Tài liệu Báo cáo khoa học: Identification of Ewing’s sarcoma protein as a G-quadruplex DNA- and RNA-binding protein ppt

Identification of Ewing’s sarcoma protein as a
G-quadruplex DNA- and RNA-binding protein
Kentaro Takahama
1,
*, Katsuhito Kino
2,
*, Shigeki Arai
3
, Riki Kurokawa
3
and Takanori Oyoshi
1
1 Department of Chemistry, Faculty of Science, Shizuoka University, Japan
2 Kagawa School of Pharmaceutical Sciences, Tokushima Bunri University, Kagawa, Japan
3 Division of Gene Structure and Function, Saitama Medical University Research Center for Genomic Medicine, Japan
Introduction
The current knowledge of Ewing’s sarcoma (EWS)
derives primarily from studies of a group of dominant
oncogenes that arise due to chromosomal transloca-
tions in which EWS is fused to a variety of cellular
transcription factors [1–3]. EWS fusion proteins are
very potent transcription activators that depend on the
EWS N-terminal domain and a C-terminal DNA-bind-
ing domain contributed by the fusion partner [4–9].
EWS ⁄ ATF1 is a potent constitutive activator of ATF-
dependent promoters [10]. The EWS N-terminal binds
directly to the RNA polymerase II subunit hsRPB7
and this interaction is thought to be important for
transactivation [11].
In contrast to EWS fusion proteins, however, the
normal function and the nucleic acid-binding proper-

ties of EWS remain poorly characterized. EWS belongs
to a family that includes the closely related proteins
translocated in liposarcoma and the TATA-binding
protein-associated factor 15 which are involved in
several aspects of gene expression [12–15]. This protein
family contains the transcriptional activation domain
in the N-terminal region and the RNA-binding domain
Keywords
Ewing’s sarcoma; G-quadruplex DNA;
G-quadruplex RNA; RGG motif; RNA-binding
protein
Correspondence
T. Oyoshi, Department of Chemistry,
Faculty of Science, Graduate School of
Science, Shizuoka University, 836 Oya,
Suruga, Shizuoka 422-8529, Japan
Fax: +81 54 237 3384
Tel: +81 54 238 4760
E-mail:
*These authors contributed equally to this
work
(Received 27 August 2010, revised 23
December 2010, accepted 13 January 2011)
doi:10.1111/j.1742-4658.2011.08020.x
The Ewing’s sarcoma (EWS) oncogene contains an N-terminal transcrip-
tion activation domain and a C-terminal RNA-binding domain. Although
the EWS activation domain is a potent transactivation domain that is
required for the oncogenic activity of several EWS fusion proteins, the nor-
mal role of intact EWS is poorly characterized because little is known
about its nucleic acid recognition specificity. Here we show that the

Arg-Gly-Gly (RGG) domain of the C-terminal in EWS binds to the G-rich
single-stranded DNA and RNA fold in the G-quadruplex structure.
Furthermore, inhibition of DNA polymerase on a template containing a
human telomere sequence in the presence of RGG occurs in an RGG
concentration-dependent manner by the formation of a stabilized G-quad-
ruplex DNA–RGG complex. In addition, mutated RGG containing Lys
residues replacing Arg residues at specific Arg-Gly-Gly sites and RGG con-
taining Arg methylated by protein arginine N-methyltransferase 3 decrease
the binding ability of EWS to G-quadruplex DNA and RNA. These find-
ings suggest that the RGG of EWS binds to G-quadruplex DNA and
RNA via the Arg residues in it.
Abbreviations
dsHtelo, human telomere duplex DNA; EAD, Ewing’s sarcoma activation domain; EMSA, electrophoretic mobility shift assay; ETS, external
transcribed spacer; EWS, Ewing’s sarcoma; FMRP, fragile X mental retardation protein; GST, glutathione S-transferase; Htelo, human
telomere DNA; mut Htelo, mutated human telomere; mut rHtelo, mutated human telomere RNA; PRMT3, protein arginine
N-methyltransferase 3; RBD, RNA-binding domain; rHtelo, human telomere RNA; RRM, RNA recognition motif; ZnF, zinc finger.
988 FEBS Journal 278 (2011) 988–998 ª 2011 The Authors Journal compilation ª 2011 FEBS
(RBD) in the C-terminal region as multiple domains
involved in nucleic acid–protein interactions: an RNA
recognition motif (RRM) flanked by two regions in
Arg-Gly-Gly repeats (RGG) and a C
2
C
2
zinc finger
(ZnF) with an RGG domain in the C-terminal [16,17].
They bind to RNA as well as single- and double-
stranded DNA [18–20]. In the case of EWS, the
C-terminal amino acids that constitute RGG specifi-
cally bind to poly G and poly U RNA in vitro [8]. On

the other hand, Hume et al. [15] suggested that EWS
binds to the proximal-element DNA of the macro-
phage-specific promoter of the CSF-1 receptor gene.
The RGG domain, initially identified as a single-
stranded RNA-binding motif in hnRNP U, is reported
to be the G-quadruplex RNA-binding motif in the
fragile X mental retardation protein (FMRP) and the
G-quadruplex DNA-binding motif in nucleolin [21–
27]. The RGG domain of FMRP, which is an RNA-
binding protein involved in nerve cell differentiation,
interacts with the G-quartet forming RNA [22–25]. In
addition, the RBD and the RGG domain of nucleolin,
a DNA-binding protein contributing to the transcrip-
tion of ribosomal RNA, bind to the G-quadruplex
forming ribosomal DNA [26]. Moreover, nucleolin
binds to the c-myc G-quadruplex DNA with high
affinity in vitro [27]. Little is known, however, about
the DNA and RNA recognition specificity of EWS,
which contains three RGG domains. To gain further
insight into the nucleic acid–EWS interaction, we per-
formed an electrophoretic mobility shift assay (EMSA)
with EWS and several G-quadruplex or single- or dou-
ble-stranded DNA and RNA. Here, we show that
EWS specifically targets G-quadruplex DNA and
RNA in vitro. We also determined that the specificity
of G-quadruplex recognition depends on the guanidini-
um group of the Arg in the RGG domain in the
C-terminal of EWS.
Results and Discussion
Several DNA-binding proteins that bind to G-quadru-

plex DNA have been investigated in vitro [28–36].
Hanakahi et al. [26] reported that the four RBD and
the Arg-Gly-Gly repeats of nucleolin, which is
involved in transcription, rRNA processing and ribo-
some assembly, can bind to G-quadruplex DNA
formed from the external transcribed spacer region of
human rDNA, ETS-1. We performed an EMSA of
EWS and ETS-1 to investigate the ability of EWS to
bind to G-quadruplex DNA (Fig. 1A, Table 1).
Recombinant EWS, which contains RBD in the C-ter-
minal region comprising RRM, ZnF and three RGG
(RGG1, RGG2 and RGG3) for binding to nucleic
acids, was expressed in Escherichia coli as proteins
fused to glutathione S-transferase (GST) and purified
using glutathione agarose.
32
P-labeled ETS-1 was first
incubated for 24 h in 100 mm KCl to allow for quad-
ruplex formation and then with GST-tag-digested
EWS for 1 h at room temperature. The EWS–DNA
complexes were resolved by 6% PAGE and visualized
by autoradiography. Binding analyses revealed that
EWS binds to the G-quadruplex formed from the
ETS-1, but not to the control single-stranded DNA.
Similar results were obtained with a human telomere
DNA (Htelo) in the presence of 100 mm K
+
(Fig. 1B,
Table 1). The results demonstrated that EWS binds to
Htelo, but not to human telomere duplex DNA (dsHt-

elo), in the presence of K
+
. The results of previous
studies indicated that Htelo in a K
+
ion-containing
solution exists as an equilibrium G-quadruplex forma-
tion of some antiparallel form of the hybrid paral-
lel ⁄ antiparallel (3 + 1) form together with the parallel
propeller form and the basket type [37–42]. These find-
ings indicate that EWS binds to G-quadruplex DNAs
formed from different synthetic oligonucleotides and
thus appears to recognize the G-quadruplex DNA
structure independently of the sequence context.
We further investigated the region of EWS that con-
tributes to the G-quadruplex binding specificity by
–
–
EWS
EWS
ssDNA L
ETS-1
ssDNA S
Htelo
dsHtelo
BA
–
–
EWS
EWS

–
EWS
Fig. 1. Affinity of EWS for binding to G-quadruplex DNA. (A) EMSA
was performed with EWS (lanes 2 and 4) and
32
P-labeled ETS-1
(lanes 3 and 4) or ssDNA L (lanes 1 and 2). (B) EMSA was per-
formed with EWS (lanes 2, 4 and 6) and
32
P-labeled Htelo (lanes 3
and 4), dsHtelo (lanes 5 and 6) or ssDNA S (lanes 1 and 2). The
structures of DNAs used as probes are indicated above each lane.
The DNA–protein complexes were resolved by 6% PAGE and visu-
alized by autoradiography.
K. Takahama et al. Identification of Ewing’s sarcoma protein
FEBS Journal 278 (2011) 988–998 ª 2011 The Authors Journal compilation ª 2011 FEBS 989
comparing the behavior of various mutant recombi-
nant proteins, i.e. the EWS activation domain (EAD),
RGG1, RRM–RGG2–ZnF and RGG3, with regard to
Htelo (Fig. 2A). RGG3 interacted with Htelo in
EMSA, whereas the proteins containing EAD, RGG1
and RRM–RGG2–ZnF did not bind to Htelo
(Fig. 2B). The RGG domain in FMRP has a closely
spaced Arg-Gly-Gly repeat, which is necessary for
G-quadruplex structure binding [23,24]. RGG3,
containing 12 RGG repeats, of EWS binds to the
G-quadruplex structure, whereas RGG1, containing
six fewer RGG repeats than RGG3, does not
(Table 2). Additional binding studies demonstrated
that recombinant RGG3 does not bind to dsHtelo or

single-stranded DNA (Fig. S1, Table 1). These studies
revealed that RGG3 of EWS binds mainly to the
G-quadruplex.
To test whether formation of the G-quadruplex is
necessary for RGG3 binding, we assayed the binding
of RGG3 with Htelo in the presence of K
+
or Li
+
.
EAD
RGG1
RRM-RGG2-ZnF
RGG3
EWS
EAD
RGG1 RRM ZnF
B
EAD
RGG3
–
RGG1
RRM-RGG2-ZnF
C
1 287 347 469 501 544 656
A
Htelo
mut Htelo
RGG3
–

RGG3
–
D
E
RGG2
RGG3
dsHtelo – – 1x 10x 100x
F
Htelo – – 1x 10x 100x
RBD
RGG3
RGG3
Full
length
Primer
RGG3
Pausing
product
Fig. 2. Structural features of EWS and
DNA-binding specificities of RGG3. (A) Sche-
matic representation of the deletion mutants
constructed to map the Htelo-binding speci-
ficity of each one of the EWS. AD (residues
1–287); RGG 1 (288–347); RRM (residues
348–469); RGG 2 (residues 450–501); ZnF
(residues 502–544); RGG 3 (residues 545–
656). (B) DNA-binding activities of EAD (lane
2), RGG1 (lane 3), RRM–RGG2–ZnF (lane 4)
and RGG3 (lane 5). EMSA was performed
with these proteins and

32
P-labeled Htelo.
(C) EMSA was performed with RGG3 (lanes
2 and 4) and
32
P-labeled Htelo (lanes 1 and
2) or mut Htelo (lanes 3 and 4). (D, E) Bind-
ing competition assay, assaying binding of
RGG3 to
32
P-labeled Htelo in the presence
of unlabeled dsHtelo (D) or Htelo (E) at the
indicated molar ratios of unlabeled ⁄ labeled
DNA. The DNA–protein complexes were
resolved by 6% PAGE and visualized by
autoradiography. (F) DNA polymerase arrest
assays. Primer extension reactions were
performed with rTaq DNA polymerase. The
primer, full-length primer extension products
and DNA polymerase arrest products are
indicated by arrows. Extension through the
template after incubation in increasing con-
centrations of RGG3. The concentrations of
RGG3 were 0 l
M (lane 1), 0.2 lM (lane 2),
0.5 l
M (lane 3) and 1 lM (lane 4).
Identification of Ewing’s sarcoma protein K. Takahama et al.
990 FEBS Journal 278 (2011) 988–998 ª 2011 The Authors Journal compilation ª 2011 FEBS
Figures S1 and S2 show that RGG3 binding to Htelo

in the presence of Li
+
, which did not form the
G-quadruplex as confirmed by CD spectroscopy, was
blocked. To further test whether RGG3 bound to
Htelo folds into a G-quadruplex, we analyzed the
binding between RGG3 and a mutated human telo-
mere (mut Htelo) that replaces G with T at positions 9
and 15, which destabilized the G-quadruplex forma-
tion, as confirmed by CD and UV spectroscopy
(Figs 2C, S2, Table 1). The analysis showed that
RGG3 binds to the folded Htelo G-quadruplex, but
not to unfolded mut Htelo despite containing one
TTAGGG sequence. Furthermore, competitive experi-
ments performed in the presence of cold competitor
Htelo or dsHtelo showed that Htelo effectively com-
peted for binding, whereas dsHtelo had no effect, even
at a 100-fold molar excess (Fig. 2D, E). These findings
suggest that RGG3 binds to G-quadruplex DNA with
structure specificity.
Having found that EWS binds to G-quadruplex con-
formations and not to single- and double-stranded
conformations by RGG3, we aimed to determine
whether RGG3 of EWS modulates the formation or
unwinding of Htelo G-quadruplex DNA. To determine
whether the RGG3 binding affected the stability of the
G-quadruplex structure of Htelo, we performed a poly-
merase stop assay as described previously [43]. The
32
P-labeled 25-mer primer annealed to the 3¢ end of

the template and could be extended by a DNA poly-
merase upon the addition of the dNTPs. If complete
extension of the primer occurred, a full-length 76-mer
product would be formed. Factors that promote and
stabilize intramolecular G-quadruplex formation,
however, led to a specific pausing site on the template.
This assay showed that the stopping site corresponded
to the base located 3¢ to the first guanine base involved
in G-quadruplex formation (Fig. 2F, lane 1). More-
over, as the RGG3 protein concentration increased,
the full-length 76-mer product decreased, and the stop-
ping site product increased (Fig. 2F, lanes 2–4). Thus,
these results indicate that RGG3 binds to and stabi-
lizes the folded G-quadruplex formation.
To test whether RGG3 contributes not only to the
G-quadruplex DNA binding, but also to G-quadruplex
RNA binding, we assayed the binding of RGG3 with
a human telomere RNA (rHtelo) in the presence of
K
+
, which exists as a G-quadruplex formation of the
parallel propeller form [44]. Figure 3(A, B) shows that
RGG3 bound to rHtelo in the presence of K
+
,
whereas binding between RGG3 and a mutated human
telomere RNA (mut rHtelo) that replaces G with T at
positions 9 and 15 destabilized the G-quadruplex for-
mation, as confirmed by CD and UV spectroscopy
(Fig. S2, Table 1). Furthermore, competitive experi-

ments performed in the presence of cold competitor
rHtelo or mut rHtelo showed that rHtelo effectively
competed for binding, whereas the mut rHtelo had no
effect, even at a 100-fold molar excess (Fig. 3C, D).
These findings suggest that RGG3 also binds to
G-quadruplex RNA with structure specificity.
To elucidate the ability of RGG3 to bind the G-quad-
ruplex, various concentrations of RGG3 were incu-
bated with 5¢
32
P-labeled Htelo or rHtelo in a K
+
solution. As the RGG3 concentration increased, the
free DNA or RNA decreased, and the higher molecu-
lar weight complex increased (Fig. 4). The mobility
shift data were fitted to a hyperbolic equation to give
a K
d
of 13 ± 3 nm (Htelo) and 10 ± 2 nm (rHtelo).
In comparison with RGG3, the full-length EWS and
RBD containing RGG3 bound to Htelo with
Table 1. Sequence of oligonucleotides used in EMSA and CD spectroscopy. Oligonucleotides were diluted to 0.5 mM (base concentration)
in 50 m
M Tris ⁄ HCl (pH 7.5) in the presence of 100 mM KCl or 100 mM LiCl, as specified. Duplex annealing or quadruplex formation was
performed by heating samples to 95 °C on a thermal heating block and cooling to 4 °C at a rate of 2 °CÆmin
–1
.
Name Sequence
ssDNAS d(CATTCCCACCGGGACCACCAC)
ssDNA L d(CATTCCCACCGGGACCACCACCATTCCCACCGGGACCACCAC)

ETS-1 d(TCTCTCGGTGGCCGGGGCTCGTCGGGGTTTTGGGTCCGTCC)
Htelo d[AGGG(TTAGGG)
3
]
dsHtelo d[AGGG(TTAGGG)
3
]⁄ d[(CCCTAA)
3
CCCT]
mut Htelo d[AGGG(TTAGTG)
2
TTAGGGJ
rHtelo r(UUAGGG)
4
mut rHtelo r[UUAGGG(UUAGUG)
2
UUAGGG]
Table 2. Amino acid sequences of RGG1 and RGG3.
RGG1 PGENRSMSGPDNRGRGRGGFDRGGMSRGGRGGGRGGMG
SAGERGGFNKPGGPMDEGPDLDLGPPVDP
RGG3 APKPEGFLPPPFPPPGGDRGRGGPGGMRGGRGGLMDRGGP
GGMFRGGRGGDRGGFRGGRGMDRGGFGGGRRGGPGGPP
GPLMEQMGGRRGGRGGPGKMDKGEHRQERRDRPY
K. Takahama et al. Identification of Ewing’s sarcoma protein
FEBS Journal 278 (2011) 988–998 ª 2011 The Authors Journal compilation ª 2011 FEBS 991
K
d
=30±5 and 14±3nm, respectively (Fig. S3).
The EAD domain therefore inhibited the high-affinity
Htelo binding of RGG3 and RBD. The ability of

RGG3 and RBD to repress transcription activation by
EAD raised the possibility that RGG3 and RBD block
the interaction between the EAD and RNA polymer-
ase II subunit [11,45]. The interaction between the
EAD and RGG3 might inhibit the high-affinity Htelo
binding of RGG3.
To gain further insight into the induction of G-quad-
ruplex formations by RGG3 of EWS, we performed a
CD spectroscopic analysis that was conducted with
Htelo in the presence of various amounts of RGG3.
The CD spectrum of Htelo, a hybrid (3 + 1) form,
showed a strong positive band at 290 nm and a nega-
tive band at around 235 nm, whereas the addition of 1
ratio excess of RGG3 led to an increase in ellipticity
and shifted the spectrum from a strong positive band
to 265 nm (Fig. 5), which is characteristic of the paral-
lel form and consistent with the results of previous CD
studies [40–42]. These data indicate that RGG3 binds
to the Htelo G-quadruplex and changes the hybrid
(3 + 1) G-quadruplex formation of Htelo. Moreover,
it may provide a model showing the change from the
hybrid (3 + 1) G-quadruplex to the parallel form with
the association of RGG3. Incubation of rHtelo with
RGG3 did not alter the G-quadruplex RNA, however,
as demonstrated by CD spectrum analysis (data not
shown).
Rajpurohit et al. [46] reported that binding of the
recombinant hnRNP A1 protein to single-stranded
nucleic acid is reduced upon enzyme methylation of
Arg. To evaluate the role of Arg in RGG3 on G-quad-

ruplex DNA recognition, we performed EMSA using
Htelo with RGG3 methylated by protein arginine
N-methyltransferase 3 (PRMT3) (Fig. 6). In vitro
methylation of the recombinant EWS with PRMT3
showed that PRMT3 is responsible for the asymmetric
dimethylations of specific Arg in the RGG region [47].
In our study, the methylation of the RGG3 by
PRMT3 with [
3
H]AdoMet as a methyl donor was
monitored with a liquid scintillation counter (Fig. S4).
RGG3-methylated Arg did not bind to Htelo (Fig. 6,
lane 6), whereas PRMT3 and AdoMet did not inhibit
the Htelo binding of RGG3 (Fig. 6, lanes 1–4). Simi-
larly, RGG3-methylated Arg did not bind to rHtelo,
whereas PRMT3 and AdoMet did not inhibit the rHt-
elo binding of RGG3 (Fig. S4). These results indicate
that enzyme methylation of Arg reduces the binding of
RGG3 to G-quadruplex DNA or RNA.
Previous results demonstrated that nine Arg are
potential methylation sites within RGG3 that react
with PRMT3 [47]. We next created mutated RGG3 to
precisely define the residues within RGG3 that bind to
G-quadruplex DNA (Fig. 6B). Simultaneous substitu-
tion of Arg by Lys in two (KGG3-2) Arg within
RGG3 reduced G-quadruplex DNA binding and in six
(KGG3-6) and four Arg (KGG3-4) within RGG3,
eliminated G-quadruplex DNA binding despite the
basic nature of the Lys side-chain (Fig. 6C). Similarly,
KGG3-2, KGG3-4 and KGG3-6 reduced G-quadru-

plex RNA binding (Fig. S5). These findings indicate
RGG3
Htelo
rHtelo
RGG3
–
–
B
A
rHtelo
mut
rHtelo
–
–
RGG3
RGG3
rHtelo – – 1x 10x 100x
RGG3
D
C
mut
rHtelo – – 1x 10x 100x
RGG3
Fig. 3. Protein–nucleic acid complexes. (A) EMSA was performed
with RGG3 (lanes 2 and 4) and
32
P-labeled Htelo (lanes 1 and 2) or
rHtelo (lanes 3 and 4). (B) EMSA was performed with RGG3 (lanes
2 and 4) and
32

P-labeled rHtelo (lanes 1 and 2) or mut rHtelo (lanes
3 and 4). (C, D) Binding competition assay assaying binding of
RGG3 to
32
P-labeled rHtelo in the presence of unlabeled rHtelo (C)
or mut rHtelo (D) at the indicated molar ratios of unlabeled ⁄ labeled
DNA. The DNA–protein complexes were resolved by 6% PAGE
and visualized by autoradiography.
Identification of Ewing’s sarcoma protein K. Takahama et al.
992 FEBS Journal 278 (2011) 988–998 ª 2011 The Authors Journal compilation ª 2011 FEBS
that Arg between amino acids 589 and 597 within
RGG3 are important for the binding of RGG3 to
G-quadruplex DNA and RNA. In ssDNA and ssRNA
recognition, the methylation of Arg in a peptide or a
protein does not affect the binding strength [25,48,49].
Methylated RGG3 of EWS inhibited G-quadruplex
Htelo binding, but was able to bind mut Htelo
(Fig. S6). These findings indicate the importance of the
guanidinium group of the Arg in RGG3 for binding to
the G-quadruplex.
In conclusion, EWS appears to be a DNA- and
RNA-binding protein that recognizes the G-quadru-
plex structure. It remains unclear, however, whether
the role of EWS in transcription or other functions is
determined by its ability to target a specific DNA and
RNA structure. Rossow & Janknecht [50] reported
that overexpression of EWS in RK13 and AKR cells
leads to the activation of the c-fos, Xvent-2 and ErbB2
promoters, indicating that EWS functions as a
transcriptional cofactor. EWS, however, has not been

reported to bind to double-stranded DNA in these
promoters. The c-fos and ErbB2 promoters contain
G-rich sequences that could potentially form G-quad-
ruplex structures [51,52]. On the basis of a combina-
tion of in silico and experimental approaches, Verma
et al. [53,54] reported an enriched sequence with the
potential to adopt the G-quadruplex motifs near tran-
scription start sites. These findings suggest that
G-quadruplex motif-mediated regulation is a more
common mode of transcription control. On the other
hand, Dejardin & Kingston [55] purified human
telomeric chromatin using proteomics of isolated
RGG3 (nM) 0 4 8 16 32 64 125 250
B
A
RGG3 (nM) 0 4 8 16 32 64 125 250
RGG3 concentration (n
M)
0.6
0.4
0.2
0
0 50 100 150 200 250
RGG3 concentration (nM)
0 50 100 150 200 250
0.8
1
0.6
0.4
0.2

0
0.8
1
Fraction of RNA bound
Fraction of DNA bound
Fig. 4. Binding affinity of RGG3 to Htelo or
rHtelo. The DNA or RNA concentration was
fixed at 1 n
M, whereas the concentration of
RNase-treated RGG3 added to the binding
reaction was varied, as shown above each
lane. The equilibrium-binding curve was
obtained by calculating the fraction of Htelo
(A) or rHtelo (B) bound at varying RGG3 con-
centrations. K
d
was determined by fitting to
the equation (see Materials and methods).
The DNA–protein complexes were
resolved by 6% PAGE and visualized by
autoradiography.
Wavelength (nm)
Molar ellipticity
(10
6
· mdeg ·
M
–1
· cm
–1

)
240
260
280
300
320
–2
0
2
4
6
220
Hybrid (3 + 1) form
Parallel form
Fig. 5. CD of Htelo in the presence of various amounts of RGG3.
Titration of Htelo with RGG3 (1.2, 1.0, 0.8, 0.6, 0.4, 0.2, 0.1 and 0
equiv. of RGG3) in 100 m
M KCl and 50 mM Tris ⁄ HCl (pH 7.5). The
concentration of DNA was 0.2 m
M base concentration. Line colors:
black = 0 equiv.; blue = 0.1 equiv.; cyan = 0.2 equiv.; green = 0.4
equiv.; light green = 0.6 equiv.; yellow = 0.8 equiv.; orange = 1.0
equiv.; red = 1.2 equiv.
K. Takahama et al. Identification of Ewing’s sarcoma protein
FEBS Journal 278 (2011) 988–998 ª 2011 The Authors Journal compilation ª 2011 FEBS 993
chromatin segments and identified that the protein
translocated in liposarcoma, which is related to EWS
as a subgroup within the RNP family of RNA-binding
proteins containing RRM and RGG domains, binds to
telomeres. Further studies are required to identify the

role of EWS and the possible function of such
G-quadruplex structures in genomic DNA.
Materials and methods
Preparation, expression and purification of GST
fusion proteins
The EWS cDNA was cloned into the pGEX6P-1 vector
between the EcoRI and XhoI sites for expression as an
N-terminal GST fusion protein (pGEX–EWS). pGEX–EAD,
pGEX–RGG1, pGEX–RRM–RGG2–ZnF and pGEX–
RGG3 vectors contain a PCR encoding EWS amino acids 1–
287, 288–347, 348–544 and 545–656, respectively, cloned in
pGEX6P-1 using the following sets of primers: EWS forward
d(CGG AAT TCA TGG CGT CCA CGG ATT ACA G)
and EWS reverse d(CGC TCG AGT CAC TAG TAG GGC
CGA TCT CTG C), for pGEX–EWS; EAD forward d(CGG
AAT TCA TGG CGT CCA CGG ATT ACA G) and EAD
reverse d(CGC TCG AGT CAT CCG GAA AAT CCT
CCA GAC T), for pGEX–EAD; RGG1 forward d(CGG
AAT TCC CAG GAG AGA ACC GGA GCA T) and
RGG1 reverse d(CGC TCG AGT CAA TCA AGA TCT
GGT CCT TCA TCC ATG G), for pGEX–RGG1; RRM–
RGG2–ZnF forward d(CGG AAT TCC TAG GCC CAC
CTG TAG ATC C) and RRM–RGG2–ZnF reverse d(CGC
TCG AGT CAC TTA CAC TGG TTG CAC TCT GTT
CTC C), for pGEX–RRM–RGG2–ZnF; and RGG3 forward
d(CGG AAT TCG CCC CAA AGC CTG AAG GCT T)
and RGG3 reverse d(CGC TCG AGT CAC TAG TAG
GGC CGA TCT CTG C), for pGEX–RGG3. pGEX–
KGG3-2, pGEX–KGG3-4 and pGEX–KGG3-6 were
obtained by replacing Arg with Lys in pGEX–RGG3 using a

KOD -Plus- mutagenesis kit (Toyobo, Japan). To construct
pGEX–KGG3-2, PCR was performed with pGEX–RGG3 as
a template and the following primers: KGG3-2 forward
d(AAA GGT GGC AAA GGT GGA GAC AGA GGT
GGC TT) and KGG3-2 reverse d(GAA CAT TCC ACC
GGG ACC ACC AC). pGEX–KGG3-4 was generated by
PCR using pGEX–KGG2 as a template and the following
primers: KGG3-4 forward d(AGA CAA AGG TGG CTT
CAA AGG TGG CCG) and KGG3-4 reverse d(CCA CCT
TTG CCA CCT TTG AAC A). PCR was conducted with
pGEX–KGG3-4 as a template and the following primers:
KGG3-6 forward d(GGC AAA GGC ATG GAC AAA
GGT GGC TTT GG) and KGG3-6 reverse d(ACC TTT
GAA GCC ACC TTT GTC TCC ACC), for pGEX–KGG3-
6. All reactions were performed according to the manu-
facturer’s protocol for the KOD-Plus- mutagenesis kit
(Toyobo). Escherichia coli strain BL21 (DE3) pLysS-
competent cells were transformed with the vectors, and the
transformants were grown at 37 °C in a Luria Bertani
medium containing ampicillin (0.1 mgÆmL
)1
). Protein expres-
sion was induced at A
600
= 0.6 with 0.1 mm isopropyl
b-d-1-thiogalactopyranoside. The cells were then grown for
an additional 16 h at 25 °C and harvested by centrifugation
(6400 g for 20 min). Pellets were resuspended in buffer A
(100 mm Tris ⁄ HCl pH 7.5, 150 mm NaCl, 1 mm EDTA acid
and 1 mm dithiothreitol) and lysed by sonication (model

UR-20P, Tomy Seiko, Tokyo, Japan) at 4 °C. The super-
natants containing the expressed proteins were centrifuged
for 15 min at 16 200 g at 4 °C, and the proteins were then
purified by glutathione agarose (Sigma, St Louis, MO,
USA). The supernatant and glutathione agarose were incu-
bated with gentle mixing for 1 h at 4 °C; resin was washed
with buffer A at 4 °C. Proteins were eluted with buffer B
(50 mm Tris ⁄ HCl pH 9.5, 20 mm reduced glutathione and
1mm dithiothreitol). Buffer B of the elution was exchanged
RGG3
PRMT3
AdoMet
+ + – – + +
– + – + – +
– – + + + +
MF RGG RGG D RGG F RGG RGMD RGG F
MF KGG KGG D RGG F RGG RGMD RGG F
MF KGG KGG D KGG F KGG RGMD RGG F
MF KGG KGG D KGG F KGG KGMD KGG F
RGG3
KGG3-2
KGG3-4
KGG3-6
RGG3
RGG3
KGG3-2
KGG3-4
KGG3-6
–
B

545 656
587 610
AC
Fig. 6. Identification of significant residues at RGG3 for G-quadruplex binding ability. (A) Ability of RGG3 to bind to Htelo in the presence (+)
or in the absence ()) of PRMT3 or AdoMet. RGG3 (lanes 2, 4 and 6) was incubated with (lanes 1, 2, 5 and 6) or without (lanes 3 and 4)
PRMT3 in a potassium buffer with (lanes 3–6) or without (lanes 1 and 2) AdoMet. (B) Schematic illustration of amino acids 587–610 within
RGG3 (residues 545–656). The point mutations are shown in bold. (C) EMSA of RGG3 (lane 2), KGG3-2 (lane 3), KGG3-4 (lane 4) and KGG3-
6 (lane 5) using Htelo. The DNA–protein complexes were resolved by 6% PAGE and visualized by autoradiography.
Identification of Ewing’s sarcoma protein K. Takahama et al.
994 FEBS Journal 278 (2011) 988–998 ª 2011 The Authors Journal compilation ª 2011 FEBS
with buffer C (50 mm Tris ⁄ HCl pH 7.5, 100 mm KCl and
1mm dithiothreitol) or buffer D (50 mm Tris ⁄ HCl pH 7.5,
100 mm LiCl and 1 mm dithiothreitol) by dialysis. The
protein concentrations were determined using the BCA
Protein Assay Kit (Thermo Scientific, USA). For all experi-
ments, GST-tag was digested according to the manu-
facturer’s instructions (GE Healthcare, Precision Protease,
Little Chalfont, UK), and 20 nmol of each protein was
incubated with 20 lg RNase A (Nippon Gene, Tokyo,
Japan) at 4 °C for 16 h before use.
EMSA
Labeled oligonucleotides were diluted to 0.2 mm (base con-
centration) in 50 mm Tris ⁄ HCl (pH 7.5) in the presence of
100 mm KCl or 100 mm LiCl, as specified. Duplex anneal-
ing or quadruplex formation was performed by heating
samples to 95 °C on a thermal heating block and cooling to
4 °C at a rate of 2 °CÆmin
)1
. Binding reactions were
performed in a final volume of 20 lL using 100 fmol of

the labeled oligonucleotide and a varying concentration (0–
2.5 lm) of purified proteins in a binding buffer (50 mm
Tris ⁄ HCl pH 7.5, 0.5 mm EDTA, 0.5 mm dithiothreitol,
0.1 mgÆmL
)1
bovine serum albumin, 1 lgÆmL
)1
calf thymus
DNA and 100 mm KCl or 100 mm LiCl). After the samples
were incubated for 1 h at 25 °C, they were loaded on a 6%
polyacrylamide (acrylamide ⁄ bisacrylamide = 19 : 1) nonde-
naturing gel; 0.5· TBE with 20 m m KCl was used, both in
the gel and as the electrophoresis buffer. Electrophoresis
was performed at 10 V Æ cm
)1
for 1 h at 4 °C. The gels were
exposed in a phosphorimager cassette and imaged (Personal
Molecular Imager FX; Bio-Rad, Hercules, CA, USA).
Bands were quantified using imagequant software. The
data were plotted as u (1 fraction of free DNA) versus the
protein concentration to determine the K
d
, which is equal to
the protein at which half of the free DNA is bound. K
d
were
extracted by nonlinear regression using Microsoft Excel
2007 and the following equation: u = [P] ⁄ {K
d
+ [P]}.

DNA polymerase stop assay
This assay was adapted from the method described by
Han et al. [43]. The 25-mer primer was 5¢-labeled with
32
P, mixed with the 76-mer template DNA and annealed
as described above. The polymerase reaction was per-
formed in a final volume of 20 lL using 20 fmol of the
duplex and various amounts of purified RGG3 in a bind-
ing buffer (50 mm Tris ⁄ HCl pH 7.5, 1 mm dithiothreitol,
100 lgÆmL
)1
bovine serum albumin, 1 lgÆmL
)1
calf thy-
mus DNA and 100 mm KCl). RGG3 was incubated with
the duplex for 1 h at room temperature. The polymerase
extension reaction was initiated by adding Taq polymer-
ase, dNTP (1 mm each) and MgCl
2
(10 mm). The reaction
was incubated at 30 °C for 10 min and then stopped by
adding an equal volume of a stop buffer (95% formamide,
10 mm EDTA, 10 mm NaOH, 0.1% bromophenol blue
and 0.1% xylenecyanol). Extension products were sepa-
rated on a 12% polyacrylamide (acrylamide ⁄ bisacryla-
mide = 19 : 1) gel; 1· TBE was used, both in the gel and
as the electrophoresis buffer. Electrophoresis was per-
formed at 1500 V for 1 h at 4 °C, and gels were visualized
on a phosphorimager.
CD spectroscopy

CD spectra were recorded on a CD spectrometer model
J-500A (Jasco). The CD spectra of Htelo (0.2 mm base con-
centration) and RGG3 (1.2, 1.0, 0.8, 0.6, 0.4, 0.2, 0.1, 0)
equivalent to Htelo DNA (RGG3 ⁄ DNA) in 50 mm
Tris ⁄ HCl (pH 7.5) and 100 mm KCl were recorded using a
0.2 cm path length cell at 25 °C. The spectra of the
Htelo–RGG3 complex were corrected by subtracting the
spectra of the free RGG3 at the same ratios.
Methylation of recombinant RGG3
This assay was adapted from the method described by Geh-
ring et al. [47]. RGG3 was incubated with PRMT3 and
AdoMet in a final volume of 50 lL with 50 mm Tris ⁄ HCl
(pH 7.5), 100 mm KCl, 1 mm EDTA and 1 mm dithiothrei-
tol for 3 h at 30 °C. The reaction solution was exchanged
with a potassium buffer containing 50 mm Tris ⁄ HCl (pH
7.5), 100 mm KCl and 1 mm dithiothreitol by dialysis.
Acknowledgements
This research was supported by the Sasakawa Scientific
Research Grant from The Japan Science Society and
a Grant-in-Aid for Young Scientists (B) (2008,
20750130) from the Ministry of Education, Science,
Sports, and Culture of Japan. We thank Dr Harvey R.
Herschman at UCLA for the PRMT3 cDNA.
References
1 Ron D (1997) TLS-CHOP and the role of RNA-bind-
ing proteins in oncogenic transformation. Curr Top
Microbiol Immunol 220, 131–142.
2 May WA & Denny CT (1997) Biology of EWS ⁄ FLI
and related fusion genes in Ewing’s sarcoma and primi-
tive neuroectodermal tumor. Curr Top Microbiol

Immunol 220, 143–150.
3 Rauscher FJ III (1997) Chromosome translocation-
mediated conversion of a tumor suppressor gene into a
dominant oncogene: fusion of EWS1 to WT1 in
desmoplastic small round cell tumors. Curr Top
Microbiol Immunol 220, 151–162.
4 Brown AD, Lopez-Terrada D, Denny CT & Lee KAW
(1995) Promoters containing ATF-binding sites are
K. Takahama et al. Identification of Ewing’s sarcoma protein
FEBS Journal 278 (2011) 988–998 ª 2011 The Authors Journal compilation ª 2011 FEBS 995
de-regulated in cells that express the EWS ⁄ ATF1
ontogeny. Oncogene 10, 1749–1756.
5 Fujimura Y, Ohno T, Siddique H, Lee L, Rao VN &
Reddy ESP (1996) The EWS-ATF-1 gene involved in
malignant melanoma of soft parts with t(12;22)
chromosome translocation, encodes a constitutive
transcriptional activator. Oncogene 12, 159–167.
6 May WA, Lessnick SL, Braun BS, Klemsz M,
Lewis BC, Lunsford LB, Hromas R & Denny CT
(1993) The Ewing’s sarcoma EWS ⁄ FLI-1 fusion gene
encodes a more potent transcriptional activator and is a
more powerful transforming gene than FLI-1. Mol Cell
Biol 13, 7393–7398.
7 Ohno T, Rao VN & Reddy ESP (1993) EWS ⁄ Fli-1 chi-
meric protein is a transcriptional activator. Cancer Res
53, 5859–5863.
8 Ohno T, Ouchida M, Lee L, Gatalica Z, Rao VN &
Reddy ES (1994) The EWS gene, involved in Ewing
family of tumors, malignant melanoma of soft parts
and desmoplastic small round cell tumors, codes for an

RNA binding protein with novel regulatory domains.
Oncogene 9, 3087–3097.
9 Sanchez-Garcia I & Rabbitts TH (1994) Transcrip-
tional activation by TAL1 and FUS-CHOP proteins
in acute malignancies as a result of chromosomal
abnormalities. Proc Natl Acad Sci USA 91,
7869–7873.
10 Pan S, Koh YM, Dunn TA, Li KKC & Lee KAW
(1998) The EWS ⁄ ATF1 fusion protein contains a
dispersed activation domain that functions directly.
Oncogene 16, 1625–1631.
11 Petermann R, Mossier BM, Aryee DN, Khazak V,
Golemis EA & Kover H (1998) Oncogenic EWS-Fli1
interacts with hsRPB7, a subunit of human RNA poly-
merase II. Oncogene 17, 603–610.
12 Wang X, Arai S, Song X, Reichart D, Du K, Pascual
G, Tempst P, Rosenfeld MG, Glass CK & Kurokawa
R (2008) Induced ncRNAs allosterically modify RNA-
binding proteins in cis to inhibit transcription. Nature
454, 126–131.
13 Hicks GG, Singh N, Nashabo A, Mai S, Bozek G,
Klewes L, Arapovic D, White EK, Koury MJ, Oltz EM
et al. (2000) Fos deficiency in mice results in defective
B-lymphocyte development and activation, high levels
of chromosomal instability and perinatal death. Nature
24, 175–179.
14 Araya N, Hirota K, Shimamoto Y, Miyagishi M, Yosh-
ida E, Ishida J, Kaneko S, Kaneko M, Nakajima T &
Fukamizu A (2003) Cooperative interaction of EWS
with CREB-binding protein selectively activate hepato-

cyte nuclear factor 4-mediated transcription. J Biol
Chem 278, 5427–5432.
15 Hume DA, Sasmono T, Himes SR, Sharma SM,
Bronisz A, Constantin M, Ostrowski MC & Ross IL
(2008) The Ewing sarcoma protein (EWS) binds directly
to the proximal elements of the macrophage-specific
promoter of the CSF-1 receptor (csf1r) gene. J Immunol
180, 6733–6742.
16 Delattre O, Zucman J, Plougastel B, Desmaze C, Melot
T, Peter M, Kovar H, Joubert I, de Jong P, Rouleau G
et al. (1992) Gene fusion with an ETS DNA-binding
caused by chromosome translocation in human
tumours. Nature 359, 162–166.
17 Li KKC & Lee KAW (2000) Transcription activation
by the Ewing’s Sarcoma (EWS) oncogene can be
cis-repression by the EWS RNA-binding domain. J Biol
Chem 275, 23053–23058.
18 Perrotti D, Bonatti S, Trotta R, Martinez R, Skorski T,
Salomoni P, Grassilli E, Iozzo RV, Copper DR &
Calabretta B (1998) TLS ⁄ FUS, a pro-oncogene
involved in multiple chromosomal translocations, is a
novel regulator of BCR ⁄ ABL-mediated leukemogenesis.
EMBO J
15, 4442–4455.
19 Bertolotti A, Lutz Y, Heard DJ, Chambon P & Tora L
(1996) hTAF
II
68, a novel RNA ⁄ ssDNA the pro-onco-
proteins TLS ⁄ FUS and EWS is associated with both
TFIID and RNA polymerase II. EMBO J 15, 5022–

5031.
20 Crozat A, Aman P, Mandahl N & David R (1993)
Fusion of CHOP to a novel RNA-binding protein in
human myxoid liposarcoma. Nature 363, 640–644.
21 Kiledjian M & Dreyfuss G (1992) Primary structure
and binding activity of the hnRNP U protein:
binding RNA through RGG box. EMBO J 11, 2655–
2665.
22 Darnell JC, Jensen KB, Jin P, Brown V, Warren ST &
Darnell RB (2001) Fragile X mental retardation protein
targets G quartet mRNAs important for neuronal func-
tion. Cell 107, 489–499.
23 Zanotti KJ, Lackey PE, Evans GL & Mihailescu MR
(2006) Thermodynamics of the fragile X mental retarda-
tion protein RGG box interactions with G quartet
forming RNA. Biochemistry 45, 8319–8330.
24 Menon L & Mihailescu M-R (2007) Interactions of the
G quartet forming semaphorin 3F RNA with the RGG
box domain of the fragile X protein family. Nucleic
Acids Res 35, 5379–5392.
25 Ramos A, Hollingworth D & Pastore A (2003) G-quar-
tet-dependent recognition between the FMRP RGG
box and RNA. RNA 9, 1198–1207.
26 Hanakahi LA, Sun H & Maizels N (1999) High affinity
interactions of nucleolin with G-G-paired rDNA. J Biol
Chem 274, 15908–15912.
27 Gonzalez V, Guo K, Hurley L & Sun D (2009) Identifi-
cation and characterization of Nucleolin as a c-myc
G-quadruplex-binding protein. J Biol Chem 284, 23622–
23635.

28 Fang G & Cech TR (1993) The b subunit of Oxytricha
telomere-binding protein promotes G-quartet formation
by telomeric DNA. Cell 74, 875–885.
Identification of Ewing’s sarcoma protein K. Takahama et al.
996 FEBS Journal 278 (2011) 988–998 ª 2011 The Authors Journal compilation ª 2011 FEBS
29 Fukuda H, Katahira M, Tsuchiya N, Enokizono Y,
Sugimura T, Nagano M & Nakagama H (2002) Unfold-
ing of quadruplex structure in the G-rich strand of the
minisatellite repeat by the binding protein UP1. Proc
Natl Acad Sci USA 99, 12685–12690.
30 Etzioni S, Yafe A, Khateb S, Weisman-Shomer P,
Bengal E & Fry M (2005) Homodimeric MyoD
preferentially binds tetraplex structures of regula-
tory muscle-specific genes. J Biol Chem 280,
26805–26812.
31 Zaug AJ, Podell ER & Cech TR (2005) Human POT1
disrupts telomeric G-quadruplexes allowing telomerase
extension in vitro. Proc Natl Acad Sci USA 102, 10864–
10869.
32 Salas TR, Petruseva I, Lavrik O, Bourdoncle A,
Mergny J-L, Favre A & Saintome C (2006) Human
replication protein A unfolds telomeric G-quadruplexes.
Nucleic Acids Res 34, 4857–4865.
33 Thakur RK, Kumar P, Halder K, Verma A, Kar A,
Paret J-L, Basundra R, Kumar A & Chowdhury S
(2009) Metastases suppressor NM23-H2 inter-
action with G-quadruplex DNA within c-MYC
promoter nuclease hypersensitive element induces
c-MYC expression. Nucleic Acids Res 37,
172–183.

34 Paramasivam M, Membrino A, Cogoi S, Fukuda H,
Nakagama H & Xodo LE (2009) Protein hnRNP A1
and its derivative Up1 unfold quadruplex DNA in the
human KRAS promoter: implications for transcription.
Nucleic Acids Res 37, 2841–2853.
35 Soldatenkov VA, Vetcher AA, Duka T & Ladama S
(2008) First evidence of a functional interaction between
DNA quadruplexes and poly(ADP-ribose) polymerase-
1. ACS Chem Biol 3, 214–219.
36 Fry M (2007) Tetraplex DNA and its interacting pro-
teins. Front Biosci 12, 4336–4356.
37 Ambrus A, Chen D, Dai J, Bialis T, Jones RA &
Yang D (2006) Human telomeric sequence forms a
hybrid-type intramolecular G-quadruplex structure with
mixed parallel ⁄ antiparallel strands in potassium solu-
tion. Nucleic Acids Res 34, 2723–2735.
38 Xu Y, Noguchi Y & Sugiyama H (2006) The new
models of the human telomere d[AGGG(TTAGGG)
3
]
in K
+
solution. Bioorg Med Chem 14, 5584–
5591.
39 Luu KN, Phan A, Kuryavyi V, Lacroix L & Patel DJ
(2006) Structure of the human telomere in K
+
solution:
an intramolecular (3 + 1) G-quadruplex scaffold. JAm
Chem Soc 128, 9963–9970.

40 Lim KW, Amrane S, Bouaziz S, Xu W, Mu Y,
Petel DJ, Luu KN & Pjan AT (2009) Structure of the
human telomere in K
+
solution: a stable basket-type
G-quadruplex with only two G-tetrad layers. JAm
Chem Soc 131, 4301–4309.
41 Singh V, Azarkh M, Exner TE, Hartig JS &
Drescher M (2009) Human telomeric quadruplex
conformations studied by pulsed EPR. Angew Chem Int
Ed 48, 9728–9730.
42 Balagurumoorthy P, Brahmachari SK, Mohanty D,
Bansal D & Sasisekharan V (1992) Hairpin and parallel
quartet structures for telomeric sequences. Nucleic Acids
Res 20, 4061–4067.
43 Han H, Hurley LH & Salazar M (1999) A DNA poly-
merase stop assay for G-quadruplex-interactive com-
pounds. Nucleic Acid Res 27, 537–542.
44 Martadinate H & Phan AT (2009) Structure of propel-
ler-type parallel-strand RNA G-quadruplexes, formed
by human telomeric RNA sequence in K solution.
J Am Chem Soc 131, 2570–2578.
45 Alex D & Lee KAW (2005) RGG-boxes of the EWS
oncoprotein repress a range of transcriptional activation
domains. Nucleic Acids Res 33, 1323–1331.
46 Rajpurohit R, Paik WK & Kim S (1994) Effect of enzy-
mic methylation of heterogeneous ribonucleoprotein
particle A1 on its nucleic-acid binding and controlled
proteolysis. Biochem J 304, 903–909.
47 Pahlich S, Bschir K, Chiavi C, Belyanskaya L &

Gehring H (2005) Different methylation characteristics
of protein arginine methyltransferase 1 and 3 toward
the Ewing sarcoma protein and a peptide. Proteins 61,
164–175.
48 Raman B, Guarnaccia C, Nadassy K, Zakhariev S,
Pintar A, Zanuttin F, Frigyes D, Acatrinei C, Vindigni
A, Pongor G et al. (2001) N
x
-arginine dimethylation
modulates the interaction between a Gly ⁄ Arg-rich
peptide from human nucleolin and nucleic acids.
Nucleic Acids Res 29, 3377–3384.
49 Valentini SR, Weiss VH & Silver PA (1999) Arginine
methylation and binding of Hrp1p to the efficiency
element for mRNA 3¢-end formation. RNA 5,
272–280.
50 Rossow KL & Janknecht R (2001) The Ewing’s
sarcoma gene product functions as a transcriptional
activator. Cancer Res 61, 2690–2695.
51 Ishii S, Imamoto F, Yamanashi Y, Toyoshima K &
Yamamoto T (1987) Characterization of the promoter
region of the human c-erbB-2 protooncogene.
Proc Natl Acad Sci USA 84, 4374–4378.
52 Coulon V, Chebli K, Cavelier P & Blanchard JM
(2010) A novel mouse c-fos intronic promoter that
responds to CREB and AP-1 is developmentally
regulated in vivo. PLoS ONE 5, e11235.
53 Verma A, Halder K, Halder R, Yadav VK, Rawal P,
Thakur RK, Mohd F, Sharman A & Chowdhury S
(2008) Genome-wide computational and expression

analyses reveal G-quadruplex DNA motifs as conserved
cis-regulatory elements in human and related species.
J Med Chem 51, 5641–5649.
K. Takahama et al. Identification of Ewing’s sarcoma protein
FEBS Journal 278 (2011) 988–998 ª 2011 The Authors Journal compilation ª 2011 FEBS 997
54 Verma A, Yadav VK, Basundra R, Kumar A &
Chowdhury S (2009) Evidence of genome-wide G4
DNA-mediated gene expression in human cancer cell.
Nucleic Acids Res 37, 4194–4204.
55 Dejardin J & Kingston RE (2009) Purification of pro-
teins associated with specific genomic loci. Cell 136,
175–186.
Supporting information
The following supplementary material is available:
Fig. S1. Affinity of RGG3 for binding to a G-quadru-
plex DNA and RNA.
Fig. S2. CD spectra of DNAs and RNAs.
Fig. S3. Binding affinity of EWS and RBD to
Htelo.
Fig. S4. In vitro arginine methylation of RGG3 by
PRMT3.
Fig. S5. Identification of significant residues at RGG3
for rHtelo binding ability.
Fig. S6. Ability of RGG3 and RGG3 methylated by
PRMT3 to bind to G-quadruplex.
This supplementary material can be found in the
online version of this article.
Please note: As a service to our authors and readers,
this journal provides supporting information supplied
by the authors. Such materials are peer-reviewed and

may be re-organized for online delivery, but are not
copy-edited or typeset. Technical support issues arising
from supporting information (other than missing files)
should be addressed to the authors.
Identification of Ewing’s sarcoma protein K. Takahama et al.
998 FEBS Journal 278 (2011) 988–998 ª 2011 The Authors Journal compilation ª 2011 FEBS