O-GlcNAcylation is involved in the transcriptional activity of ... - Nature

Oncogene (2009) 28, 1280–1284

& 2009 Macmillan Publishers Limited All rights reserved 0950-9232/09 $32.00 www.nature.com/onc

SHORT COMMUNICATION

O-GlcNAcylation is involved in the transcriptional activity of EWS-FLI1 in Ewing’s sarcoma R Bachmaier1, DNT Aryee1, G Jug1, M Kauer1, M Kreppel1, KA Lee2 and H Kovar1 1 Children’s Cancer Research Institute, St Anna Kinderkrebsforschung, Vienna, Austria and 2Department of Biology, Hong Kong University of Science and Technology, Kowloon, Hong Kong, SAR China

The oncogene EWS-FLI1 encodes a chimeric transcription factor expressed in Ewing’s sarcoma family tumors (ESFTs). EWS-FLI1 target gene expression is thought to drive ESFT pathogenesis and, therefore, inhibition of EWS-FLI1 activity holds high therapeutic promise. As the activity of many transcription factors is regulated by posttranslational modifications, we studied the presence of modifications on EWS-FLI1. The immuno-purified fusion-protein was recognized by an antibody specific for O-linked b-N-acetylglucosaminylation, and bound readily to a phosphoprotein-specific dye. Inhibition of Ser/Thrspecific phophatases increased EWS-FLI1 molecular weight and reduced its O-GlcNAc content, suggesting that phosphorylation and O-GlcNAcylation of EWSFLI1 interact dynamically. By mutation analysis, O-GlcNAcylation was delineated to Ser/Thr residues of the amino-terminal EWS transcriptional-activation domain. Metabolic inhibition of the hexosamine biosynthetic pathway abrogated O-GlcNAcylation of EWSFLI1 and interfered specifically with transcriptional activation of the EWS-FLI1 target Id2. These results suggest that drugs modulating glycosylation of EWSFLI1 interfere functionally with its activity and might, therefore, constitute promising additions to the current ESFT chemotherapy. Oncogene (2009) 28, 1280–1284; doi:10.1038/onc.2008.484; published online 19 January 2009 Keywords: EWS-FLI1; Ewing’s sarcoma; post-translational modification; O-GlcNAc; phosphorylation; transcription

Post-translational modifications (PTMs) regulate the subcellular localization, stability and activity of most proteins. PTMs of transcription factors interact frequently with each other in a spatially and temporally ordered manner, and serve the integration of intra- and extra-cellular signals in the regulation of gene expresCorrespondence: Professor H Kovar, Children’s Cancer Research Institute, St Anna Kinderkrebsforschung, Kinderspitalgasse 6, A1090 Vienna, Austria. E-mail: [email protected] Received 12 August 2008; revised 1 December 2008; accepted 9 December 2008; published online 19 January 2009

sion. Mutations or polymorphisms in the coding region of a gene may affect sites of PTM, resulting in altered and potentially pathogenic protein activity. De novo combination of unrelated protein domains in fusionproteins, as a consequence of chromosomal rearrangements, subject the activity of individual partner domains to the mutual effects of their PTMs and may, thus, affect their regulation and function. The translocation t(11;22)(q24;q12) fuses EWS and FLI1 genes resulting in a chimeric transcription factor that drives the pathogenesis of Ewing’s sarcoma family tumors (ESFTs) (Kovar, 2005). Modifications of EWS have been described at domains replaced by FLI1 in the fusion-protein. EWS is phosphorylated by protein kinase C within the IQ domain interfering with the RNA-binding activity of EWS (Deloulme et al., 1997). The EWS C-terminal RGG boxes are dimethylated extensively, resulting in cytoplasmic retention of EWS (Araya et al., 2005). Additionally, cell-cycle-dependent phosphorylation by Bruton’s tyrosine kinase and glycosylation at unknown sites have been described for EWS (Guinamard et al., 1997; Matsuoka et al., 2002; Wells et al., 2002). FLI1 is phosphorylated on serine residues by a calcium-dependent process, though modification sites have not been mapped (Zhang and Watson, 2005). So far, nothing is known about PTMs of the oncogenic EWS-FLI1 protein. However, several lines of indirect evidence suggest that PTMs may be involved in specific EWS–FLI1 protein interactions. EWS and EWSFLI1, despite sharing identical N-terminal domain (NTD) primary structure, differ in their ability to interact with the RNA Pol II subunit hsRPB7 (Petermann et al., 1998) and with antibodies directed to the N-terminus of the protein (Aryee et al., 2006), which might be explained by distinct higher-order structure and specific PTMs of the fusion-protein. Also, although able to dimerize, EWSFLI1 binds to DNA as a monomer (Spahn et al., 2003), similar to EWS-WT1 in desmoplastic small-round-cell tumors, in which oligomerization and DNA binding of EWS-WT1 is regulated by phosphorylation (Kim et al., 1999). Thus, it is likely that the function of EWS-fusionproteins is regulated by cellular signaling mechanisms. As the pharmacological targeting of nuclear-proteins is difficult, and knowledge about the type and physiological consequences of their PTMs may unravel upstream signaling molecules as potential therapeutic targets, we investigated PTMs of EWS-FLI1.

Post-translational modification of EWS-FLI1 R Bachmaier et al

1281

The EWS-FLI1 NTD (amino-acids 1–265) shares distant homology with the CTD (C-terminal domain) of eukaryotic RNA polymerase II (RNA Pol II.) (Delattre et al., 1992). Heptapeptide repeats >50 of RNA Pol II CTD with the consensus YSPTSPS appear in EWSFLI1 as 31 degenerate hexapeptide repeats with the consensus sequence SYGQQS. As the RNA Pol II CTD repeats are modified alternatively by phosphorylation on serines 2 and 5 of the repeats and, in the absence of phosphorylation, by O-GlcNAc on threonine and serine residues (Kelly et al., 1993), we hypothesized that the EWS-NTD in EWS-FLI1 may be subject to a similar interplay between O-GlcNAc modification and phosphorylation. EWS-FLI1 immunoprecipitated from the ESFT cell lines SK-N-MC (Figure 1a) and STA-ET-7.2 (not +

-

GlcNAcase

62

-FLI1 Ab

62

-O-GlcNAcAb

flag-FLI1 + flag-EWS-FLI1 83

+

+

1

2

3

4

flag-EWS-FLI1 flag-FLI1

* -FLI1 -O-GlcNAc

6

7

-FLI Ab

-flag Ab

+

62

5

62

-O-GlcNAcAb

*

80 56 47 30 25 80 56 47 30 25

Figure 1 EWS-FLI1 is modified at N-terminal Ser/Thr residues by O-GlcNAc. (a) EWS-FLI1 reactivity with O-GlcNAc-specific antibody RL2 is sensitive to treatment with GlcNAcase. EWSFLI1 immunoprecipitated from SK-N-MC cell extracts using antiFLI1 C-19 antibody (Santa Cruz, CA, USA) was either mocktreated () or incubated with GlcNAcase ( þ ) (Calbiochem, La Jolla, CA, USA), and probed on the Western blot with either antiFLI1 monoclonal antibody 7.3 or RL2 (Abcam, Cambridge, UK). (b) Almost all EWS-FLI1 is modified by O-GlcNAcylation. SK-NMC extracts were incubated with either FLI1 C-19 antibody or irrelevant isotype-matched antibody coupled to anti-rabbit Dynabeads M-280 (Dynal Biotech ASA, Oslo, Norway). Upon C-19 precipitation, no detectable EWS-FLI1 remained in the supernatant. Isolated complexes were eluted from the precipitation matrix, and the entire eluate was subjected to a second round of immunoprecipitation using RL2. Precipitates and supernatants were probed with anti-FLI1 7.3 monoclonal antibody (gift from Olivier Delattre). Lane 1: input (SK-N-MC extract); lane 2: equivalent amount of flow-through of first precipitation; lane 3: EWS-FLI1 immunoprecipitation; lane 4: immunoprecipitation with irrelevant antibody; lane 5: supernatant from precipitation with irrelevant antibody; lane 6: immunoprecipitated EWS-FLI1 eluted and subjected to O-GlcNAc precipitation; lane 7: entire supernatant of O-GlcNAc precipitation. (c) FLI1 is not O-GlcNAcylated. SJ-Nb-7 cells were transfected with expression constructs for flag-FLI1 (lanes 1 and 3) or flag-EWS-FLI1 (lanes 2 and 4), and total protein extracts were incubated with Dynabeadcoupled antibody C-19. Immuno-complexes were probed with antibodies 7.3 and RL2. *Immunoglobulin band from immunoprecipitation. (d) O-GlcNAcylation of the EWS-NTD occurs at Ser/Thr residues. Extracts of SJ-NB-7 cells transfected with either full-length flag-tagged EWS-FLI1 (EF) or flag-EWS-NTD constructs encoding alanine substitutions of EWS N-terminal Gln residues (QA) or of Ser/Thr residues (STA) (Ng et al., 2007) were immunoprecipitated with anti-flag antibody and probed with antiflag and RL2 antibodies. Positions of specific bands are indicated by arrowheads. *Antibody band from immunoprecipitation.

shown) was recognized readily by antibody RL2 which detects specifically single O-GlcNAc moieties on proteins (Snow et al., 1987). Treatment of purified EWSFLI1 protein with b-N-acetylglucosaminidase (GlcNAcase) abolished RL2 reactivity without affecting recognition by the FLI1 antibody 7.3 (Figure 1a). To evaluate the proportion of EWS-FLI1 carrying O-GlcNAc modifications, immunoprecipitated EWS-FLI1 was subjected to a second round of immunoprecipitation with RL2 and, for control, unspecific antibody, and precipitates and supernatants were probed for EWS-FLI1 protein content. Almost no FLI1 antibody-reactive EWS-FLI1 remained in the supernatant after immunoprecipitation with RL2, whereas no EWS-FLI1 precipitated with the irrelevant antibody (Figure 1b). This result suggests that most EWS-FLI1 proteins in ESFT cells is O-GlcNAc modified. By contrast, intact FLI1 lacked reactivity with RL2, consistent with the modification being confined to the EWS-NTD (Figure 1c). Mutation of residues predicted to serve as the most probable O-GlcNAc acceptor sites by ‘YinOYangmotif’ search (Gupta and Brunak, 2002) (T185A, S202A, S239A, T244A), alone or in combination, did not alter reactivity of EWS-FLI1 with RL2 antibody (data not shown). However, combined conversion of Ser residues 40, 51, 69, 87, 111, 162, 168, 171 and of Thr residues 8, 22, 32, 48, 64, 79, 95 and 120, but not of all conserved glutamine residues in the EWS-NTD to alanine (Ng et al., 2007), abolished completely reactivity with the O-GlcNAc-specific antibody RL2 (Figure 1d). Variations in RL2 reactivity of the different EWS-NTD constructs as compared with full-length EWS-FLI1 result likely from structure-based differences of the proteins in accessibility for either RL2 antibody, GlcNAc transferase, or GlcNAcase, despite retaining full transcriptional activity (Ng et al., 2007). These results suggest that O-GlcNAc modification of EWSFLI1 is confined to serine–threonine residues within the EWS transactivation domain. For RNA Pol II CTD, an intimate interplay of O-GlcNAc modification and phosphorylation has been shown. These two modifications occur on identical amino-acid residues and appear to be mutually exclusive and regulating each other (Hart et al., 2007). Immunoprecipitated EWS-FLI1 was probed with Pro-Q Diamond phosphoprotein reagent on an SDS polyacrylamide-gel (Martin et al., 2003). Figure 2a shows that EWS-FLI1 stained readily with this dye. The precipitating FLI1 antibody-band containing abundant protein and BSA stained negative with the phosphoprotein-dye indicating the specificity of the EWS-FLI1 signal. In addition, pre-treatment of immunoprecipitated EWS-FLI1 with alkaline phosphatase reduced staining with the phopsphoprotein dye (Figure 2b). This result identifies EWS-FLI1 as a phosphoprotein. Treatment of SK-N-MC cells with the serine–threonine phosphatase-inhibitor ocadaic acid increased EWS-FLI1 molecular weight and reduced reactivity with RL2 antibody, suggesting that forcing EWS-FLI1 phosphorylation reduces O-GlcNAc modification of EWS-FLI1 in line with a Oncogene


62

Protein stain

62 * 45 Phospho stain

Protein stain 0

60

90

120 min OA EWS-FLI1

62 -FLI1 Ab

EWS-FLI1

62 -GlcNAcAb

DON (mM) Glucosamine (mM) GlcNAcstatin (mM) STZ (mM)

-

0 -

Oncogene

2 -

20 -

0.02 -

5

-O-GlcNAcAb -FLI1 Ab

62

extracts

175

IPs

dynamic interplay between these two modifications on the EWS-FLI1 fusion-protein (Figure 2c). The fact that both the higher and the lower molecular weight bands observed at 120 min of ocadaic-acid treatment still retain some reactivity with the RL2 antibody, indicates that hyperphosphorylation does not abrogate completely GlcNAc modification of EWS-FLI1. No effect on EWS-FLI1 mobility and RL2 reactivity was observed upon treating SK-N-MC cells with the tyrosine phosphatase-inihibitor orthovanadate (data not shown). These results suggest N-terminal O-GlcNAcylation of EWS-FLI1 communicating with Ser/Thr phosphorylation. b-N-acetylglucosamine is transferred to proteins by O-GlcNAc-transferase from UDP-GlcNAc, which is generated by the hexosamine biosynthetic pathway. Inhibition of glutamine:fructose-6-phosphate amidotransferase by the glutamine analog 6-diazo-5-oxo-Lnorleucine (DON) inhibits UDP-GlcNAc biosynthesis and reduces consequently protein O-GlcNAcylation, whereas inhibition of cellular O-GlcNAcase by GlcNAcstatin or Streptozotocin (STZ) stabilizes protein glyco-

0.1 0.1 0.5 -

62

---

Figure 2 Cross-talk between O-GlcNAc modification and phosphorylation of EWS-FLI1. (a) Evidence for phosphorylation of EWS-FLI1 by staining with Pro-Q Diamond phosphoprotein stain. Immunoprecipitated EWS-FLI1 from SK-N-MC cells was resolved on an 8% acrylamid gel and stained with Pro-Q Diamond phosphoprotein stain (Molecular Probes, Invitrogen, Lofer, Austria) (left) and subsequently with SYPRO ruby protein stain (Molecular Probes) (right). 45 kD ovalbumin and the 63 kD BSA marker bands served as positive and negative phosphoprotein controls. *Immunoglobuline band from immunoprecipitation. (b) Treatment of immunoprecipitated EWS-FLI1 with 20 U of calf intestinal phosphatase (CIP, New England Biolabs, Frankfurt, Germany) for 1 h at 37 1C reduced reactivity with Pro-Q Diamond phosphoprotein stain. (c) Treatment of SK-N-MC cells with the Ser/Thr phosphatase inhibitor ocadaic acid (OA) increases the molecular weight and reduces O-GlcNAcylation of EWS-FLI1. Cells were treated with 0.25 mM OA for the indicated times. Immunoprecipitated EWS-FLI1 was probed with antibodies 7.3 and RL2. Note the occurrence of a second EWS-FLI1 band and the loss of RL2 reactivity of the original EWS-FLI1 band after 120 min of OA treatment.

STZ

Phospho stain

DON

62

---

EWS-FLI1

+

STZ

CIP (20U) -

sylation. Treating SK-N-MC cells with DON abrogated EWS-FLI1 O-GlcNAcylation, whereas addition of increasing amounts of exogenous glucosamine restored EWS-FLI1 O-GlcNAcylation (Figure 3a). In contrast, GlcNAcstatin and STZ increased moderately RL2 reactivity of the fusion-protein (Figures 3a and 4a). Loss of O-GlcNAc modification was accompanied by a moderate decrease in overall protein levels of EWSFLI1, which was restored by addition of exogenous glucosamine. O-GlcNAcylation has been reported to play a role in the subcellular distribution of several transcription factors (Hart et al., 2007). Therefore, we tested whether modulation of O-GlcNAcylation in ESFT cells with DON and STZ alters the subcellular localization of EWS-FLI1 (Figure 3b). Nuclear PARP and cytoplasmic tubulin were used as surrogate markers for purity of subcellular fractions. In untreated cells, most EWSFLI1 was found in the nuclear extract, but also the cytosolic fraction contained low amounts of the fusion-protein. DON treatment reduced the size and abundancy of EWS-FLI1, but did not change its proportional distribution to nuclear and cytosolic fractions. Conversely, STZ treatment increased slightly the EWS-FLI1 abundancy in both the nuclear and the cytoplasmic compartments. Consequently, the level of O-GlcNAcylation does not affect the subcellular localization of EWS-FLI1.

DON

EWS-FLI1 IP

Marker

EWS-FLI1 IP

Marker

1282

Inhibitor PARP

83 62

Tubulin

EWS-FLI1

62 cytosol

nuclei

Figure 3 O-GlcNAc modification does not change the subcellular distribution of EWS-FLI1. (a) O-GlcNAc modification of EWSFLI1 is sensitive to inhibitors of the hexosamine biosynthetic pathway. SK-N-MC cells were either left untreated (lane 1) or incubated overnight with 0.1 mM 6-diazo-5-oxo-L-norleucine, DON (Sigma-Aldrich, St Louis, MO, USA) (lanes 2–6) in the absence or presence of increasing amounts of glucosamine as indicated, 0.02 mM GlcNAcstatin (gift of DMF van Aalten, University of Dundee, Scotland, UK) (lane 7), or 5 mM Streptozotocin, STZ (Sigma-Aldrich) (lane 8), and probed with either RL2 or 7.3 antibody. (b) Subcellular distribution of EWS-FLI1. SK-NMC cells were either left untreated or treated with 100 mM DON or 5 mM STZ overnight. Nuclear and cytoplasmic extracts were prepared, and the purity of subcellular fractions and homogenous loading were probed with anti-PARP (Pharmingen-BD, San Diego, CA, USA) and anti-a-tubulin (Calbiochem, San Diego, CA, USA) antibodies. EWS-FLI1 subcellular localization was monitored by probing immunoprecipitated EWS-FLI1 with antibody 7.3 and densitometric analysis.


1283 160 140 120 100 80 60 40 20 0

% of mock treated

% of mock treated

EWS-FLI1 protein levels

350 300 250 200 150 100 50 0 -50

0h 3h 6h 9h 12h 16h 16h DON STZ

-8

1 *

*

*

-12 -16

* 3h 6h 9h 12h 16h 16h DON STZ

4 fold change

fold change

0 -4

0h 3h 6h 9h 12h 16h 16h DON STZ

ID2 expression levels SK-N-MC SU-DHL-1

4

16 h DON

EWS-FLI1 O-GlcNAc levels

0

TGFRII expression levels SK-N-MC SU-DHL-1 1

-4 -8 -12 -16 3h 6h 9h 12h 16h 16h DON STZ

16h DON

Figure 4 O-GlcNAc modification alters the transcriptional activity of EWS-FLI1. Changes in EWS-FLI1 protein expression (a) and O-GlcNAcylation levels (normalized to protein levels) (b) upon 6-diazo-5-oxo-L-norleucine (DON) and Streptozotocin (STZ) treatment of SK-N-MC cells for the indicated times were monitored by densitometric scanning of Western blots probed with antibodies 7.3 and RL2, respectively. Id2 (c) and TGFbRII expressions (d) were measured by Q-PCR and normalized to b-2microglobulin RNA and to EWS-FLI1 protein expression at the indicated time points. As an EWS-FLI1 negative control, the cell line SU-DHL-2 was used. Changes in expression are shown as the fold change deduced from the difference between Ct-values of inhibitortreated and untreated cells. Columns and error bars represent mean values and s.d. of 4 and 3 independent experiments for SK-N-MC and SU-DHL-2 cells each performed in triplicate, respectively. Fold changes in expression were found to be significantly different from zero by Student’s t-test for Id2 at time points 2 h (P ¼ 0.026), 6 h (P ¼ 0.009), 12 h (P ¼ 0.027) and 16 h (P ¼ 0.046) as indicated by the asterisk. Primer and probe sequences for Q-PCR are available upon request.

Id2 and TGFbRII are activated directly and suppressed (Kovar, 2005), respectively, by EWS-FLI1. Therefore, we asked whether O-GlcNAcylation affects the transcriptional activity of EWS-FLI1 by studying Id2 and TGFbRII expressions after DON- and STZtreatment of SK-N-MC cells. As an EWS-FLI1-negative control, we used the anaplastic large cell lymphoma cell line SU-DHL-2. In SK-N-MC cells, EWS-FLI1 protein levels were reduced to about 50% within 16 h of DON treatment, but remained unchanged by incubation with STZ (Figure 4a). During this period, DON led to an almost complete disappearance of O-GlcNAc modification on EWS-FLI1, whereas STZ increased RL2 reactivity by about 50% (Figure 1b). Id2 and TGFbRII RNA expressions were measured by Q-PCR and were normalized to EWS-FLI1 protein levels to account for effects of inhibitor treatment on general gene expression. Interestingly, inhibition of O-GlcNAcylation in SK-NMC cells resulted in an up to 10-fold decrease of Id2 expression within 16 h, whereas stabilization of O-GlcNAc modifications did not change Id2 RNA levels when compared with mock-treated cells (Figure 4c). In contrast, TGFbRII suppression by EWS-FLI1 was not affected by either of the two inhibitors (Figure 4d). However, in EWS-FLI1-negative SU-DHL-2 cells, both Id2 and TGFbRII levels remained unchanged largely. These results suggest that O-GlcNAcylation specifically supports the transcription-activating function of EWS-FLI1.

This is the first report describing PTMs of the transcriptional-activation domain of EWS-FLI1 in ESFT and its functional consequences. We identified phosphorylation and O-linked glycosylation of this protein domain. In an earlier study (Wang et al., 1999), it was found that inhibitors of N-linked glycosylation and of the mevalonate pathway impacts on the expression and function of EWS-FLI1. However, under their experimental conditions, the addition of single dynamic O-linked N-acetylglucosamine modification was not addressed and it was concluded that EWSFLI1 is not a glycoprotein. PTMs had previously been shown to regulate the transcriptional activity of other EWS-fusion-proteins. For EWS-WT1, tyrosine-phosphorylation by the c-ABL kinase modulates dimerization and DNA-binding activity (Kim et al., 1999). In addition, the protein tyrosine kinase v-SRC has been shown to interact with the EWS-NTD modulating the transcriptional activity of EWS-WT1 (Kim et al., 2000). For both kinases, the sites of phosphorylation have not been mapped. In contrast, phosphorylation by protein kinase C modulating the transcriptional activity of rare EWS-ATF and EWS-ETS fusion-proteins that retain EWS exons 7 and 8, has been delineated to the aminoacid residue 266, which is not contained in the majority of EWS-FLI1 fusion-proteins (Ohlson et al., 2005). Here, we provide evidence that EWS-FLI1 is phosphorylated generally and O-GlcNAcylated in ESFT. As ocadaic acid, but not orthovanadate, affected the Oncogene


1284

electrophoretic mobility and the interaction with O-GlcNAcylation of EWS-FLI1, the fusion-protein appears to be phosphorylated at Ser/Thr residues. The sites of phosphorylation overlap likely with those of O-GlcNAcylation. Our results map O-GlcNAc modifications to the EWS transcriptional-activation domain. The EWS-NTD is disordered intrinsically and overall protein structure, but not specific peptide sequences or individual repeats, has been reported to drive the transcriptional activity of EWS fusion-proteins (Ng et al., 2007). For O-GlcNAcylation, no consensus sequence has been defined, and it is possible that due to the disordered structure of the EWS-NTD, this modification may occur at any serine or threonine residue dependent on the three-dimensional protein structure or by a stochastic process. Although the presence of multiple tyrosine residues distributed over the NTD was shown to be essential for the EWS transcriptional-activation function, our results using the glutamine:fructose-6-phosphate amidotransferase inhibitor DON suggest that glycosylation of serine–

threonine residues might be equally important. We found that nearly all EWS-FLI1 is O-GlcNAcylated, suggesting this type of modification as an intrinsic characteristic of EWS-FLI1. The demonstration that inhibition of EWS-FLI1 glycosylation by DON impairs severely transcriptional EWS-FLI1 target gene activation, which is thought to mediate EWS-FLI1-driven oncogenesis, gives hope that this or related drugs may be of therapeutic value in ESFT. DON has been tested in several phase I/II clinical trials in adults and children since 1957 with dose-limiting toxicities (Lynch et al., 1982; Rahman et al., 1985; Sullivan et al., 1988). Our finding, of a potentially specific DON-effect on EWS-FLI1 activity, suggests that DON might be a reasonable addition to polychemotherapy in children with ESFT. Acknowledgements This study was supported in part by Grant P18046-B12 of the Austrian Science Fund (FWF).

References Araya N, Hiraga H, Kako K, Arao Y, Kato S, Fukamizu A. (2005). Transcriptional down-regulation through nuclear exclusion of EWS methylated by PRMT1. Biochem Biophys Res Commun 329: 653–660. Aryee DN, Kreppel M, Bachmaier R, Uren A, Muehlbacher K, Wagner S et al (2006). Single-chain antibodies to the EWS NH(2) terminus structurally discriminate between intact and chimeric EWS in Ewing’s sarcoma and interfere with the transcriptional activity of EWS in vivo. Cancer Res 66: 9862–9869. Delattre O, Zucman J, Plougastel B, Desmaze C, Melot T, Peter M et al (1992). Gene fusion with an ETS DNA-binding domain caused by chromosome translocation in human tumours. Nature 359: 162–165. Deloulme JC, Prichard L, Delattre O, Storm DR. (1997). The prooncoprotein EWS binds calmodulin and is phosphorylated by protein kinase C through an IQ domain (in process citation). J Biol Chem 272: 27369–27377. Guinamard R, Fougereau M, Seckinger P. (1997). The SH3 domain of Bruton’s tyrosine kinase interacts with Vav, Sam68 and EWS. Scand J Immunol 45: 587–595. Gupta R, Brunak S. (2002). Prediction of glycosylation across the human proteome and the correlation to protein function. Pac Symp Biocomput, 310–322. Hart GW, Housley MP, Slawson C. (2007). Cycling of O-linked betaN-acetylglucosamine on nucleocytoplasmic proteins. Nature 446: 1017–1022. Kelly WG, Dahmus ME, Hart GW. (1993). RNA polymerase II is a glycoprotein. Modification of the COOH-terminal domain by O-GlcNAc. J Biol Chem 268: 10416–10424. Kim J, Lee JM, Branton PE, Pelletier J. (1999). Modification of EWS/ WT1 functional properties by phosphorylation. Proc Natl Acad Sci USA 96: 14300–14305. Kim J, Lee JM, Branton PE, Pelletier J. (2000). Modulation of EWS/ WT1 activity by the v-Src protein tyrosine kinase. FEBS Lett 474: 121–128. Kovar H. (2005). Context matters: the hen or egg problem in Ewing’s sarcoma. Semin Cancer Biol 15: 189–196. Lynch G, Kemeny N, Casper E. (1982). Phase II evaluation of DON (6-diazo-5-oxo-L-norleucine) in patients with advanced colorectal carcinoma. Am J Clin Oncol 5: 541–543. Martin K, Steinberg TH, Cooley LA, Gee KR, Beechem JM, Patton WF. (2003). Quantitative analysis of protein phosphorylation status Oncogene

and protein kinase activity on microarrays using a novel fluorescent phosphorylation sensor dye. Proteomics 3: 1244–1255. Matsuoka Y, Matsuoka Y, Shibata S, Yasuhara N, Yoneda Y. (2002). Identification of Ewing’s sarcoma gene product as a glycoprotein using a monoclonal antibody that recognizes an immunodeterminant containing O-linked N-acetylglucosamine moiety. Hybrid Hybridomics 21: 233–236. Ng KP, Potikyan G, Savene RO, Denny CT, Uversky VN, Lee KA. (2007). Multiple aromatic side chains within a disordered structure are critical for transcription and transforming activity of EWS family oncoproteins. Proc Natl Acad Sci USA 104: 479–484. Ohlson J, Enstero M, Sjoberg BM, Ohman M. (2005). A method to find tissue-specific novel sites of selective adenosine deamination. Nucleic Acids Res 33: e167. Petermann R, Mossier BM, Aryee DN, Khazak V, Golemis EA, Kovar H. (1998). Oncogenic EWS-Fli1 interacts with hsRPB7, a subunit of human RNA polymerase II (In Process Citation). Oncogene 17: 603–610. Rahman A, Smith FP, Luc PT, Woolley PV. (1985). Phase I study and clinical pharmacology of 6-diazo-5-oxo-L-norleucine (DON). Invest New Drugs 3: 369–374. Snow CM, Senior A, Gerace L. (1987). Monoclonal antibodies identify a group of nuclear pore complex glycoproteins. J Cell Biol 104: 1143–1156. Spahn L, Siligan C, Bachmaier R, Schmid JA, Aryee DN, Kovar H. (2003). Homotypic and heterotypic interactions of EWS, FLI1 and their oncogenic fusion protein. Oncogene 22: 6819–6829. Sullivan MP, Nelson JA, Feldman S, Van Nguyen B. (1988). Pharmacokinetic and phase I study of intravenous DON (6-diazo5-oxo-L-norleucine) in children. Cancer Chemother Pharmacol 21: 78–84. Wang M, Xie Y, Girnita L, Nilsson G, Dricu A, Wejde J et al. (1999). Regulatory role of mevalonate and N-linked glycosylation in proliferation and expression of the EWS/FLI-1 fusion protein in Ewing’s sarcoma cells. Exp Cell Res 246: 38–46. Wells L, Vosseller K, Cole RN, Cronshaw JM, Matunis MJ, Hart GW. (2002). Mapping sites of O-GlcNAc modification using affinity tags for serine and threonine post-translational modifications. Mol Cell Proteomics 1: 791–804. Zhang XK, Watson DK. (2005). The FLI-1 transcription factor is a short-lived phosphoprotein in T cells. J Biochem (Tokyo) 137: 297–302.