 was quantified by real-time RT–PCR at 2 and 4 days after doxycycline stimulation. Levels of EWS/FLI1 and DAX1 mRNA were also analysed to confirm the efficiency of gene silencing. mRNA expression was normalized to cells cultured in absence of doxycycline. The expression profile of cell-cycle genes was similarly regulated by EWS/FLI1 and DAX1 in both Ewing's cell lines.

Full figure and legend (138K)

Finally, we analysed if DAX1 was able to induce these cell-cycle genes in other cell types. 293 (human embryonic kidney) and IMR32 (neuroblastoma) cells, which do not express DAX1 mRNA, were transfected transiently (293) or stably (IMR32) with an expression plasmid harboring the complete DAX1 cDNA. As shown in the Supplementary Figure S1, only MCM10 was significantly upregulated by DAX1 overexpression in the neuroblastoma cell line IMR32. These results suggest that the regulation of cell-cycle genes by DAX1 depends on the cell background, probably because different cofactors are present in different cellular contexts or because of effects of chromatin structure.

DAX1 has been shown to interact with several nuclear receptors and to act as a negative transcriptional regulator in the context of steroid hormone synthesis. In an attempt to identify putative DAX1 partners in Ewing's cells, we screened a public microarray expression profile dataset of Ewing's tumors (GEO ID: GDS1713; Staege et al., 2004), as well as a Ewing's tumor cells microarray dataset collected in our laboratory (data not shown). This screening shows that most established DAX1 partners and downstream targets such as SF1 (NR5A1; Ito et al., 1997), Nur77 (NR4A1; Song et al., 2004), LRH1 (NR5A2; Suzuki et al., 2002), androgen receptor (Holter et al., 2002) or estrogen receptor (ESR1; Zhang et al., 2000) are not expressed in Ewing's cells and tumors. Thus, it is most unlikely that DAX1 exerts their action through interaction with these nuclear receptors and consequently the physiological partners of DAX1 in Ewing's cells remain to be identified. DAX1 has also been show to interact with several corepressors such as N-CoR or Alien (Crawford et al., 1998; Altincicek et al., 2000). These corepressors are expressed nearly ubiquitously in all tissues and thus their expression in Ewing's cells is not unexpected. Nevertheless, our data demonstrate that DAX1 induces several cell-cycle genes in Ewing's tumor cells. A simple explanation for this apparent discrepancy is that a transcriptional repressor regulating the expression of cell-cycle genes was repressed by DAX1, which subsequently would induce the expression of those genes. Alternatively, DAX1 could act as a transcriptional activator in the context of Ewing's tumor cells. It is at the moment largely unknown which of the genes regulated by DAX1 in Ewing's cells are direct targets and if these are repressed or activated in the Ewing's cell background. ChIP-on-chip and yeast two hybrid technologies would be useful to determine, respectively, what genes are truly DAX1 direct targets and the physiological partners of DAX1 in Ewing's tumor cells, to establish the cascade of events that ultimately lead to EWS/FLI1 and DAX1-mediated cell proliferation.

The experimental data published in the last 2 years support an importance of DAX1 in the pathogenesis of Ewing's tumors (Kinsey et al., 2006; Mendiola et al., 2006 and this study): DAX1 is highly expressed in EFTs, is a direct target of the EWS/FLI1 oncoprotein, its expression is necessary for EWS/FLI1-mediated proliferation, transformation and tumorigenesis and regulates the expression of key genes implicated in cell-cycle progression. These findings attribute new and unexpected functions to DAX1 in cell-cycle regulation, provide new clues about Ewing's oncogenesis and make DAX1 an attractive protein for targeted therapeutics. In this sense characterization of DAX1 domains implicated in DAX1 function should be helpful to design small molecules able to interact with DAX1 and to block their function in Ewing's tumor cells.

Top

Materials and methods

Expression and reporter plasmids

The expression vectors pSRα EWS/FLI1 and the EWS/FLI1 mutants pSRα EWS/FLI1 Δ65 and pSRα EWS/FLI1 triple mutant, which are deficient in DNA binding were a generous gift of Dr CT Denny (UCLA, LA, USA) and are described in detail elsewhere (Welford et al., 2001). The expression vector pSG5-DAX1 has been previously described (Zanaria et al., 1994). p1.6kb DAX1 PROM harbors 1.6kb (−1595/+46) of the DAX1 promoter sequence cloned into the pBLCAT3 vector (Luckow and Schutz, 1987). p1.2kb DAX1 PROM harbors 1.2kb (−1241/+46) of the DAX1 sequence cloned into the pGL2basic vector (Promega, Madison, WI, USA).

Transient and stable transfection experiments and luciferase reporter assays

293 cells (350000 cells per well into six-well dishes) were transfected with 1.2μg of expression vector (pSRα EWS/FLI1, pSRα EWS/FLI1 Δ65 or pSRα EWS/FLI1 triple mutant) using Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA, USA). After 48h, cells were lysed and expression of DAX1 and EWS/FLI1 proteins analysed by western blot with antibodies anti-DAX1 (clone 2F4; Tamai et al., 1996) and anti-FLI1 (C-19, Santa Cruz Biotechnology, Santa Cruz, CA, USA), as previously described (Mendiola et al., 2006).

293 and IMR32 neuroblastoma cells were transfected with the expression vector pSG5-DAX1 and empty vector as a control. IMR32 cells were incubated in presence of G418 to obtain stable clones. Total RNA was extracted and analysed by real-time quantitative RT–PCR as previously described (Ferreira et al., 2008).

For luciferase reporter assays, 293 cells were transfected with EWS/FLI1 expression vectors as described above and reporter plasmids (0.4μg). pSV-β-galactosidase (Promega; 20ng) vector was used as a control for transfection efficiency. At 48h after transfection, cells were lysed and assayed for CAT (CAT ELISA kit, Roche Applied Sciences, Basel, Switzerland) and luciferase activity (Promega) using a Lumat LB 9507 luminometer (Berthold Technologies, Bad Wildbad, Germany). CAT and luciferase activity were normalized by β-galactosidase expression.

Establishment of Ewing's cell lines expressing doxycycline-inducible small-hairpin RNA

Establishment of Ewing's cell lines harboring doxycycline-inducible shRNAs against EWS/FLI1 is described in detail elsewhere (Carrillo et al., 2007). The target shRNA sequences for DAX1 (shDAX1) were chosen using the BLOCK-iT RNAi Designer web application (Invitrogen) and correspond to nucleotides 1320–1340 of the DAX1 mRNA (GenBank accession number NM_000475). One shRNA against the GFP was used as a negative control. Oligonucleotide sequences were: shDAX1 forward, 5′-CACCGAGATTCATCAATGCCAATGTCGAAACATTGGCATTGATGAATCTC-3′; shDAX1 reverse, 5′-AAAAGAGATTCATCAATGCCAATGTTTCGACATTGGCATTGATGAATCTC-3′; shGFP forward, 5′-CACCGCATGGACGAGCTGTACAAGTCGAAACTTGTACAGCTCGTCCATGC-3′; shGFP reverse 5′-AAAAGCATGGACGAGCTGTACAAGTTTCGACTTGTACAGCTCGTCCATGC-3′. Oligonucleotides were annealed and inserted into the pENTR-BLOCK-iT plasmid and transferred by recombination to the pLenti4-BLOCK-iT plasmid. Then, A673/TR and SK-PN-DW/TR Ewing's tumor cells, expressing high levels of the tetracycline repressor, were infected with lentiviruses containing the pLenti4-shRNAs and selected with zeocin (100μg/ml). Upon doxycycline (1μg/ml) stimulation, mRNA and protein knockdown were assayed by quantitative RT–PCR and by western blot (Mendiola et al., 2006). Clones displaying the highest levels of mRNA and protein knockdown were chosen for additional studies.

Chromatin immunoprecipitation assay

A total of 5 × 107 A673/TR/shEF cells, incubated in absence or presence of doxycycline (1μg/ml) for 72h to knockdown EWS/FLI1 expression, were used for each experimental condition. After fixation with formaldehyde and sonication, DNA-protein complex were immunoprecipitated with 15μg of anti-FLI1 antibody C-19 or nonspecific rabbit immunoglobulin G Santa Cruz Biotechnology). DNA-protein complexes were then reverse cross-linked and PCR-amplified using DAX-PF1 (5′-CCTCTCACAGGCAGAATGAAAT-3′) and DAX-PR3 (5′-TATACCAGCTGATACA GAATCATT-3′) primers.

Cell proliferation assays and cell-cycle analysis

DNA synthesis was determined by BrdU incorporation into DNA (Cell proliferation ELISA, Roche Applied Science) as previously described (Carrillo et al., 2007). For cell-cycle analysis, 500000 cells per 100mm dishes were incubated in absence or presence of doxycycline (1μg/ml) for 72h. Subsequently, both floating and adherent cells were collected, fixed with cold 70% ethanol, incubated with RNase (100μg/ml) and propidium iodide (40μg/ml) for 30min at 37°C and analysed by flow cytometry (FACSCAN, BD Biosciences, San Jose, CA, USA). Results were analysed with WinMDI v2.8 software (The Scripps Research Institute, La Jolla, CA, USA).

Colony formation and tumor formation assay in nude mice

Cells were plated by triplicate (5 × 104 cells per 60mm dishes) in soft agar and cultured in presence or absence of doxycycline during 25 days. Fresh culture medium was added to plates each 2 days. At the end of the experiment, three random fields for each plate were photographed. The number of colonies for field and its area were calculated using the image analysis software analysis (Olympus, Hamburg, Germany). Tumor formation assay in nude mice was performed as previously described (Carrillo et al., 2007)

CodeLink microarrays for analysis of gene expression profiles

Microarray CodeLink Human Whole Genome (Applied Microarrays Inc., Tempe, AZ, USA), containing approximately 55000 gene probes derived from annotated mRNA sequences and expressed sequence tags, was used to analyse gene expression profiles. RNA isolation, labeling, hybridization and scanning were performed at the Genomic core facility of the Instituto de Investigaciones Biomédicas (Madrid, Spain) as previously described (Carrillo et al., 2007; Ferreira et al., 2008). Data were analysed using CodeLink Expression Analysis Software, median-normalized, log2 transformed and exported to the MeV software (TIGR, Rockville, MD, USA; Saeed et al., 2003) for additional statistical analysis and graphical visualization. To identify genes regulated by EWS/FLI1 and DAX1 in Ewing's tumor cells, we analysed the RNA isolated from three separate clones and a polyclonal population stimulated with doxycycline for 72h to induce the expression of GFP (control), EWS/FLI1 and DAX1-specific siRNAs. Genes that reach statistical significance (P<0.001, Student's t-test) in relation to control cells were considered regulated by EWS/FLI1 or DAX1.

Functional genomic analysis of gene expression profiles

Gene Ontology annotations (cellular component, molecular function, biological process) were obtained for differentially expressed genes using the EASE (http://david.abcc.ncifcrf.gov/ease/ease.jsp) implemented in the MeV software. Statistical differences between the percentages of GO terms were calculated by Fischer's exact test.

Gene Set Enrichment Analyses (http://www.broad.mit.edu/gsea/; (Subramanian et al., 2005) were carried out using publicly available Biocarta pathways (www.biocarta.com). GSEA is a computational method that determines whether an a priori defined set of genes (in our case Biocarta pathways) is overrepresented at the top or bottom of a ranked list of genes (in our case the list of genes differentially regulated by EWS/FLI1 or DAX1 ranking using the Student's t-test metric). Statistical significance (P-values) was determined using phenotype testing (1000 permutations). When a whole database of gene sets (that is, Biocarta) is evaluated, we adjust the estimated significance level (P-value) to account for multiple hypothesis testing. This adjustment was carried out using FDR (Reiner et al., 2003) that estimates the probability that a gene set with a given normalized enrichment statistic (NES) represents a false-positive finding. NES is the primary statistic for examining gene set enrichment results and is obtained by normalizing the enrichment score (which reflects the degree to which a gene set is overrepresented at the top or bottom of a ranked list of genes), to compare results across gene sets. In this study, we got conclusions from those gene sets (pathways) reaching a FDR less than 25% (FDR<0.25) among classes (EWS/FLI1 or DAX1 versus GFP), which is a well-established cutoff for the identification of biologically relevant gene sets (Subramanian et al., 2005).

Top

References

  1. Altincicek B, Tenbaum SP, Dressel U, Thormeyer D, Renkawitz R, Baniahmad A. (2000). Interaction of the corepressor Alien with DAX-1 is abrogated by mutations of DAX-1 involved in adrenal hypoplasia congenita. J Biol Chem 275: 7662–7667. | Article | PubMed | ChemPort |
  2. Arvand A, Denny CT. (2001). Biology of EWS/ETS fusions in Ewing's family tumors. Oncogene 20: 5747–5754. | Article | PubMed | ISI | ChemPort |
  3. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM et al. (2000). Gene ontology: tool for the unification of biology. The gene ontology consortium. Nat Genet 25: 25–29. | Article | PubMed | ISI | ChemPort |
  4. Bardoni B, Zanaria E, Guioli S, Floridia G, Worley KC, Tonini G et al. (1994). A dosage sensitive locus at chromosome Xp21 is involved in male to female sex reversal. Nat Genet 7: 497–501. | Article | PubMed | ISI | ChemPort |
  5. Carrillo J, Garcia-Aragoncillo E, Azorin D, Agra N, Sastre A, Gonzalez-Mediero I et al. (2007). Cholecystokinin down-regulation by RNA interference impairs Ewing tumor growth. Clin Cancer Res 13: 2429–2440. | Article | PubMed | ChemPort |
  6. Crawford PA, Dorn C, Sadovsky Y, Milbrandt J. (1998). Nuclear receptor DAX-1 recruits nuclear receptor corepressor N-CoR to steroidogenic factor 1. Mol Cell Biol 18: 2949–2956. | PubMed | ISI | ChemPort |
  7. De Alava E, Gerald WL. (2000). Molecular biology of the Ewing's sarcoma/primitive neuroectodermal tumor family. J Clin Oncol 18: 204–213. | PubMed | ISI | ChemPort |
  8. Ferreira BI, Alonso J, Carrillo J, Acquadro F, Largo C, Suela J et al. (2008). Array CGH and gene-expression profiling reveals distinct genomic instability patterns associated with DNA repair and cell-cycle checkpoint pathways in Ewing's sarcoma. Oncogene 27: 2084–2090. | Article | PubMed | ChemPort |
  9. Guo W, Burris TP, Zhang YH, Huang BL, Mason J, Copeland KC et al. (1996). Genomic sequence of the DAX1 gene: an orphan nuclear receptor responsible for X-linked adrenal hypoplasia congenita and hypogonadotropic hypogonadism. J Clin Endocrinol Metab 81: 2481–2486. | Article | PubMed | ChemPort |
  10. Hahm KB, Cho K, Lee C, Im YH, Chang J, Choi SG et al. (1999). Repression of the gene encoding the TGF-beta type II receptor is a major target of the EWS-FLI1 oncoprotein. Nat Genet 23: 222–227. | Article | PubMed | ISI | ChemPort |
  11. Holter E, Kotaja N, Makela S, Strauss L, Kietz S, Janne OA et al. (2002). Inhibition of androgen receptor (AR) function by the reproductive orphan nuclear receptor DAX-1. Mol Endocrinol 16: 515–528. | Article | PubMed | ISI | ChemPort |
  12. Hosack DA, Dennis Jr G, Sherman BT, Lane HC, Lempicki RA. (2003). Identifying biological themes within lists of genes with EASE. Genome Biol 4: R70. | Article | PubMed |
  13. Ito M, Yu R, Jameson JL. (1997). DAX-1 inhibits SF-1-mediated transactivation via a carboxy-terminal domain that is deleted in adrenal hypoplasia congenita. Mol Cell Biol 17: 1476–1483. | PubMed | ISI | ChemPort |
  14. Janknecht R. (2005). EWS-ETS oncoproteins: the linchpins of Ewing tumors. Gene 363: 1–14. | Article | PubMed | ChemPort |
  15. Kim S, Denny CT, Wisdom R. (2006). Cooperative DNA binding with AP-1 proteins is required for transformation by EWS-Ets fusion proteins. Mol Cell Biol 26: 2467–2478. | Article | PubMed | ChemPort |
  16. Kinsey M, Smith R, Lessnick SL. (2006). NR0B1 is required for the oncogenic phenotype mediated by EWS/FLI in Ewing's sarcoma. Mol Cancer Res 4: 851–859. | Article | PubMed | ChemPort |
  17. Kovar H. (1998). Ewing's sarcoma and peripheral primitive neuroectodermal tumors after their genetic union. Curr Opin Oncol 10: 334–342. | Article | PubMed | ChemPort |
  18. Kovar H. (2005). Context matters: the hen or egg problem in Ewing's sarcoma. Semin Cancer Biol 15: 189–196. | Article | PubMed | ChemPort |
  19. Lalli E, Sassone-Corsi P. (2003). DAX-1, an unusual orphan receptor at the crossroads of steroidogenic function and sexual differentiation. Mol Endocrinol 17: 1445–1453. | Article | PubMed | ISI | ChemPort |
  20. Luckow B, Schutz G. (1987). CAT constructions with multiple unique restriction sites for the functional analysis of eukaryotic promoters and regulatory elements. Nucleic Acids Res 15: 5490. | Article | PubMed | ISI | ChemPort |
  21. Mendiola M, Carrillo J, Garcia E, Lalli E, Hernandez T, De Alava E et al. (2006). The orphan nuclear receptor DAX1 is up-regulated by the EWS/FLI1 oncoprotein and is highly expressed in Ewing tumors. Int J Cancer 118: 1381–1389. | Article | PubMed | ISI | ChemPort |
  22. Nakatani F, Tanaka K, Sakimura R, Matsumoto Y, Matsunobu T, Li X et al. (2003). Identification of p21WAF1/CIP1 as a direct target of EWS-Fli1 oncogenic fusion protein. J Biol Chem 278: 15105–15115. | Article | PubMed | ChemPort |
  23. Nishimori H, Sasaki Y, Yoshida K, Irifune H, Zembutsu H, Tanaka T et al. (2002). The Id2 gene is a novel target of transcriptional activation by EWS-ETS fusion proteins in Ewing family tumors. Oncogene 21: 8302–8309. | Article | PubMed | ISI | ChemPort |
  24. Prieur A, Tirode F, Cohen P, Delattre O. (2004). EWS/FLI-1 silencing and gene profiling of Ewing cells reveal downstream oncogenic pathways and a crucial role for repression of insulin-like growth factor binding protein 3. Mol Cell Biol 24: 7275–7283. | Article | PubMed | ISI | ChemPort |
  25. Reiner A, Yekutieli D, Benjamini Y. (2003). Identifying differentially expressed genes using false discovery rate controlling procedures. Bioinformatics 19: 368–375. | Article | PubMed | ISI | ChemPort |
  26. Saeed AI, Sharov V, White J, Li J, Liang W, Bhagabati N et al. (2003). TM4: a free, open-source system for microarray data management and analysis. Biotechniques 34: 374–378. | PubMed | ISI | ChemPort |
  27. Song KH, Park YY, Park KC, Hong CY, Park JH, Shong M et al. (2004). The atypical orphan nuclear receptor DAX-1 interacts with orphan nuclear receptor Nur77 and represses its transactivation. Mol Endocrinol 18: 1929–1940. | Article | PubMed | ChemPort |
  28. Staege MS, Hutter C, Neumann I, Foja S, Hattenhorst UE, Hansen G et al. (2004). DNA microarrays reveal relationship of Ewing family tumors to both endothelial and fetal neural crest-derived cells and define novel targets. Cancer Res 64: 8213–8221. | Article | PubMed | ChemPort |
  29. Subramanian A, Tamayo P, Mootha VK, Mukherjee S, Ebert BL, Gillette MA et al. (2005). Gene set enrichment analysis: a knowledge-based approach for interpreting genome-wide expression profiles. Proc Natl Acad Sci USA 102: 15545–15550. | Article | PubMed | ChemPort |
  30. Suzuki T, Kasahara M, Yoshioka H, Umesono K, Morohashi K. (2002). LXXLL motifs in Dax-1 have target specificity for the orphan nuclear receptors Ad4BP/SF-1 and LRH-1. Endocr Res 28: 537. | Article | PubMed |
  31. Tamai KT, Monaco L, Alastalo TP, Lalli E, Parvinen M, Sassone-Corsi P. (1996). Hormonal and developmental regulation of DAX-1 expression in Sertoli cells. Mol Endocrinol 10: 1561–1569. | Article | PubMed | ISI | ChemPort |
  32. Tanaka K, Iwakuma T, Harimaya K, Sato H, Iwamoto Y. (1997). EWS-Fli1 antisense oligodeoxynucleotide inhibits proliferation of human Ewing's sarcoma and primitive neuroectodermal tumor cells. J Clin Invest 99: 239–247. | Article | PubMed | ISI | ChemPort |
  33. Wasylyk B, Hahn SL, Giovane A. (1993). The Ets family of transcription factors. Eur J Biochem 211: 7–18. | Article | PubMed | ISI | ChemPort |
  34. Welford SM, Hebert SP, Deneen B, Arvand A, Denny CT. (2001). DNA binding domain-independent pathways are involved in EWS/FLI1-mediated oncogenesis. J Biol Chem 276: 41977–41984. | Article | PubMed | ISI | ChemPort |
  35. Zanaria E, Muscatelli F, Bardoni B, Strom TM, Guioli S, Guo W et al. (1994). An unusual member of the nuclear hormone receptor superfamily responsible for X-linked adrenal hypoplasia congenita. Nature 372: 635–641. | Article | PubMed | ISI | ChemPort |
  36. Zhang H, Thomsen JS, Johansson L, Gustafsson JA, Treuter E. (2000). DAX-1 functions as an LXXLL-containing corepressor for activated estrogen receptors. J Biol Chem 275: 39855–39859. | Article | PubMed | ISI | ChemPort |
Top

Acknowledgements

We are grateful to B Bardoni for gift of the pDAX-1 PROM.CAT plasmid and CT Denny for gift of the pSRα EWS/FLI1 and the EWS/FLI1 mutants pSRα EWS/FLI1 Δ65 and pSRα EWS/FLI1 triple mutant used in this study. This study was funded by Ministerio de Educación y Ciencia grants SAF2006-07586 and SAF2007-62101; Fundación Inocente Inocente, Fundación Enriqueta Villavecchia and Fundación Científica Asociación Española Contra el Cáncer. E Lalli is supported by Fondation pour la Recherche Médicale, Association Recherche sur le Cancer and a CNRS-ATIP grant; E García-Aragoncillo and J Carrillo are predoctoral fellows from the Ministerio de Educación y Ciencia. N Agra is a predoctoral fellow from the Fundación General de la U.A.M; G Gómez-López is supported by a contract from de Fondo de Investigaciones Sanitarias; J Alonso has been supported by a Ramón y Cajal contract from the Ministerio de Educación y Ciencia.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)